Report · Department for Education · Ask about this document

Independent review into prevalence and support for mental health conditions, ADHD and autism: final report

Independent review into prevalence and support for mental health conditions, ADHD and autism

Final Report

Professor Peter Fonagy (Chair)

Professor Gillian Baird and Professor Sir Simon Wessely (Vice-Chairs)

October 2026

Foreword

A foreword is usually a formality. I would like this to be something else, because I cannot explain why this Review matters without saying something about the people who made it, and something about my own life.

The people first. I have worked alongside our two Vice-Chairs, Professor Gillian Baird and Professor Sir Simon Wessely. Gillian, a paediatrician, has kept the child whose difficulties rarely arrange themselves into the categories we bring to them, at the forefront of our work. Simon, the psychiatric epidemiologist, has kept the numbers and evidence honest, whilst encouraging us to think boldly and of a world beyond our immediate preoccupations. Kate Headington and Charlotte Kume- Holland have led a secretariat that has held together our work; Kate’s standing instruction to us all – start with the person, then talk about the system – has shaped more of this report than a foreword can show.

Behind the four of them stand over 120 colleagues: clinicians, researchers, teachers, commissioners, people who work in the voluntary sector, and people who came to this because of what they or their children have been through. They did not agree with one another. That was the point, and it is the reason I trust what follows. Some of the sharpest disagreements are recorded in these pages rather than smoothed away. We were asked some very plain questions, and much of the disagreement came from discovering that they do not have plain answers. I have chaired a good many things. I have not before been part of a group that disagreed so productively, or so kindly. It has been an honour.

Now the harder part.

I have spent forty years studying how children develop and how development goes awry, and more than twenty leading a national children’s mental health charity. Those are the credentials that put my name on the cover. They are not why I said yes.

I came to England in 1967, fifteen years old, with almost no English, and was lodged with a family in Kew Gardens. Within a year I had stopped wanting to eat or to leave my room. I was severely depressed and I had made decisions about my future that would worry me a great deal if a young person described them to me today. I was at a failing school and couldn’t bridge a connection to the other boys who were solely interested in football.

What saved me was not a system. It was a neighbour who noticed. She sent me to a clinic in Hampstead that Anna Freud had founded for children displaced by war. In remarkable symmetry, decades later, I was asked to lead what had by then become the Anna Freud Centre.

The therapy saved my life. What made it work was not a technique. It was that the person seeing me could look past everything that was wrong and see that there was someone there who might yet do quite well. I have told the story before of arriving at a session in a battered Ford Anglia that I loved beyond reason and babbling about it; she walked to the window, looked out and said, “Peter, that is a wonderful car”. She understood that what I needed was not her insight into my psyche but someone willing to be interested in what I was excited about. That is when I began to believe she might have something worth telling me. Everything I have done in treatment development since is, I now realise, a long footnote to that afternoon.

There is a second debt. I was born with a neurodevelopmental difficulty: severe dysgraphia. My school made nothing of it, to them I was a failure. University College London, where good fortune took me next, allowed my exam papers to be transcribed, years before any law obliged it to make such an adjustment. It is no coincidence that I have worked at that university ever since.

My own difficulty happened to be writing. Over the years I have met countless people whose specific difficulty was different, but who shared the sense of being misunderstood, of working much harder than others realised, of wondering why ordinary things seemed so much more difficult than they appeared to be for everyone else. The Review was asked to focus on mental health, autism and ADHD, but these conditions are part of a much wider landscape of neurodevelopmental differences that includes dysgraphia, speech and language disorders, dyslexia, dyspraxia, learning disability, tic disorders and others. We have not been able to do justice to all of them. Yet the experiences they bring, struggling to be understood, difficulties in school, bullying, exclusion, poorer mental and physical health, and obstacles to participation in everyday life, are often strikingly similar.

At the same time, they are not all the same. Mental health conditions, ADHD and autism overlap in many people’s lives, but they differ in important ways: in how they arise, how they develop, how they affect day-to-day life, and how they are best understood and supported. Each exists across a spectrum of severity and impact, and the overlap and the difference between them together is what makes questions of prevalence, diagnosis and support so difficult. Simple answers are rarely adequate. I hope this Review will encourage people to see mental health and neurodevelopmental conditions as part of a connected developmental picture, in

which early recognition, understanding and evidence-based support matter at least as much as diagnosis itself.

I want to be careful not to turn my own history into a lesson for other people’s children. Life has become harder since I was a young man: more uncertain, more relentlessly observed and appraised, with less belonging and fewer opportunities to take time-out, to recuperate and recover. The cacophony of voices now available to a young person – most of them digital, many of them confident, few of them accountable – has made the ordinary developmental process of understanding oneself far harder than it was for me, and the consequences for how young people function are correspondingly greater. As a clinician who still sees young people, and as a proud father of two children, I am in awe of their courage. Young people today are contending with more powerful forces than I ever faced. They are not fragile. They are outnumbered.

Mental ill health is a feature of being human. No report will change that. But we have learnt a great deal and have much more to offer since that fifteen-year-old arrived in London with few words of English, and we are capable of far more than we currently do.

I know, not only from the research literature but from the inside, that help which arrives in time, is offered with real care, and is matched to what a person needs in that moment, can have effects that feel close to magical. If the recommendations that follow are adopted, my hope is that they will mean more people with mental health and neurodevelopmental needs are met the way mine were so that they may be as grateful, every day, as I am.

Professor Peter Fonagy

Chair, Independent Review into prevalence and support for mental health conditions, ADHD and autism

Executive Summary

  • Our country is experiencing a profound change in the need for support for people with mental health problems and neurodevelopmental conditions.
  • Mental health is part of everyone’s health, not a diagnosis. Mental ill health

includes conditions such as anxiety, depression and eating disorders, which are recognised patterns of thought, feeling or behaviour that cause significant distress or impairment. They can arise at any stage of life, may come and go, and often respond well to treatment and support.

  • Mental distress and common mental health conditions have increased

substantially, particularly among children and young people. The consequences are felt in families, schools, workplaces, communities and the economy. The total annual cost of mental ill health in England has been estimated at around £300 billion[1].

  • Attention deficit hyperactivity disorder (ADHD) and autism are

neurodevelopmental conditions, strongly genetically influenced and present from early life, influencing attention, communication, behaviour, learning and how people experience the world. With the right adjustments many people live full lives; some face substantial difficulties and lifelong high support needs. Traits related to these conditions, including inattention, social difficulties and emotional regulation, occur to varying degrees across the population, but having some of these traits, or identifying as neurodivergent, is not the same as having a diagnosable neurodevelopmental condition. Diagnosis depends not simply on the early developmental presence of traits, but on their severity, persistence and impact on functioning.

  • Recognition of ADHD and autism has risen dramatically, leading to

hundreds of thousands of people awaiting diagnostic assessment and many others struggling to access support. Many people also continue to face barriers to timely treatment, increasing the risk that difficulties escalate and become more severe.

  • Continuing as we are is not sustainable: not for the people waiting weeks,

months or years, becoming more distressed because help is locked behind waiting for a diagnosis; not for those with the most severe needs who cannot access specialist care; and not for a society paying the human and economic

price of failing to support people to participate in family, community, education and working life.

  • The opportunity now is not simply to manage rising demand. It is to

build a better system: reducing unacceptable waits now, while reshaping both clinical and non-clinical support, taking a serious approach to prevention and wellbeing, and improving mental health and neurodevelopmental care within and beyond the NHS, particularly in the community.

  • There should be a fundamental shift from a diagnosis-dependent system

to a needs-led plus diagnosis system. The heated debate about whether there is too much diagnosis or too little risks reinforcing the centrality of diagnosis rather than need and timely access to support and treatment. The government must address two failures at the same time: people wait far too long for support, diagnosis and treatment, and diagnosis has become the primary gateway to support that could and should be available earlier. These failures sit alongside insufficient capacity for people with the most severe, enduring and complex needs.

  • Diagnosis matters but is not without risks. For many people it is clinically

necessary, personally meaningful and essential to treatment. It can help explain and address difficulties that significantly impair a person’s functioning, participation and quality of life. People with significant functional impairment should be able to access diagnostic assessment within a reasonable timeframe, and specialist care should reach those who need it most. Support should be available according to a person’s needs before, during and after assessment, and not be withheld while waiting for a diagnosis. Diagnosis can also do harm when it is inaccurate or used poorly. It can direct someone towards the wrong treatment and obscure co-occurring conditions or the social and environmental causes of their difficulties. A diagnosis can inform practical support people need to remain in education, participate in work, sustain relationships and live well. But those same outcomes can also be delayed or denied if a diagnostic label defines a person by their difficulties rather than by their strengths, circumstances and potential.

  • Increasing participation in relationships, education, work, family and

community life should be a central outcome and measure of success. These are core pillars of good mental health and wellbeing. Success should not be measured by assessments, diagnoses or waiting lists alone, but by increased participation and improved quality of life, recognising that the outcomes that matter most will differ from person to person. For example, the benefits of employment, and the adverse consequences of lack of

employment on mental health, have been known for over a century. The vast majority of individuals with mental health and neurodevelopmental conditions want to work and go to school and are able to, if given appropriate support and reasonable adjustments. It is the job of the NHS and other public services to recognise the importance of engagement in education and work for good health, and their role in providing timely, and comprehensive informed support and treatment to help people to participate. Work, when secure and well supported, offers purpose, structure, financial security and social connection, and can help people recover rather than simply following their recovery.

  • Change requires reform across mental health and neurodevelopmental

care, increased capacity, transformation through digitally enabled care, clearer national standards and sustained investment. It also requires government to tackle the social and economic conditions, and the environmental context, that shape mental health and neurodevelopmental conditions and to measure whether public services are improving people’s lives rather than simply moving them between waiting lists. Some changes can begin immediately through existing powers and programmes; others require deliberate workforce growth, investment and implementation over time.

  • This Executive Summary sets out how we have interpreted the evidence;

what we found about prevalence, need and demand across mental health conditions, ADHD and autism; what is driving these changes; how well the current system responds; and our recommendations for a fairer and more effective response.

Our approach to interpreting the evidence

  • We set out to understand how many people are affected and how this is

changing; what is driving these changes; how they affect different groups; the benefits and risks associated with diagnosis; the growth of the independent sector; the risks of medicalisation; and what works in treatment and prevention. The Review distinguishes between population prevalence, administrative prevalence, demand for assessment and unmet need because these measures answer different questions and should not be treated as interchangeable.

In short: population prevalence is not administrative prevalence; distress is not disorder; diagnosis is not need; and support is broader than treatment.

  • Our conclusions do not rest on any single source. We have drawn on

population surveys, administrative and clinical data, systematic reviews, and testimony from people with lived experience and those delivering services. Where evidence is incomplete, outdated or uncertain, we say so. This is particularly important for ADHD and autism, where population prevalence data remain limited and less up to date than for many other conditions.

  • Because the same words are often used to mean different things, we define

clearly what we mean by distress, need, prevalence, demand, functioning and impairment. We are also candid about the limitations of the available data and what can currently be measured robustly. Functional impairment in particular is inconsistently defined and recorded, so trends in impairment should be interpreted with caution.

Chapters 1–4 provide more detail on the context for the Review and the approach we have taken, and Chapter 5 explores overdiagnosis, misdiagnosis and underdiagnosis in more detail.

Our assessment of the evidence on prevalence, need and service demand

  • Two kinds of data tell different stories. Population prevalence estimates how

many people have a condition, whether or not it has been recognised or recorded. Administrative prevalence records those who have been seen and recorded by services and therefore reflects demand on the system, including waiting lists. Administrative data are essential for planning services but are not measures of population prevalence. Some people who could benefit from support appear in neither.

Mental Health Conditions

  • Mental distress has increased substantially across the population. Both

the population and administrative prevalence of common mental health conditions, particularly anxiety and depression, has risen significantly across all age groups, but at a substantially faster rate for young people. The proportion of adults aged 16-64 years old meeting criteria for a common mental disorder increased by almost half, from 15.5% in 1993 to 22.6% in 2024[2]. Among 16–24-year-olds it increased from around one in six to one in four, with young women now reporting the highest rates of any group. Self- harm and eating disorders have also increased among young adults[3, 4]. The rise in mental health problems is real, is affecting everyday functioning, and cannot be explained by greater awareness alone (Chapter 6).

  • We are not seeing the same increase in severe mental illness. Population

rates of schizophrenia and bipolar disorder, for example, have remained broadly stable[5, 6]. However, too many people with severe, enduring illness still do not receive the treatment and care they need and continue to experience some of the poorest outcomes in society. Improving care for this group, while reducing the risk of harm to themselves and others, must remain a priority.

ADHD

  • The available evidence suggests that the population prevalence of

ADHD has remained broadly stable or increased only modestly over recent decades, although uncertainty remains because England lacks up-to-date population prevalence data. The most recent Mental Health of Children and Young People (MHCYP) survey estimated ADHD prevalence at 3% among 5–15-year-olds in 2017 according to the Diagnostic and Statistical Manual of Mental Disorders, 5th Edition (DSM-5) criteria. We estimate by projecting from the existing MHCYP data points that the current prevalence among 5–15-year-olds might be around 3.8%, but this is at best a rough estimate in the absence of contemporary data based on the modest rise between 2004 and 2017 continuing unchanged. No equivalent adult population survey has been undertaken. Administrative prevalence of ADHD has increased among children and young people aged 6–11 years by more than six-fold (from 0.45% in 2000 to 2.70% in 2025) and among those aged 12–17 years by more than fifteen-fold (from 0.26% in 2000 to 3.88% in 2025)[7]. Socioeconomic disadvantage has long been associated with higher ADHD incidence and prevalence; however, more recent data show the gap between the most and least deprived groups has narrowed (Chapter 7).

Figure Es.1. Adhd recorded administrative prevalence by age band Recorded ADHD prevalence, active 5-year (%) Age group (years) 0-5 6-11 12-17 18-24 25-34 35-44 45-54 55-64 65+ 2000 2005 2010 2015 2020 2025 Calendar year Source: Clinical Practice Research Datalink (CPRD) analysis for the Review.

  • Demand for ADHD assessment has increased dramatically. Monthly

referrals more than tripled, from around 5,000 in 2019 to around 17,000 in 2024. By the end of 2025, open referrals exceeded 560,000, more than seventeen times the number recorded in 2019. Community paediatric waiting lists, the majority relating to ADHD assessment, reached approximately 173,000[8]. Waiting lists have grown rapidly, reflecting not only rising need but a system unable to keep pace with demand. Even when an appropriate diagnosis of ADHD is made, there are still unacceptable delays to support and treatment for those who need it (Chapter 7).

Autism

  • Similarly to ADHD, there is little evidence of a substantial change in

underlying population prevalence, although uncertainty remains because there has been no recent England-wide prevalence study. The best available estimate remains around 1.2% among young people, based on the 2017 MHCYP survey. Again, there is no equivalent adult population survey data. The administrative prevalence of autism has increased among children and young people aged 6–11 years by more than twenty-fold (0.24% in 2000 to 4.58% in 2025) and among those aged 12–17 by more than forty- fold (0.10% in 2000 to 4.06% in 2025)[9, 10].

Figure ES.2. Autism recorded administrative prevalence by age band Recorded ASD prevalence, active 5-year (%) Age group (years) 0-5 6-11 12-17 18-24 25-34 35-44 45-54 55-64 65+ 2000 2005 2010 2015 2020 2025 Calendar year Source: CPRD analysis for the Review.

  • Autism related SEND rose from 1.5% of pupils in 2016 to 3.3% in 2024[11].

Demand for assessment has also risen sharply. By December 2025, more than 254,000 people had an open autism referral, with over 90% waiting longer than 13 weeks[12]. Yet 40% of clinicians surveyed reported no ongoing post-diagnostic support, and only 12.6% described support extending beyond a single session[13] (Chapter 8). National data collection on people with lifelong support needs, autism with severe or profound intellectual disability for example, is required so that broader recognition of autism does not make those requiring the most intensive support less visible.

ADHD and autism

  • As shown above, administrative prevalence for ADHD and autism

among children and young people is now approaching and for some groups already exceeding population prevalence estimates, and is continuing to rise. The administrative prevalence figures also exclude some private-sector assessments that are not captured in routine data. If administrative prevalence were simply converging with underlying population prevalence, rates would be expected gradually to stabilise. Instead, both recorded diagnoses and demand for assessment continue to increase rapidly.

  • This does not, by itself, establish that overdiagnosis is currently

occurring at scale. Population prevalence estimates are dated, use different methods and should not be treated as fixed biological ceilings. However, the evidence suggests we are entering a phase in which the balance of risks between underdiagnosis, misdiagnosis and overdiagnosis may be changing, with the risks of misdiagnosis and overdiagnosis becoming increasingly important considerations for policy and service delivery. To balance this concern, the evidence does not show that people presenting today are less impaired than those presenting previously. This point is explored in more detail below in paragraphs 29 to 36.

  • One of the most important changes over recent decades has been the

increased recognition of ADHD and autism among girls and women. Research suggests that at least part of the historical sex difference reflected lower rates of recognition rather than true differences in prevalence.

  • There has also been a marked growth in public awareness of

neurodevelopmental differences or traits that can occur to varying degrees across the population on a continuum. This is accompanied by a far richer, more accepting public language for mental health and diversity, shared online and through social media. This is reflected in the term neurodiversity, the idea that variation in brain function, cognition and behaviour is part of natural human diversity rather than pathology or disorder. The expression emerged in the 1990s, and is a social, not a clinical diagnostic term (Chapter 4). However, the line between the social and the medical is beginning to blur. Neurodiversity describes the ordinary breadth of human variation rather than something to be diagnosed; yet some providers now offer ‘neurodiversity assessments’, which risk turning a welcome recognition of difference into a formal diagnosis. Self-identification in general and as neurodiverse in particular, usually based on the self-recognition of traits, is increasingly common. In the OxWell secondary school survey, around 1 in 5 pupils described themselves as neurodivergent[14] – well above the proportion with a formal neurodevelopmental diagnosis. Clinical diagnosis is determined not by the presence of traits alone, but by the extent to which they cause persistent difficulties in multiple settings and lead to functional impairment, including distress.

Multiple and increasing difficulties

  • More young people are experiencing multiple, interacting difficulties

rather than a single condition. Co-occurrence is the rule rather than the exception. By 2024, among 16–24-year-olds with a common mental disorder,

57% also screened positive for ADHD and 61% for autistic traits, although screening positive indicates increased likelihood rather than confirmed diagnosis[15]. Anxiety, depression, eating disorders, self-harm and suicidality are all substantially more common among autistic people than in the general population. ADHD likewise commonly co-occurs with autism and emotional disorders. Most people do not present with a single problem but with several interacting difficulties, and systems organised around single conditions cannot see the whole person (Chapter 9).

The impact

  • The consequences are already being borne by individuals, families and

carers, and by society as a whole: distress, impairment, withdrawal, and reduced participation in relationships, education and work. Around one million 16–24-year-olds are not in education, employment or training (NEET). Between 2015 and 2025, the proportion of young people who were NEET and reported a work-limiting health condition increased from 26% to 44%, a rise of almost 70% over the decade[16]. Mental health conditions and autism accounted for around 68% of these work-limiting health conditions. Children diagnosed with ADHD before the age of 17 are around four times more likely to be NEET in early adulthood than those without ADHD[17] – a risk comparable with autism and greater than that associated with many other long-term health conditions. This is an important and growing challenge. Participation is not only an outcome of good health; it is one of the ways good health and wellbeing is built and sustained. Good work, when secure and well supported, offers purpose, structure, income and social connection and can itself aid recovery, while the conditions that help children stay and thrive in education are the same ones that protect their mental health. A system that helps people take part, rather than waiting until they have withdrawn, is fairer, more effective and less costly: earlier participation supports wellbeing and reduces the need for more intensive support later, while wellbeing in turn sustains participation (Chapters 13 and 16).

Drivers of increases in diagnosis and prevalence

Underdiagnosis, misdiagnosis and overdiagnosis

  • Public debate is often reduced to a single question: is there too much

diagnosis or too little? Underdiagnosis, misdiagnosis and overdiagnosis are different problems, and all three can occur within the same system at the same time. They occur across many areas of medicine and are not unique to ADHD, autism or mental health conditions. Diagnosis is often complex, and in

mental health and neurodevelopmental care there is no biological test that provides a simple yes-or-no answer. What matters is whether people receive an accurate and comprehensive assessment that identifies the causes of their difficulties and the degree of functional impairment they are experiencing, appropriate diagnosis where needed, and that this informs the right support that improves their functioning, participation and quality of life.

  • Underdiagnosis occurs when someone has a condition but does not receive

a diagnosis when they should, which can reduce opportunities to access appropriate treatment and support. Misdiagnosis occurs when someone has a diagnosis, but it is either the wrong one or only partially correct, for example missing additional relevant diagnoses. Mental health conditions, ADHD, autism, trauma, sleep problems and adversity can produce overlapping symptoms, and very frequently people experience several conditions at once. The risk of misdiagnosis is more likely to occur where assessment is brief, not face to face, fragmented, made without access to informants or records, and is focused on a single condition. Overdiagnosis is the labelling of a person with a disease or condition that either they do not have, or that would not have caused them harm or significant difficulties if left undiscovered. It also includes turning ordinary, healthy life experiences into medical problems (‘over medicalisation’) by expanding disease definitions or lowering diagnostic thresholds without evidence of better health outcomes. Overdiagnosis, misdiagnosis and underdiagnosis are all system failures.

  • Both misdiagnosis and overdiagnosis are harmful. They may delay

appropriate support, lead to unnecessary or inappropriate treatment, or focus attention on the wrong solutions. People understandably ask whether ‘ordinary’ stress, worry and low mood are increasingly being classified as disorders. That may happen because the threshold between normal variation or having certain traits versus having a clinical condition or disorder can be difficult to define in practice and is a clinical judgement. But over- medicalisation specifically occurs when every difficulty is assumed to require a medical response. For example, a young person who is lonely but not clinically depressed may benefit more from rebuilding social connection and participation than from antidepressant medication.

  • Misdiagnosis is an unwelcome but inescapable fact of modern medicine,

including psychiatry and psychology. However, we are confident that the rapid increases in administrative prevalence and demand for diagnostic assessments, together with the system failures in responding to this, are creating the conditions for increasing risk of misdiagnoses. As was noted

above in paragraph 23, recorded diagnoses in some younger age groups are now exceeding older population prevalence estimates (Chapters 7 and 8).

  • On the basis of evidence available to us we cannot conclude that

overdiagnosis is occurring at scale, particularly given the gaps in population prevalence data and data reporting. However, if present trends continue we are clearly entering a phase in which the balance of risks is changing, from previous merited concerns with under-recognition and underdiagnosis towards an increased risk of misdiagnosis and overdiagnosis, which are becoming increasingly important considerations for policy and service delivery.

  • Under-recognition nevertheless continues for some groups. The

increasing risk of overdiagnosis or misdiagnosis cannot mean neglecting those often overlooked. The requirement is for comprehensive, high-quality assessment and timely diagnosis where needed, while making practical and clinical support available according to need rather than diagnosis alone.

  • None of this should be taken to suggest that people’s distress or

difficulties are not real. If anything, the evidence points the other way: the rising numbers coming forward for assessment have not been accompanied by any fall in severity or impairment – those now presenting are no less affected than those identified before. People are seeking answers and support for very real problems, and they are not the problem. The problem is a system that leans too heavily on diagnosis as the route to help and cannot keep pace. Rising need, greater awareness and recognition, and the demand that follows are running up against limited capacity, fragmented pathways and uneven quality. As a result, need too often goes unmet – waiting lists have become unmanageable, thresholds for services keep rising, and evidence-based, National Institute for Health and Care Excellence (NICE) recommended support is frequently not provided.

  • We have been clear that current evidence cannot yet adequately

quantify the scale of each of these problems at the population level. Regular prevalence surveys, together with better information on assessment quality and outcomes, are needed to monitor how the diagnostic landscape changes over time: the balance between underdiagnosis, misdiagnosis and overdiagnosis; the effects of changing diagnostic concepts and thresholds; the recognition of co-occurring conditions; and the relationship between

population prevalence, administrative prevalence and demand for assessment.

Social, cultural, environmental and system drivers of rising prevalence, need and demand

  • The wider drivers of rising prevalence, need and demand include social,

cultural and environmental change, alongside socio-cultural changes in awareness, recognition and the organisation of services. No single cause explains the increases observed.

  • The context in which people live, and in which children and young people

grow up, has changed. COVID-19 accelerated these trends, but they began before the pandemic. Mental health and wellbeing are shaped not only by exposure to adversity and stress but by opportunities for connection with other people, increased participation and a sense of purpose. The drivers can broadly be categorised as follows:

  • More uncertainty and insecurity: poverty, financial insecurity,

unemployment and housing pressures have continued to affect mental health, compounded by uncertainty associated with the cost of living, climate change and wider political instability.

  • More evaluation and comparison: educational, work and social

environments have become more continuously evaluative, increasing opportunities for comparison and scrutiny of performance, appearance and social standing. We find that education has become a more significant contributor to distress than for previous generations, but not that it is the principal cause of the rise. Social media can intensify these pressures, and can be a channel for bullying, contributing to greater distress and poorer mental health – particularly for those already living with anxiety, depression or eating disorders.

  • Less belonging and more loneliness: reductions in youth and

community provision have reduced opportunities to connect, while loneliness and social isolation have become increasingly important contributors to poor mental health.

  • Reduced opportunities for rest and recovery: more of daily life is spent

online, with less opportunity for physical activity, time in nature, play and uninterrupted sleep.

  • Understanding has also changed. Greater awareness has reduced stigma

for some and helped many people understand their own experiences, bringing previously unrecognised needs into view and increasing demand for support.

  • The risks of misdiagnosis and overdiagnosis increase where an

individual’s difficulties overlap, assessment quality varies, data is withheld, thresholds are applied inconsistently, and diagnosis becomes the principal gateway to support.

  • Diagnosis has increasingly become a requirement to access help

through education health and care plans (EHCPs), Access to Work, and in parts of the welfare system. The Review argues that support should be available according to identified need rather than depending solely on diagnosis (Chapter 13).

  • Diagnostic concepts have also broadened which has contributed to

some increase in administrative prevalence, but are not the sole factor in driving the increases. Changes introduced in DSM-5 and the International Classification of Diseases, 11th Revision (ICD-11) changed the diagnostic criteria for ADHD and autism. This combined previously separate autism subtypes into a single spectrum diagnosis and, for the first time, formally recognised that ADHD and autism can co-occur, while increasing recognition of ADHD in adults. As a result, some people who would not previously have met diagnostic criteria, or who might previously have received a different diagnosis, are now eligible for diagnosis. These changes are therefore likely to have contributed to increases in recorded prevalence and diagnostic rates, although they do not by themselves explain the full scale of the increases now being observed.

  • Many people experience several interacting difficulties that do not fit

neatly within a single diagnosis. This increases the risk of misdiagnosis and reinforces the need for comprehensive assessment capable of identifying multiple interacting needs rather than directing people through separate single-condition pathways. Someone may be diagnosed with autism while clinically significant anxiety is overlooked; equally, undiagnosed and untreated ADHD may be mistaken for anxiety or depression alone.

  • Across both mental health and neurodevelopmental conditions, traits

are continuously distributed across the population. Most people are sometimes inattentive, anxious or socially uncomfortable. There is no biological test. Diagnosis depends on the persistence of difficulties, their developmental history, their presence across different settings, and the degree of functional impairment they cause. Circumstances influence how

these traits are expressed and how impairing they become. Increasing psychological distress and environmental demands may interact with underlying neurodevelopmental vulnerabilities, making difficulties more impairing and more likely to come to clinical attention. Having ADHD- or autism-related traits does not necessarily mean someone has a diagnosable condition. Diagnosis therefore remains a matter of informed clinical judgement, drawing together multiple sources of information and recognised diagnostic criteria. Equally important, diagnosis and support needs are not the same. People may require support without meeting diagnostic thresholds, while people with the same diagnosis may require very different kinds of help. The aim is to provide the right support at the right time, rather than making diagnosis the only route to help.

  • There are strong concerns about the consistency and quality of ADHD

and autism assessments being carried out by independent providers, and the financial incentives associated with assessment activity. Independent providers now carry out more than half of all NHS-funded ADHD assessments and more than a third of autism assessments. Spending on independent-sector ADHD provision has risen by around 252% in three years, with a projected overspend of £164 million across the areas analysed alone[18]. Yet this large and fast-growing use of public money is running without consistent oversight, transparency or quality assurance – commissioners often cannot see what they are paying for (Chapter 12). This is not acceptable. It is no criticism of independent provision in principle – independent providers are often used successfully to cover gaps in service or reduce waiting lists – nor is it a criticism of the people who turn to them after long NHS waits. But what has happened here is that a major, publicly funded activity has been allowed to grow faster than the arrangements meant to guarantee its quality and value. Public money must be properly commissioned, regulated, monitored and assured, and everyone deserves to trust that the assessment they receive is consistent and of high quality, wherever it is carried out.

How well the current system is set up to respond

  • The current system is not working for the people it is meant to serve, the

patients or the professionals who work in it, and not for the public who fund it. Judged against what people actually need, it falls short (Chapters 13–17). Incremental fixes will not close that gap: what is needed is a new model of care and radical reform.

  • Investment in mental health and neurodevelopmental support has not

kept pace with need. The result is a system that too often responds late, and at greater human and economic cost. The consequences are already being felt across health, education, welfare, employment, local government and families.

  • Support across health, care, education, employment and welfare

remains fragmented. This is not only distressing for the people caught within it; it is also inefficient. Some people are funnelled towards expensive specialist services after long delays, when earlier, simpler support would have been more appropriate. Others cannot access specialist care at all when they need it most.

  • Waiting lists are too long and delay itself causes harm. National evidence

suggests that the proportion of people whose difficulties worsen while waiting rises almost three-fold, from around one in four among those waiting less than two weeks to more than seven in ten among those waiting longer than six months[19] (Chapter 12).

  • A diagnosis is not in itself an outcome. Too often diagnosis is followed by

no timely offer of evidence-based treatment or support. At precisely the point where the system appears to have succeeded, it often fails the person it was intended to help (Chapter 15).

  • Unfairness, inequality and unwarranted variation must be addressed.

Rising need is not evenly distributed. It falls disproportionately on people experiencing deprivation, discrimination and disadvantage, and interacts with ethnicity, disability and geography. Those with the greatest need are often the least able to secure timely support. Nearly half of looked-after children have a diagnosable mental health condition[20], and around eight in ten children cautioned or sentenced have identified special educational needs[21], yet these are among the groups least well served by existing pathways. Patterns of access also risk creating new inequalities, with services often responding most effectively to those best able to navigate complex systems or pay for alternatives (Chapters 13–17).

Our call to action and the opportunity for government

  • This Review recommends moving from a system that too often requires

people to wait for diagnosis before receiving help to one that recognises need earlier and responds sooner, while protecting specialist capacity for

those who need it most. Diagnosis remains essential, but it should sit within a broader system organised around need rather than acting as the principal gateway to support.

  • Our recommendations focus on intervening earlier, preventing avoidable

distress and impairment, and strengthening the specialist services on which people with the most severe and complex needs depend. Resources should be used more effectively across the whole system so that support is proportionate to need and addresses the drivers of the substantial increases in distress, demand and disability described throughout this Review.

  • Increasing participation in relationships, family life, communities,

education and work should become a central measure of success, while recognising that the outcomes that matter most will differ from person to person. But participation is more than an outcome; it is essential to how health and wellbeing are created and sustained. Belonging, purpose, routine, learning and good work are the rewards of recovery as well as being the conditions that make recovery possible. A system that enables people to participate, rather than waiting until they have recovered before they can do so, is therefore fairer as well as being more likely to be effective. Participation and wellbeing reinforce one another.

  • For children and young people, this is a generational opportunity. The

government should reduce avoidable distress and improve wellbeing while ensuring that children are not left to adapt to systems and environments that create or compound their difficulties. Neurodevelopmental difference is not a problem to be eliminated. The challenge is to create the conditions in which children, young people and adults are understood, supported, treated where treatment is needed, and able to learn, work, participate and flourish.

  • The recommendations that follow combine actions that can begin within the

next 12 to 24 months with a longer-term programme of reform to build capacity, strengthen prevention and improve outcomes over the coming decade. They will require investment, national leadership and sustained partnership across government, public services, employers, education and the voluntary and community sector. It will also require public resources to be used more effectively, reducing fragmentation and duplication, and shifting the balance towards prevention and early intervention, which cost far less than the crises they are able to prevent. The potential return is substantial: improved wellbeing and participation, better use of specialist services, and greater value from public investment. The government now has an opportunity to begin that transformation.

Recommendations for government and local systems

Chapter 18 explains how support can be organised around need, functioning and participation, including through shared descriptions of need. Chapter 19 sets out our vision for a new model of care and support, together with the evidence on what more effective support looks like for children and young people, and where the same principles can be applied across the life course. See chapter 20 for further detail on the ten priority recommendations below.

Recommendation 1: Tackle the drivers of mental distress and ill health through an ambitious approach to prevention, early intervention and improving public mental health and wellbeing.

A range of specific actions is set out in Chapter 20, focusing on creating the conditions, opportunities and communities that protect mental health and wellbeing. These actions should be underpinned by

  • stronger governance and accountability for public mental health within the Department of Health and Social Care (DHSC) and the Department for Education (DfE), across government and at local level;
  • long-term investment in youth, community and voluntary organisations;
  • meaningful partnership with people with lived experience.

Recommendation 2: Move from a ‘diagnosis-dependent’ to a ‘needs-led plus diagnosis’ system by developing a digitally and AI-enabled shared needs assessment and prioritisation framework for people with mental health and neurodevelopmental needs.

This framework should be based on the ‘no wrong front door’ principle, with prioritisation according to each person’s needs, level of functional impairment and risk. There should also be a clinically led review of current ADHD and autism assessment waiting lists to:

  • identify the primary presenting concern
  • to prioritise the most relevant existing pathway
  • organise assessments for other potential co-existing conditions around this core assessment to determine the best course of action for each individual
  • offer evidence-based support while waiting.

The purpose is not to remove people from waiting lists, but to make sure each person is offered the right response to their needs. Crucially, no one should be taken

off a waiting list without being connected to appropriate help, and alternative support should be actively arranged with the person, not simply signposted.

Recommendation 3: Commission and deliver integrated digitally-enabled mental health and neurodevelopmental assessment and support that is age- appropriate and responds to the needs of local communities.

Support should be:

  • easy to access
  • as close to home as possible
  • organised around people through multidisciplinary teams, following the ‘expert at hand’ principle
  • joined up from childhood through to adulthood

There should be a focus on parity of access and supporting participation. Integrated assessment of the possible reasons for mental health and neurodevelopmental problems should minimise misdiagnosis and ensure appropriate treatment.

Recommendation 4: Close treatment gaps while protecting, strengthening and investing in specialist services to improve access and outcomes for those with the most severe, complex and enduring needs, including specialist support delivered in the community, and crisis and inpatient services.

Access to NICE-recommended and evidence-based interventions and therapies should be expanded; specialist capacity should be focused where it adds greatest value; timely medication reviews and safe deprescribing practices should be strengthened and standardised; and there should be a focus on improving social inclusion and quality of life for people with the highest support needs.

Recommendation 5: Overhaul the commissioning, funding, oversight and regulation of mental health and neurodevelopmental services, including making all diagnostic assessments for ADHD and autism Care Quality Commission (CQC) regulated activity and banning direct-to-public advertising of NHS-funded ADHD and autism assessments.

Move away from separately commissioned ADHD, autism and mental health assessment and support services towards outcome-based commissioning of integrated neurodevelopmental and mental health assessment and support; ensure consistent quality standards and reporting; and shift from episodic specialist care for ADHD to long-term management in primary care where required.

Recommendation 6: Build capacity and capability across the multidisciplinary workforce to deliver more integrated mental health and neurodevelopmental support, backed by investment in training, supervision and career development.

Multidisciplinary workforce strategies should be co-produced with relevant professional bodies, practitioners and people with lived experience; neighbourhood multidisciplinary teams (MDTs) should be built with representation across the health sector and beyond, including education, social care and the voluntary sector; all staff should have a core competence in mental health and neurodevelopmental needs, supported by training and a nationally agreed competency framework.

Recommendation 7: Take an ambitious approach to scaling up safe digital and AI-enabled tools within a digitally enabled and integrated care pathway, while sharing and using data to improve quality of care, efficiency and productivity.

Build an integrated digital infrastructure platform for the orchestration of care pathways, using AI to support delivery of the needs assessment and prioritisation framework; extend regulated digital therapies safely and securely; improve information and support available on the NHS App; build interoperable data infrastructure; and preserve in-person care rather than replacing it.

Recommendation 8: Improve data collection and evaluation for mental disorders, ADHD and autism, including through regular population prevalence studies, and update NICE guidelines to strengthen the evidence-base for policy development and implementation.

Invest in more regular national population prevalence surveys and improve the national data infrastructure with disaggregated data to better understand inequalities. Disaggregated data is data that has been broken down according to factors such as age, sex, ethnicity or region.

Recommendation 9: Commission a national public information and education campaign to help people better understand, support and accommodate neurodevelopmental conditions and manage their mental health, supported by a trusted digital platform providing evidence-based information, tools and support without requiring diagnosis.

This campaign should be co-produced with people with lived experience and evaluated against mental health and wellbeing outcomes as well as awareness and knowledge.

Recommendation 10: Publish a cross-government mental health strategy this year setting out how this report’s recommendations will be implemented through immediate actions in the next 12 to 24 months and a 10-year reform plan.

This strategy should put improving mental health and wellbeing and improving participation at the centre of government decision-making, underpinned by transparent departmental accountability, investment and workforce requirements.

Contents

Section One: Context and approach 27

Chapter 5. Overdiagnosis, misdiagnosis, under-recognition, underdiagnosis and

  • medicalisation 53

Section Two: What is changing in population prevalence, need and service

  • demand? 76

Chapter 6. Population mental health: what is changing, for whom, and with what

Chapter 10. Environmental drivers of change in mental health for children and young

  • people 213

Chapter 11. Changing understanding of mental health and neurodevelopmental

Chapter 15. Service models, mental health treatment delivery and

Chapter 19. A model of care matching support to need, functioning and

Annexes 451

Section One: Context and approach

Chapter 1. Introduction

This Review is about how we understand, and respond to, the needs of people with mental health conditions, ADHD and autism in England.

The Review was commissioned in response to the growing difficulties many people face in obtaining recognition and support for mental health and neurodevelopmental conditions. The Review seeks to bring clarity through evidence to a debate that is contested, has brought frustration to individuals and sustained pressure on public services.

1.1 Why this Review was established

Mental health conditions, ADHD and autism have moved to the centre of public life in England. Over the past two decades, and with particular speed in the last ten years, greater openness has reduced some of the stigma that surrounded these conditions, encouraged more people to seek help, and brought recognition to some groups whose difficulties had long been overlooked. These are real gains, and it is right to begin by affirming them. That more people now feel able to speak about their mental health and neurodevelopmental conditions, and our public services increasingly treat them as central to health, education, employment and participation in ordinary life, is a change to be welcomed.

Yet alongside this progress, the demand for assessment and support has risen to a degree that few anticipated and that current services were never designed to meet. Requests for additional support in schools have grown year upon year. Primary care, specialist services, universities, employers and community organisations all describe sustained pressure. And as the numbers have grown, so the public conversation about them has grown louder and more polarised, organised increasingly around a single, contested question: what does the rise actually mean?

1.2 A debate that has become polarised

Three broad interpretations have come to dominate that conversation. The first holds that the rise reflects an increase in mental health and neurodevelopmental conditions. The second holds that much of the rise reflects not a change in the underlying mental health and neurodevelopmental conditions but a change in how we recognise, describe and respond to them. Some argue that greater awareness,

combined with lower diagnostic thresholds and growing digital, economic, social and educational pressures that earlier generations did not face, has encouraged ordinary unhappiness to be reframed in clinical language. They suggest this has inflated diagnostic rates and led to diagnostic ‘mission creep’, with the boundaries between normal emotion, behaviour and clinically significant disorder becoming increasingly blurred. This position is usually summed up as overmedicalisation and overdiagnosis, and it is nothing new in medicine or psychiatry. The third holds that a system in which diagnosis has become the key that unlocks support for the individual will understandably generate greater demand for diagnosis.

None of these positions can be easily dismissed, and each captures something real. None, on its own, explains the evidence for the rise in impairment and deterioration in functioning, particularly among the young. More people are struggling, and struggling more deeply, than before. It increasingly bears on their capacity to take part in ordinary life – to learn, to work and to sustain relationships.

The questions about what is changing and why have acquired a national dimension. Rising distress among young people now shows itself in school absence and, increasingly, in the labour market and the finding that mental ill health has become one of the principal drivers of economic inactivity among young adults[22]. It is partly this – the movement of a health question into the heart of education and the economy – that prompted the Government to commission this Independent Review. The Terms of Reference were to examine the rising numbers of mental health conditions, ADHD and autism; the drivers of that rise; inequalities in prevalence and in access to support; the role of diagnosis, medication and medicalisation; the opportunities for prevention, early intervention and support, both within the NHS and beyond it[23].

The Review examined the available evidence on the prevalence of mental health conditions, autism and ADHD, looking at not just whether the numbers have risen, which they have, but also at how the different things measured – population prevalence, psychological distress, diagnosis, self-identification, service demand and need – relate to one another. It considered what those relationships imply from the perspective of the individuals involved and how a system of support could be fairer, earlier, and better matched to what is actually needed to manage their lives. What began as an inquiry into prevalence became, of necessity, a broader inquiry into how we understand and respond to mental health and neurodevelopmental need.

1.3 Why this matters now

The experience of young people is a core concern. Where psychological distress was once greater in midlife than in youth, young adults now report the highest levels

of all. The strong association between mental health and neurodevelopmental conditions and being out of education, employment or training can narrow a life before it has properly begun, with profound consequences for individuals and for society as a whole, not just the health services[24, 25]. It is this human cost, more than the pressure on services, that lends the Review its urgency.

Chapters 6 to 9 set out the current evidence on population mental health, ADHD and autism, and on co-occurrence and complexity.

1.4 The Review’s approach

No single dataset or discipline can explain the changes we are witnessing. Mental health conditions, ADHD and autism sit at the meeting point of biology, psychology, education, social policy and lived experience. The Review draws on different forms of evidence – national surveys and epidemiological studies, routine health and education data, systematic reviews, international comparisons, questionnaire surveys, focus groups, expert testimony, and sustained engagement with people who live with these conditions and those who support them – on the principle that different sources can answer different questions. To structure the work, multidisciplinary Topic Groups were established across the areas set out in the Terms of Reference, informed by many hundreds of written submissions, with input from people with lived experience working as part of the Review and charities and professional organisations concerned with these conditions. Chapter 3 sets out this approach, and its limits.

1.5 From the interim report to the final report

The Review published an interim report in March 2026[26]. Its purpose was to set out the questions the evidence was raising, to be honest about what remained unknown, and to describe how the second phase of the work would proceed.

That report made several observations that continue to anchor this one. It confirmed that psychological distress among young people has risen, and that demand across mental health, ADHD and autism services has risen with it. It showed that different sources of evidence should be understood as the product of several processes unfolding at once. Distress is real, it insisted, whatever difficulty we may have in classifying it; the question is not whether it is genuine, but whether it is being understood and responded to in the right way.

The interim report also drew attention to a feature of the present system that this Review examines closely. Diagnosis has become the practical gateway through which people obtain recognition, adjustments and support in education and the

workplace, even though the question a diagnosis answers, ‘what is the nature of this condition?’, is not the same as the question support must answer: ‘what does this person need now to function?’. We agreed that the second phase should look more closely at the relationships between distress, diagnosis, functioning, impairment and need, and ask whether a system organised around need might respond earlier, and more proportionately, than one organised around diagnosis alone.

This Review builds directly on those foundations. Some of the interim observations have since been confirmed; others have been refined as further evidence has come in. Most importantly, the second phase has allowed what were separate observations to be drawn together into a more coherent way of understanding mental health and neurodevelopmental need – the framework that the chapters ahead set out.

1.6 What the Review does, and does not, attempt

A word is needed about the boundaries of this Review, for they shape everything that follows. Its subject is mental health conditions, ADHD and autism, considered across the life course – sometimes together, where the questions and the systems that respond to them are shared, and sometimes separately, where they differ. Its focus is England, although it draws on international evidence where that helps to show what is, and is not, distinctive about our own situation. And it is concerned throughout with the whole of the system that surrounds a person – not health services alone, but education, employment, welfare, social care, family and community – because need does not confine itself to the boundaries of any one department of life or government, and neither can any adequate response to it.

It is equally important to be clear about what this Review does not attempt. It is not a clinical guideline, and it does not adjudicate individual diagnoses or prescribe how any individual should be assessed or treated. It does not question the biological processes of mental health and neurodevelopmental conditions or underlying causes of any specific disorder. Nor does it question the reality of anyone’s experience, nor the value of a diagnosis to those for whom it has brought understanding, or access to treatment; its purpose is to improve the match between need and support. And it is careful, in a field where the structured description of need can slide into the rationing of it, to keep those two things apart. To describe need more clearly is not to build a new barrier to support.

Chapter 2. Key questions

The key questions examined by the Review are outlined: What is changing in the population? What is changing in systems? What is changing in how distress is understood? How are prevalence, diagnosis, demand and need related? What kinds of support are needed, by whom, when and for how long? What should government and public services do differently?

Many of the contradictions that so animate current debate stem from conflating distinct ideas and meanings: population prevalence and administrative prevalence; diagnosis and need; functioning and participation; demand and distress. These are related, but they are not the same, and when they are treated as interchangeable, confusion follows for researchers, policymakers, practitioners and the public alike. Separating them is the Review’s first task, in Chapters 3 and 4.

Diagnosis is fundamental to good clinical care and individual autonomy. It explains the nature of a condition, guides evidence-based treatment, and gives people and professionals a shared language for describing experiences. However, a diagnosis speaks to a defined condition; it does not, on its own, settle what a person needs now, which depends on how they are functioning, on what they are able to take part in, on their circumstances, their own hopes and goals, and on whether the world around them helps or hinders. Two people who share a diagnosis may need very different things; two who need much the same thing may share no diagnosis at all. The goal is to ensure that people are understood and supported according to their needs.

This report examines whether the strain currently felt across health, education and social care reflects that more people are being diagnosed with a condition, or that more people have these conditions and need diagnoses, or that our systems have organised themselves around diagnosis as an administrative key to support without which that support does not happen. The choice is unlikely to be the stark one the public debate assumes. The spirit in which the Review approaches this is a simple and human one. It is not to diagnose less where diagnosis helps, nor more where it does not, but to ask how people with problems might be understood and supported according to what they need. That means the right understanding, the right help, at the right time, within public systems that are able to learn and improve as they go. This is the golden thread that the Review pursues in the chapters that follow.

2.1 What is changing in the population?

The first question sounds simple: is there more mental ill health, autism and ADHD in the population than there was, and if so, in whom, why, and with what consequences? The strongest evidence comes from studies of the population as a whole rather than of those who reach services. Chapters 6 to 9 present evidence on where there are changes in prevalence. Chapters 10 to 12 discuss the drivers of change, including increasing awareness and understanding, and the ways in which systems have shaped recognition, diagnosis and support.

The public argument over how to read this has become fierce. One account holds that a generation has become genuinely more distressed, under digital, economic, social and educational pressures that earlier generations did not face. Another argues that ordinary unhappiness has been reframed in clinical language, inflating the figures. Both accounts may hold some truth, investigated in the chapters that follow. But the Review’s reading is that relabelling alone cannot explain what we observe: the rising numbers coming forward have not been accompanied by any fall in severity or impairment, including among those reporting the highest levels of distress.

The more revealing question is whether symptoms are increasingly bearing on people’s capacity to take part in ordinary life – to learn, to work, to sustain relationships. That question has acquired a national dimension: rising distress among young people now shows itself in school absence and, increasingly, in the labour market, where mental ill health has become one of the principal drivers of economic inactivity among young adults[22]. It is partly this – the movement of a health question into the heart of education and the economy – that prompted the Review.

2.2 What is changing in systems?

The next question moves from the population to the machinery that responds to it. Here, the simplistic reading that services are simply struggling to keep pace with more patients, is overturned by the evidence. Systems do not merely absorb demand; they shape it. Demand has grown; the question is how much of it has been generated by the system’s own architecture. The Review therefore examines whether current arrangements have made diagnosis the single reliable key to support, whether constrained access has amplified demand rather than merely delaying it, and why a growing share of diagnoses among children are no longer followed by evidence-based treatment. These are the questions taken up in Chapters 13 to 15 on the systems through which people seek and receive support.

2.3 What is changing in how distress is understood?

The third question concerns the current cultural context, the frame through which a society interprets its own inner life, and it is here that the debate has become the most contested. Change in populations, systems and societal understanding now unfolds against a charged political backdrop – a rising working-age disability and incapacity bill, and a welfare debate in which the same evidence is read by some as proof of unmet need and by others as evidence of over-medicalisation. Over two decades, falling stigma and a greater mental health literacy have made it far more possible for some people to recognise difficulty and to seek help[27] – these are positive changes, although the picture is not one of simple progress. Some evidence suggests that public attitudes may in places be moving backwards, with lower expectations of recovery and reduced acceptance of caring for people with severe mental illness in the community[28] – a reminder that greater openness has not settled into a stable or uniformly benign consensus.

The Review’s task is to understand the interaction of these components and the perspectives of the individuals trying to navigate systems.

The Review distinguishes between underlying prevalence, diagnosis, function and need, and clarifies these concepts in Chapter 4: (i) diagnosis, which asks whether a person meets the criteria for a defined disorder or condition; (ii) functional impact, which asks how far their difficulties affect everyday activities and participation; and (iii) need, which asks what support is required at a given moment in order to function in particular environments, and which may rise and fall over time.

Two tensions are named rather than resolved here. The first is that we continue to use categorical labels defined by two international classification systems, the Diagnostic and Statistical Manual of Mental Disorders (DSM) and the International Classification of Diseases (ICD), even though many of the symptoms they classify are continuously distributed across the population. This issue is discussed in Chapters 4, 5, 7 and 8.

The second tension concerns functional impairment. Impairment of function is a requirement for the diagnosis of autism and ADHD. At the same time, changes to international classification systems have broadened some diagnostic categories, increasing recognition and diagnosis of autism and ADHD among previously under-

recognised groups, including girls and women. Whether changes in diagnostic thresholds have contributed to rising diagnosis remains uncertain because the health service does not routinely record severity, functioning or impairment among people receiving diagnoses. That absence – the single most important knowledge gap for understanding exactly what an individual’s support needs are and measuring change over time – means the very data needed to settle the ‘overdiagnosis’ argument within services largely do not exist. The distinctions between diagnosis, functional impairment and need are defined in Chapter 4 on language and concepts and revisited throughout the chapters that follow, which examine why diagnosis, though necessary, is not sufficient on its own.

The Review strongly recommends a shared language to describe function and need across all services using the same principles as the International Classification of Functioning, Disability and Health (ICF)[29] developed by the WHO, which also provides a coding system for data analysis.

2.5 What kinds of support are needed, by whom, and when?

The fifth question asks about the nature of need itself, including complexity. People rarely present with a single, well-bounded difficulty. The Review asks what should determine support: how functioning, participation, context, risk and co-occurring difficulties ought to be weighed alongside diagnosis; how support should change as a person’s circumstances change; and how systems might respond in proportion to complexity without forcing everyone to wait for the same specialist assessment.

It asks these questions in full awareness of a danger it takes seriously – that any move to describe need more systematically can slide, under fiscal pressure, into a new instrument for rationing it, or can be read by clinicians as an attempt to dilute the diagnostic process. How to describe need more clearly while preserving the integrity of diagnosis – how, in other words, to keep diagnosis clinical and trustworthy rather than asking diagnosis to carry the whole weight of deciding who gets help – is one of the questions the Review will address. These questions are developed in Chapters 18 and 19.

2.6 What should government and public services do differently?

Finally, the Review addresses current support pathways to determine what a better system might look like. This is an open question, to be supported by the evidence rather than assumed at the outset. It does not assume that simply expanding existing

pathways is the answer – this has already happened with the introduction of the Right to Choose initiative, which has led to soaring costs that the NHS has had to meet without necessarily meeting an individual’s needs any better. Given the data discussed in Chapters 5 to 9, simply carrying on as before is creating a serious risk of overdiagnosis in the strict use of the word – a system failure, not a personal one. Recommendations are put forward to develop a system that moves from a diagnosis-dependent to a needs-led plus diagnosis system in which support starts when need is identified.

The Review’s answers will not stand in isolation. They will need to sit alongside the reform of the SEND (special educational needs and disabilities) system, and wider work on young people and their participation in work, including the review led by the Rt Hon Alan Milburn. The task is less to widen any single pathway than to build a system able to answer many different levels of need – and one that can also recognise when what a person needs is not a clinical response at all, but the support of the people and institutions already around them: family, friends, colleagues, schools, employers and a well-supported general practice. Most importantly, meeting need should achieve better outcomes for individuals and their families.

Chapter 3. How the Review approached the evidence

This chapter explains the Review’s evidence approach, highlighting that no single dataset can answer the questions central to the Review.

Population surveys are treated as the strongest source for understanding epidemiological prevalence, while administrative data are used to understand diagnosis, access and service activity, and qualitative and lived-experience data are used to help interpret and extract meaning from statistics. Conclusions are reached through triangulating data sources, weighing evidence according to strength and being explicit about uncertainty.

The Review identifies current gaps in the evidence, such as limited routine data on severity, impairment and outcomes, weak linkages across health, education and welfare, and incomplete capture of co-occurring conditions. These limitations are revisited in the recommendations but should be considered when interpreting the Review’s findings.

How one answers a question depends a great deal on how one looks for evidence. The questions for this Review, about prevalence, diagnosis, demand and need, cannot be settled by a single source. This chapter describes the Review’s approach: using multiple evidence types, clarifying what each can and cannot show, how sources that did not always agree were weighed, and the graded way of moving from evidence to conclusion.

3.1 Why no single dataset can answer the Review’s questions

Mental health conditions, ADHD and autism sit at the meeting point of epidemiology, clinical practice, education, culture, social policy and lived experience, and each of these sees a part of the whole that the others cannot. The Review’s first principle is that different forms of evidence address different questions:

  • Population surveys are the most reliable guide to whether the underlying prevalence of distress or disorder has changed in the community.
  • Administrative datasets describe how systems recognise, record and respond to those who reach them.
  • Qualitative and lived-experience evidence explains how those systems are encountered in practice.

The characteristic error is to let one form stand in for another. Measurement and interpretation are distinct. Data do not simply mirror reality; they are produced within systems that determine what is recorded, how it is classified and how it is used. Symptoms, diagnosis, functioning and need are related but not identical.

3.2 Population surveys: the strongest guide to epidemiological prevalence

For questions of prevalence, the Review prioritised population-based studies using measures that are comparable over time, because these are not restricted to people who have sought help or received a diagnosis. The principal sources were the Adult Psychiatric Morbidity Survey (APMS), the Mental Health of Children and Young People survey (MHCYP), the UK Household Longitudinal Study (Understanding Society) and the Health Survey for England, with the Global Burden of Disease study used for international context[30-33]. These surveys use different instruments, and the instrument matters as much as the sample: common mental disorders in APMS are assessed with the Clinical Interview Schedule–Revised (CIS-R); psychological distress in Understanding Society with the General Health Questionnaire (GHQ-12); child difficulties in MHCYP with the Strengths and Difficulties Questionnaire (SDQ) and the more detailed Development and Well-Being Assessment (DAWBA); adult ADHD with the Adult ADHD Self-Report Scale (ASRS); and autism through screening with the Autism Spectrum Quotient (AQ) followed, in APMS, by the Autism Diagnostic Observation Schedule (ADOS). Given that consistency of measurement, far more than any single estimate, is what allows trends to be read with confidence, the Review prioritised measures collected by consistent methods across repeated waves.

These surveys have limitations. Response rates have fallen over time, and groups with the highest levels of need (people in insecure housing, residential care or prison) are often under-represented or excluded altogether. The mode of administration also shapes findings: the shift in the MHCYP from face-to-face to online collection between 2017 and 2020 requires caution, because non-face-to-face methods can overstate prevalence compared with interviews and may not be comparable with earlier waves[34]. Willingness to disclose is not fixed; as stigma falls and awareness grows, people may simply report more of what was always there as social understandings of conditions change.

Where the condition of interest lies on a continuum rather than either side of a discrete cut-off, a judgement must be made about when a threshold has been

crossed; the greatest difficulty arises for people ‘on the margins’, and variation between and within studies is inevitable. Trends must therefore be read with attention to measurement consistency. Differences in baseline prevalence estimates across studies usually reflect not error, but different tools capturing different aspects of the same underlying reality.

3.3 Administrative and service data

For questions of diagnosis, access and service use, the Review drew on administrative data: primary care records, secondary mental health datasets, prescribing data and education data. These show how systems function, but they do not measure underlying prevalence, because a recorded diagnosis depends not only on symptoms but on impairment, help-seeking, professional recognition, referral, diagnostic thresholds, service availability, recording practice and, in some settings, the incentives created when diagnosis unlocks support. The single clearest methodological message of the Review is that administrative prevalence must not be confused with epidemiological (or ‘population’) prevalence.

The Review also distinguished between administrative sources, which are not equally trustworthy for a given purpose. Primary care records, such as the CPRD (Clinical Practice Research Datalink) Aurum database, are the most useful national source for diagnostic trends, since they cover defined populations registered with general practitioners, but also can capture diagnoses made elsewhere in the system once these are communicated to a general practice. Secondary mental health datasets are invaluable for understanding referrals, contacts and pathways, but are considerably less reliable as measures of diagnostic prevalence, given that, for conditions like depression or anxiety, most cases never reach secondary care. Secondary care mental health datasets may also miss data, for example from community child health services, and coding may be incomplete. Analyses presented to the Review indicate that, for ADHD and autism, a substantial share of diagnoses recorded in clinical notes never reach structured data fields in secondary care data, and become visible only when free text is interrogated using natural language processing – as in the Akrivia Health analyses used in the Review.

A further track of activity is harder still to see. Assessments delivered through private and Right to Choose pathways may surface in NHS records only when a person later seeks prescribing or shared care, so that a growing component of diagnostic activity sits partly outside the national data altogether. Given that the independent sector now delivers well over half of all NHS-funded ADHD assessments, this is a significant gap[35].

3.4 Clinical, pathway and linked data

No single dataset tracks a person through all stages from distress to diagnosis, treatment and support. The Review also drew on linked datasets and pathway analyses, which revealed waiting times, movement through the system, and the role of co-occurring conditions. These are crucial for understanding real experiences and system bottlenecks. They also allow a view of the relationship between clinical treatment and the wider forms of support that aim to maintain stability, improve functioning or enable participation, and that are delivered across health, education and social care.

3.5 Qualitative evidence and lived experience

Alongside the quantitative work, the Review drew on qualitative evidence and on sustained engagement with people who live with these conditions and those who work with them. Consistent themes were fragmented pathways, the repeated retelling of one’s story, long waits, variable thresholds, diagnosis as the necessary route to support, and concerns about under-recognition and misdiagnosis. Such testimony cannot by itself establish how widespread a problem is; what it can do, and what no dataset can, is show how the patterns in the numbers are actually lived.

The engagement was substantial and structured, including over 50 organisations across the mental health and neurodevelopmental sectors, a large volume of written contributions from individuals, and lived-experience partners embedded within established Topic Groups, providing ongoing challenge throughout. Lived experience thus consistently informed how the evidence was read.

While many of the survey tools of epidemiological prevalence rely on self-completed questionnaires, the Review was also able to draw on studies of self-reported diagnoses of both mental health conditions and neurodivergence through the OxWell study covering 20,000–30,000 individuals[14]. We also conducted a national survey of healthcare professionals and educators asking about their perspectives on ADHD, autism and mental health presentations and their experiences in managing these conditions. The Review conducted secondary analyses of qualitative interview data from the national i-THRIVE programme evaluation involving interviews with staff working in twenty child mental health services across England[36], and from the Timely research programme on developing preventative pathways for child mental health[37]. With the support of Mind, Anna Freud and Ambitious about Autism, we conducted focus groups with those with lived experience from a range of age groups and parents and carers of those with lived experience, to explore how our research and interim findings corresponded to their experience.

3.6 Weighing the evidence: stronger and weaker conclusions

Not all of the Review’s findings carry equal weight. Some patterns recur across independent datasets and can be stated with real confidence; others are suggestive but incomplete; others still are best regarded as hypotheses awaiting further analysis. To keep interpretation honest, the Review worked with a simple, explicit evidence framework that distinguishes stronger from weaker conclusions and guards against overstatement. That discipline matters most in exactly this field, where debate is polarised and where there is standing pressure from every direction to claim more than the evidence can bear. Where the evidence is uncertain, or where expert judgement diverges, the Review says so.

Table 3.1. Structured framework for grading evidence quality and calibrating language used to inform the Independent Review

Evidence quality

Evidence profile Appropriate language

Strong, unequivocal evidence

- Multiple independent studies (replication across teams / settings) - Low risk of bias and clear methods - Convergent designs pointing the same way (e.g. cohort, registry and quasi- experimental) - Consistency in effect direction; heterogeneity explained - Precision (reasonably narrow uncertainty for key outcomes) - Temporal plausibility: longitudinal evidence where time / order matters - Robustness: sensitivity analyses do not change the conclusion materially

‘there is strong evidence

  • that ’

‘consistent findings

  • across studies show ’

‘the balance of high- quality evidence indicates…’

Some evidence, not decisive

‘evidence suggests…’ ‘evidence is consistent

  • with ’

‘findings are mixed…’ ‘we cannot rule out…’ ‘the most plausible

  • interpretation is ’

- Limited replication (one or two studies; mixed findings) - Methodological constraints (selection bias, measurement issues, unmeasured confounding) - Indirectness (proxy outcomes, wrong population, different context) - Imprecision (wide confidence intervals) - Heterogeneity that is real but not understood (e.g. effects differ by age / setting) - Rapidly changing systems (policy / practice changes make older data less informative)

Absence of- No relevant studies for the claim /‘there is currently
evidencepopulation / outcomeinsufficient evidence to
- Evidence exists but is too biased orconclude...’
indirect to support any inference‘we found no robust
- Data are systematically missing (noevidence addressing...’
linkage, no follow-up, no‘this remains an
denominators)important evidence
- Only anecdote / opinion / single casesgap…’
or uncontrolled observations, when the
claim requires more

3.7 The limits of the evidence base

Some of the Review’s conclusions concern what cannot yet be known, because the absence of evidence is itself a finding. The largest gap is the near-total lack of routine national data on severity, functional impairment and outcomes among those being diagnosed. This means that the question at the centre of public debate, whether and how far diagnostic thresholds have shifted over time, cannot presently be settled, since the data needed to settle it are not collected. Furthermore, there are no recent large-scale diagnostic surveys for adult ADHD, and co-occurrence of conditions is poorly captured in structured data. No survey of children and adolescents has been carried out using the same tools such as DAWBA since 2017, and that covered 5- to 15-year-olds, not younger children. Fragmented data across health, education and welfare make it difficult to follow individual trajectories. These gaps define the confidence with which some of the current data can be interpreted and inform the Review’s recommendations about data and measurement – which can be summarised as ‘commission these studies now’ (Chapter 20).

3.8 UK data and international comparison

The Review’s remit is England, and its core analysis rests on English population, administrative and service data. However, some of the datasets also included the Devolved Administrations and we indicate when that is the case. International comparisons, drawn from sources such as the Global Burden of Disease study[38] and comparative epidemiology, are used only for a specific purpose: to show that similar underlying pressures can produce markedly different patterns of diagnosis, demand, and support depending on how systems are organised. These comparisons illuminate what makes England’s situation distinctive, while recognising the limits of transferring findings across different systems of language, history, health, education, and welfare. Specifically, it is unwise to use such studies to compare simple prevalence figures.

3.9 Bringing it together: triangulation and holding distinctions

No single source being sufficient, the Review reached its conclusions by triangulation, seeking consistency across many forms of evidence. Where independent sources converged, confidence grew. Where they diverged, the Review treated the divergence as itself informative: the gap between epidemiological and administrative prevalence provided an opportunity to further analyse the relationships between recognition, access and recording. Central to the Review’s approach was distinguishing five concepts that public debate often conflates: population prevalence, psychological distress, recorded diagnosis (administrative prevalence), service demand and underlying need – and from these, three separate lines of enquiry: what is changing in the population, what is changing in systems, and how the two are related.

Chapter 4. Language, concepts and classification

This chapter discusses the terms that are used throughout the Review: prevalence, diagnosis, need, functioning, impairment, participation, neurodiversity, difference and disability.

Language is contested and the Review has tried to handle it with care. In wider discussions around prevalence, the same words are often used to mean different things. Without clear shared definitions – as outlined here – disputes around the evidence cannot be meaningfully resolved and solutions developed.

4.1 Why language matters

Terms such as distress, disorder, diagnosis, need, impairment, functioning, medicalisation and support appear constantly in public debate, policy, research and professional discussion, though rarely with a shared understanding of what precisely each denotes. The result is that people argue past one another – answering different questions in the belief that they are addressing the same one. For example, if a waiting list is read as a measure of prevalence, rising demand will be mistaken for rising illness. If the presence of symptoms is taken to establish disorder, ordinary and understandable distress may be drawn into clinical pathways that were designed for something else. When we treat diagnosis as the same thing as need, systems tend to organise themselves around diagnosis.

It is important to set this within the larger change in the understanding and use of language of the past two decades. The rise in mental health literacy (the public’s working knowledge of psychological concepts) has been the result of profound social and cultural changes, aided by active campaigning[39, 40]. Greater openness has reduced stigma for some, bringing many more people to seek help, and allowed groups whose needs were long overlooked – among them many women and girls, older adults, and people from some minoritised communities – to be recognised and supported. The Review affirms this without reservation.

The expansion of vocabulary has had a second effect. Concepts that began in clinical practice have entered everyday speech, where they are used with different meanings and for different purposes. Thus the language of diagnosis is now enlisted not only to describe illness but to explain identity, difference, adversity and ordinary experience.

4.2 Symptoms, distress and disorder in mental health

The first distinction the Review draws, and on which much else depends, is between symptoms, distress and disorder. These are closely related and routinely conflated.

Symptoms are individual experiences or observable features: low mood, anxiety, poor concentration, inattention, disturbed sleep, irritability, impulsivity, intrusive thoughts or differences in social communication. They are extremely common in the general population, and often intertwined with everyday human experiences[41]. Most people pass through periods of anxiety, sadness or distraction, particularly at times of loss, uncertainty or change. Their presence alone establishes neither illness nor a need for clinical intervention.

Distress, by contrast, is the subjective experience of suffering or difficulty – the extent to which symptoms or circumstances are felt as painful, overwhelming or hard to bear. It too is an inescapable part of human life, accompanying bereavement, disappointment, insecurity and illness, and much of it resolves with time, support and changed circumstance. But distress should never be dismissed because it is common. The key distinction is between distress that can reasonably be expected to ease through ordinary recovery and distress that becomes persistent, severe or impairing enough to warrant clinical attention – a line that is often blurred, and that depends on the interplay of symptoms, duration, severity, context, risk and effect on daily life.

Disorder is a further step. The ICD-11 (International Classification of Diseases, 11th Revision) and DSM-5 (Diagnostic and Statistical Manual of Mental Disorders, Version 5) do not define disorder by the presence of symptoms alone, but require that symptoms be persistent, occur in characteristic patterns, and be associated with clinically significant distress or impairment in important areas of functioning[42, 43]. A diagnosis reflects a judgement that difficulties have crossed a threshold. That judgement is unavoidably both scientific and clinical, because the underlying characteristics do not fall into naturally distinct groups. Rather, anxiety, mood, attention and social communication are distributed continuously across the population. Dimensional language is often used to describe this (blood pressure is a familiar example), and it recurs throughout this Review. But it is necessary for our diagnostic systems to impose categorical boundaries on dimensions for practical reasons: to guide treatment, enable communication, structure research and organise access to expertise. The categories retain great practical value, but they should not be mistaken for natural boundaries, and this tension between a dimensional reality and a categorical classification is widely recognised.

Two consequences follow. The first is that a population may grow more symptomatic, or more distressed, without necessarily a proportionate rise in diagnosable disorder, and, conversely, that changes in recognition, referral or diagnostic threshold may raise diagnoses without any change in the underlying distribution of symptoms. The second concerns functioning: people with very similar symptoms may differ enormously in how far those symptoms disrupt their lives, because functioning reflects not symptom severity alone but the interaction between a person, their environment, their supports and the demands upon them. This resists two equal and opposite simplifications: that all distress is evidence of disorder requiring diagnosis, and that distress falling short of a diagnostic threshold is somehow less real or less deserving of help. Symptoms describe experiences; distress describes suffering; disorder describes a clinically significant and impairing pattern of the two. None of them, by itself, establishes what a person needs.

4.3 Neurodevelopmental conditions, neurodiversity, difference and disability

Neurodevelopmental conditions describe differences in the development of the brain and nervous system that are present from early in life, though they may not become apparent immediately, and that are associated with enduring but changing patterns of cognition, communication, behaviour or learning. Alongside autism and ADHD, they include developmental language disorder, intellectual disability, tic disorders, dyslexia, dyspraxia and others, and, as Chapter 9 shows, they co-occur with one another, and with mental and physical health conditions, far more often than they appear in isolation. Like the characteristics underlying mental disorder, the traits associated with autism and ADHD are continuously distributed across the population – some people with more, some with less – rather than marking off a discrete group, so that here too, a diagnostic line is drawn across a continuum[44, 45].

The concept of neurodiversity, the idea that variation in brain function, cognition and behaviour is part of natural human diversity rather than pathology and disorder, emerged in the late 1990s and has reshaped the public discussion of autism and ADHD, with the related terms neurodivergent and neurotypical now in wide use[46, 47]. A neurodevelopmental difference is not in itself a disorder; whether and how far it becomes disabling depends on the fit between a person and the environment in which they are expected to function[48]. This is the insight formalised in the social model of disability and in the World Health Organization’s biopsychosocial framework, in which disability arises from the interaction between an individual and their surroundings (educational, social, sensory, occupational) rather than residing wholly within the person[29, 49]. The same person, autistic or with ADHD, may be able to participate in one setting and struggle profoundly in another; the difference lies as

much in the environment as in the individual, although environment alone is not the whole story for people with severe or profound learning disability and/or autism. The law reflects this logic: under the Equality Act 2010 a person is disabled if a physical or mental impairment has a substantial and long-term adverse effect on their ability to carry out normal day-to-day activities, so that a neurodevelopmental condition may amount to a disability[50].

Language use can be contested. Many autistic people and autism organisations prefer identity-first language (‘autistic person’ rather than ‘person with autism’) and prefer that autistic characteristics be described as features or traits rather than symptoms; others, including many parents, prefer person-first language. The evidence shows that no single term commands universal assent[51, 52]. The Review avoids implying that any one usage is correct for everyone, aims to be precise and respectful, and describes specific difficulties rather than reaching for global labels. The Review also recognises that the neurodiversity framing has been valuable and, for many, liberating; it has reduced stigma and enabled self-understanding.

In seeking a diagnosis, many people aren’t looking for a ‘label’ but for a credible account of how they think, learn, regulate attention and emotion, communicate, relate to others, and perform at their best. They want greater agency in describing their needs. They increasingly expect schools, employers, and health services to respond to them as individuals rather than as approximations of an average person. If that interpretation is correct, expanding diagnostic capacity, although necessary, cannot be the whole long-term answer. A broader language and infrastructure for understanding difference before, alongside, and beyond diagnosis, is also needed.

This begins by treating neurodiversity as a universal property of human populations rather than as a description of a defined minority. Every person has a distinctive pattern of cognitive, emotional, behavioural, and sensory characteristics. Clinical diagnoses remain important, but only in identifying thresholds of impairment or need within a much wider landscape of variation. The social neurodiversity paradigm recommends using the term ‘condition’ over ‘disease’ or ‘disorder’; ‘comparison group’ not ‘compared with normal’; and ‘challenges’ or ‘needs’ over ‘deficits’[53]. However, while we find this description of neurodiversity to have many benefits, some populations have experienced increased levels of judgement because of this changing narrative and there is a legitimate counter-argument that overly neutral language can obscure the suffering, impairment and need for evidence-based support experienced by some people with ADHD and autism[54]. In particular, the reality for some autistic people, particularly those with severe or profound intellectual disability, is that they have life-long very high support needs and experience serious, sometimes life-limiting, impairment. A framework adequate to these conditions needs

to honour difference and take impairment seriously – a balance the Review returns to throughout.

The Review has to make a decision and has concluded that we will continue to describe conditions such as common mental disorders, autism and ADHD – and symptoms and traits – rather than using the term ‘neurodiversity’.

4.4 Why clearer distinctions matter for policy and practice

These distinctions are not academic refinements; they change how evidence is read, how problems are recognised, and which solutions become thinkable at all. Much of the confusion in current debate can be traced to four key conflations. Firstly, when administrative prevalence is treated as though it were population prevalence, the ordinary workings of recognition and recording get mistaken for changes in the health of the population. Secondly, when administrative diagnosis is read as a direct measure of need, a record of recognition and categorisation is mistaken for a measure of what support a person requires. Thirdly, when a waiting list is taken as a measure of how common a condition is, a sign of pressure within a system is misread as a fact about the population it serves. Fourthly, when diagnosis is used by systems as a proxy for support need, the door to help is placed behind a clinical judgement that was never meant to guard it.

Distinguishing the concepts also changes how a successful system is measured. If diagnosis is the principal measure of performance, attention flows to waiting times, numbers assessed and diagnoses recorded. If the goal is instead framed as improved functioning and participation, different questions need to be asked and answered: are people staying in education, holding on to work, sustaining relationships, and living the lives they want? These are the outcomes that people with lived experience most often say they care about. However, no amount of conceptual clarity will resolve the pressures on services, which are also matters of workforce, funding, inequality and the wider determinants of health. Clearer thinking is the ground on which those harder problems can be addressed, and a system which can tell these concepts apart is better placed to respond intelligently to complexity than one that cannot.

4.5 Working definitions used throughout this Report

Diagnosis is vital to modern healthcare. It gives clinicians a shared language, guides evidence-based treatment, supports research and prognosis, and, as people with

lived experience told the Review repeatedly, offers many individuals and families an explanation for experiences that had seemed confusing or shaming; a moment of sense-making that can itself be a relief. The Review affirms the continuing importance of accurate, timely, high-quality diagnosis; nothing in this Report should be read as diminishing it.

But as emphasised throughout, people who share a diagnosis often differ markedly in their circumstances, functioning, risks, symptoms, aspirations and support needs. The Review therefore keeps five terms distinct:

Diagnosis – the process of identifying the most likely explanation for a person’s difficulties according to internationally recognised criteria for diagnostic categorisation; a basis for communication, treatment and research.

Impairment – limitations associated with a health condition or developmental difference that affect functioning; understood in interaction with the environment rather than as a fixed property of the individual, and which may or may not result in a disability (see below).

Functioning – describes how a person carries out everyday activities and roles (learning, communicating, self-care, emotional regulation, relationships, work, parenting) and reflects the interaction between the person and their environment, so that two people with similar symptoms or impairments may function very differently.

Participation – a person’s involvement in education, employment, family life, relationships and community. This is the outcome that functioning serves, and the one that people with lived experience most often said mattered to them, thus what support should seek to improve.

Need – the gap between an individual’s status and the support required to improve health, functioning or participation at a particular time; shaped by symptoms and diagnosis but also by functioning, context, risk, co-occurring conditions, goals and available support; dynamic rather than fixed.

Because a diagnosis is often the most reliable route to an educational adjustment, workplace accommodation, welfare support or specialist care, it can come to serve not only as a clinical explanation but as an administrative key. Where that happens, a clinical judgement is asked to carry an administrative load it was never designed for.

How far this has come to characterise the systems examined here, and with what consequences for pressures on diagnostic pathways, is one of the central empirical questions of the Review, taken up across all chapters but particularly in Chapter 12. Diagnosis and need answer different questions, and a system that treats them as one will struggle to respond to the diversity of people’s lives.

To keep use of important concepts consistent, the Review adopts the working definitions below. They are not intended to displace established clinical or scientific usage, but to provide a common framework for interpreting the evidence and recommendations that follow. A fuller glossary appears in the Annex.

Overdiagnosis – when someone is given a diagnosis for a medical condition they do not have to explain the difficulties and impairment they are experiencing[55].

Misdiagnosis – when someone has a condition, or several conditions, but is given the wrong diagnosis.

Under-recognition and underdiagnosis – these terms are closely linked; the former is the precursor to the latter. Under-recognition refers to a condition not being recognised as a possible explanation for a person’s difficulties. Underdiagnosis occurs when someone has a condition but does not receive a diagnosis when they should.

Population (epidemiological) prevalence – the proportion of the population who are estimated to have a condition, or to be above a pre- defined level of symptoms at a given time, whether or not they have sought help or been diagnosed; estimated principally through surveys of a defined population.

Administrative prevalence – the proportion of people recorded as having a condition within a health, education or other administrative system; reflects recognition, access, diagnosis, recording practice and eligibility as well as underlying need, and is not equivalent to population prevalence.

Incidence – the number of new cases arising in a defined population over a specified period.

Symptoms – subjective experiences which may be accompanied by observable manifestations (signs), such as anxiety, low mood, or inattention. Many vary dimensionally across the population. Their presence alone does not establish a disorder; clinical significance

generally depends on their pattern, severity, persistence, context, associated distress or functional impairment, and relevant diagnostic exclusions.

Distress – the phenomenological experience of emotional suffering and the extent to which symptoms or circumstances are felt as painful, overwhelming or hard to bear; a normal part of human life, which may accompany but is not synonymous with disorder.

Mental disorder – a clinically significant, persistent pattern of symptoms that gives rise to significant levels of distress and/or impairment that meets recognised diagnostic criteria.

Neurodevelopmental conditions – differences in how the brain and nervous system develop that are present from early life, though not always apparent immediately, and shape enduring patterns of cognition, attention, communication, behaviour and learning. They exist through complex genetic, biological and environmental factors. They include ADHD and autism, alongside developmental language disorder, intellectual disability and tic disorders.

Neurodiversity – the concept that variation in brain function, cognition and behaviour (including variation associated with conditions such as autism and ADHD) is part of natural human diversity rather than pathology and disorder; neurodivergent and neurotypical are the associated terms.

Disability – under the Equality Act 2010, a physical or mental impairment with a substantial and long-term adverse effect on the ability to carry out normal day-to-day activities.

Demand – the expression of need through requests for help, referrals, assessments or service use; reflects awareness, help-seeking, accessibility and system design as well as underlying need.

Support – the full range of interventions intended to improve health, functioning, acquisition of skills and participation, extending well beyond healthcare to include educational adjustments, family and community support, workplace accommodations and practical assistance.

Treatment – clinical interventions intended to improve health outcomes through established therapeutic approaches; one component of support, not a synonym for it.

Medicalisation / Professionalisation – the process by which experiences once understood in social, educational or personal terms come to be interpreted through medical or other professional clinical frameworks; used descriptively and recognised as having both beneficial and unintended consequences depending on context.

4.6 Chapter synthesis

Much of the apparent disagreement surrounding mental health conditions, ADHD and autism arises less because the evidence is contradictory than because different concepts are treated as though they were one. Symptoms, distress, disorder, diagnosis, need, functioning, participation and support each describe a different aspect of a person’s life; they are closely related, but they answer different questions. Running them together has produced not only muddled debate but, at times, systems less coherent than they might be.

The purpose of drawing these distinctions is not to erect walls between concepts that inevitably overlap, but to show how they relate without collapsing into one another. Each concept says more when read alongside the others.

The chapters that follow put these terms to use to ask whether support might be better organised to align diagnosis, need, functioning and participation more closely with the lives people are trying to lead. Language shapes understanding, understanding shapes systems, and systems shape the help people can reach.

Chapter 5. Overdiagnosis, misdiagnosis, under-recognition, underdiagnosis and medicalisation

Debates about mental health, ADHD and autism diagnoses are too often framed as ‘overdiagnosis vs underdiagnosis’, when the reality is more complex: overdiagnosis, misdiagnosis, under-recognition, and underdiagnosis can all coexist within the same healthcare system.

Diagnosis is important because it can provide an explanation for someone’s problems and validation of their experiences, guide treatment and open access to support. But being diagnosed may also carry risks, including stigma, narrowed expectations, diagnostic overshadowing, inappropriate pathways, and potential ‘nocebo’ effects (where a person experiences negative symptoms simply because they expect harm, even when the intervention, or in this case label, is inert) when labels are communicated poorly or applied too broadly.

The issue is not just whether there is too much or too little diagnosis, but whether diagnosis is being used appropriately. Moreover, as with all medical interventions, we must ask: is a diagnosis doing more good than harm?

This chapter addresses one of the most contested aspects of the current debate. Rather than asking whether there is ‘too much’ or ‘too little’ diagnosis, it considers a broader set of questions: how diagnoses are made, where uncertainty arises, why overdiagnosis, misdiagnosis, under-recognition and underdiagnosis occur, and how the organisation of health, education and wider public systems uses diagnosis not just to guide treatment but for a wide variety of purposes for which it was never intended. Throughout, we start from the premise that people’s distress and difficulties are real experiences, even though the ways in which they are classified and responded to are not always well aligned with what they need.

Like any clinical intervention, diagnosis has potential benefits and potential harms. For many people it provides understanding, guides treatment and opens access to support. In other circumstances it may prove less helpful, particularly if it delays more appropriate forms of support, is the wrong or incomplete diagnosis, narrows what they and those around them believe is possible, and defines people by what they cannot do. So rather than focus on under versus over diagnosis, we should

instead be asking whether, in any particular context, the likely benefits outweigh the possible disadvantages.

5.1 Why the current debate is insufficient

5.1.1. Overdiagnosis vs underdiagnosis: conceptual points

Debate about the growth in mental health, ADHD and autism diagnoses is often framed as a binary choice between over and underdiagnosis. This is incorrect. ‘Overdiagnosis’ – the most commonly used term in popular and media discourse, but also the most contentious – means different things to different people.

Awais Aftab[56] identified a range of distinct concerns gathered under this single label, including diagnostic error, poorly performing diagnostic instruments, expanding diagnostic thresholds, the diagnosis of transient or subthreshold difficulties, concern about the medicalisation of ordinary experience, and unease about the growing visibility of mental health conditions.

To these we can add the anger and upset felt by our Lived Experience Partners about the way debates about overdiagnosis are often associated with framings that minimise, trivialise or deny lived distress and challenges, or lead to simplistic and negating conclusions; for example, ‘it’s become too easy to get a diagnosis’.

By contrast, a survey of 752 GPs found considerable concern about the rising number of mental health diagnoses, with 442 (58.8%) believing that mental health disorders were either being overdiagnosed a little (259, 34.4%) or overdiagnosed a lot (183, 24.3%).

Figure 5.1 Responses to a survey of 752 GPs answering the question ‘On balance do you think mental health problems are being over-diagnosed or under-diagnosed?’

Source: BBC questionnaire, September 2025[57] While being aware of the above views, overdiagnosis as a term can be legitimately used in three common scenarios. First, diagnosing something asymptomatic that would never progress if left alone: Monoclonal gammopathy of undetermined significance (MGUS) is a non-cancerous condition found in 3% to 5% of people over age 50, and most people live with it without ever developing health problems. However, receiving the diagnosis can generate intense anxiety and a ‘cascade’ of unnecessary invasive testing with sometimes serious side effects[55].

Second, overdiagnosis is one of the well-known risks of screening programmes. For example, when a particular test starts to wrongly identify increasing numbers of people as either being at risk of, or having, a cancer (‘false positive’), leading to adverse psychological and physical consequences.

Third, and overlapping with the term ‘medicalisation’, is labelling genuine symptoms and difficulties as a medical disorder requiring a medical solution when they could be better managed in other ways. For example, a student experiencing marked loneliness but not major depression could be prescribed antidepressants when a better option would be anything that increases their social networks. In these scenarios there are unequivocal symptoms and/or difficulties, but they have been medicalised to their detriment.

All these different but legitimate uses of the word overdiagnosis share one thing in common. In none of these is the patient at fault: neither the person told they have an immune abnormality of little consequence, nor the person given an inaccurate cancer screening result, nor the person prescribed antidepressants when there are better and safer options. These are all system or professional failings, not personal ones. These are different arguments, and each raises scientific, clinical and ethical questions that arise across many areas of health and are not unique to mental health and neurodevelopmental conditions. Social media is a further complication: public interest in a condition can rise sharply following media coverage or a celebrity disclosure, heralding a wave of referrals.

Other factors particularly relevant to the Review include the influence and use of psychiatric language as part of everyday life, how the boundary between distress and disorder is negotiated, how thresholds are determined for dimensional conditions, or how classification systems evolve over time; the increasing role of commercial interests in diagnosis, and how diagnosis has assumed an increasingly central role across health, education and wider public systems.

Concerns about overdiagnosis must also not obscure under-recognition and under- diagnosis. Many people with substantial difficulties have historically remained unrecognised, such as girls and women with autism or ADHD, children and young people in Children and Young People’s Mental Health Services (CYPMHS), those in care and the secure estate, people from some minoritised communities, or adults whose difficulties were not identified during childhood,[10, 58-60].

5.1.2 The evolution of diagnosis

Part of the reason why these debates are difficult to resolve is that diagnostic systems are continuously evolving. There is no area of medicine in which diagnostic categories have stood still; revision is the rule rather than the exception, and psychiatry has been especially prone to expansion. The first attempt to classify mental illness for the US Census in 1880 used seven categories; the American Psychiatric Association’s early nomenclature in 1917 listed 22; the first Diagnostic and Statistical Manual (DSM), in 1952, contained 128; by 1980 the much-discussed DSM-III listed 227, and by 2013 DSM-5 reached 541 – a total that looked smaller on the shelf only because it was printed on thinner paper.

We like to think that such changes are driven by advances in scientific knowledge and clinical understanding, but this is not always the case. Diagnostic thresholds and classifications have changed across many branches of medicine, reflecting changing evidence, but also changing concepts of illness, societal changes and clinical priorities. Nor are debates about overdiagnosis confined to psychiatry. An annual international conference on overdiagnosis co-organised by Oxford’s Centre for

Evidence-Based Medicine[61] is dominated by presentations from other areas of medicine, illustrating that questions about thresholds, uncertainty and appropriate intervention extend well beyond mental health.

Diagnosis is also more complex than public debate often assumes. It is not a fixed or permanent statement about a person’s future and even where lifelong impairment is part of the diagnosis as with severe and profound learning difficulties, individual future functioning is variable. Many conditions improve, some remit completely and many fluctuate across the life course. Diagnosis cannot by itself convey prognosis or future participation. The meaning attached to diagnosis depends on how it is understood and how it is used, themes returned to in this chapter around identity and in Chapter 16 around work and participation.

5.1.3 Self-recognition, identity and self-understanding

Another part of the current landscape is self-identification, which is not the same as diagnosis. The OxWell survey found that around one in five pupils identified themselves as neurodivergent, while only around half of those identifying as having autism and around one quarter identifying as having ADHD reported having received a formal diagnosis[14]. Polling commissioned for this Review found that self- identification among younger adults exceeded commonly cited prevalence estimates and that many who believed they had ADHD or autism had not yet received professional confirmation. Thus, many people experience difficulties they recognise in themselves while remaining outside the formal diagnostic system.

Figure 5.2 Percentage of students by sex and school year group self-reporting that they have ADHD, autism or are neurodivergent in any other way

Source: OxWell 2025 Summary Report[14] Figure 5.3 Percentage of respondents self-identifying by condition and age group Source: General population survey conducted for the Review, commissioned in collaboration with the King’s College London Policy Institute and Focaldata. 95% confidence intervals shown.

Many clinicians, especially general practitioners, report that people increasingly seek assessment not only for treatment or support but also to better understand themselves[62].

Diagnosis has, in this sense, become closely connected with identity, belonging and self-interpretation. For some people, particularly within the autistic community, diagnosis becomes part of how they understand themselves rather than simply something they have[63]. It may also open access to communities built around shared experience, offering recognition, mutual support and collective advocacy[64, 65].

People may adopt diagnostic language without formal assessment because they face long waits for services, identify with neurodevelopmental perspectives, are dissatisfied with existing pathways or are searching for explanation and belonging[66]. Social media has amplified these developments by providing unprecedented access to information, personal narratives and communities organised around shared experience[67, 68]. This has clear benefits, particularly where it helps people recognise previously overlooked difficulties. It also carries risks where information is incomplete, inaccurate or lacks clinical context.

5.1.4 Diagnostic systems: ICD-11 and DSM-5

Contemporary diagnoses of mental health and neurodevelopmental conditions are based on one of two classification systems: the International Classification of Diseases[42], produced by the World Health Organization, and the Diagnostic and Statistical Manual of Mental Disorders[43], produced by the American Psychiatric Association. Together they provide the common language through which professionals describe patterns of difficulty. Classification supports communication, research, treatment, service planning and the monitoring of population trends. Without shared diagnostic frameworks it would be impossible to organise services, evaluate interventions or improve scientific knowledge.

The two systems, ICD and DSM, overlap. Both describe conditions in a broadly similar way and draw on much the same evidence base. DSM generally provides more detailed operational criteria and symptom thresholds, whereas ICD places greater emphasis on clinical judgement, context and functional impairment. Historically, DSM criteria have also been broader in some neurodevelopmental conditions, particularly ADHD and autism. These differences may contribute to variation in prevalence estimates and patterns of diagnosis. ICD is the classification used in the UK and most other countries; DSM dominates in the USA, where it is central to the ways in which the health insurance system defines and treats illnesses.

‘True’ underlying current population prevalence in the UK of ADHD is unclear and prevalence estimates rely on pooled international studies and consensus (Chapter 7). Variation in prevalence between countries points to the importance not just of diagnostic criteria but of socio-cultural expectations of appropriate child behaviour, which influence how symptoms and behaviours are recorded or framed, and which

apply as much in this country as anywhere else. This makes deciding what constitutes under- or over-diagnosis more challenging than in some other disorders.

Clinician judgement is key to diagnosis across medicine, and not simply a ‘readout’ of a set of biological data. Yet judgement is not exercised in a vacuum: it operates within the pressures, needs and narratives of the society in which clinicians work[69,

70]. This is important because a diagnosis of ADHD or autism is not made from a blood test or a brain scan, but by clinical judgement based on a defined pattern of symptoms being present across more than one setting, persisting over time, evident in childhood, and associated with significant functional impairment. Because both symptoms and impairment lie on continua, different clinicians in different settings will inevitably reach different conclusions about whether a threshold has been met; and a diagnosis, once made, still does not describe function or need.

Successive revisions of DSM broadened eligibility for ADHD by modifying age-of- onset criteria, reducing symptom thresholds for adults and recognising that ADHD frequently co-occurs with autism rather than excluding it[59, 71, 72]. Each change had a stated scientific or clinical rationale. Together, they increased the number of people eligible for diagnosis and altered the characteristics of the diagnosed population.

A similar process has occurred in autism, with subtypes removed and all presentations brought within a single spectrum, making wider diagnostic interpretation possible although that was not the intention. Earlier descriptions of autism focused primarily on people with marked communication difficulties, often alongside intellectual disability. Subsequent revisions recognised the much wider range of autistic presentations, including people with fluent language, average or above-average intellectual ability, and those whose difficulties become apparent only as social demands increase[13, 73, 74].

Diagnostic systems have always reflected an interaction between scientific evidence, clinical observation and wider social understanding. Conditions have been added, revised and occasionally removed as knowledge has developed and societal attitudes have changed. Post-traumatic stress disorder (PTSD) emerged from the actions of committed clinicians who had opposed the war in Vietnam. They believed that the psychological consequences of war were not just short-term (which was accepted) but long-term as well. Their efforts were rewarded when PTSD was entered in a rush into the DSM in 1980, but epidemiological evidence to support this did not appear for another decade. This does not mean that PTSD was invented. The point is that such changes are not driven solely by the dispassionate march of science. ‘Diagnoses are concepts, not things…’; they evolve because understanding evolves. Future evidence will almost certainly refine current classifications, just as previous generations revised those that came before.

A further challenge is that many of the difficulties these systems seek to classify do not fit neatly within the categories they define. As Chapter 9 demonstrates, co- occurrence between mental health, neurodevelopmental and physical health conditions is common. Symptoms frequently cross diagnostic boundaries and many people experience several overlapping difficulties simultaneously. Classification simplifies a reality that is often more complex than any single diagnosis can capture. Pathways organised around one condition alone may overlook important aspects of a person’s needs. Chapters 6 to 9 examine population and administrative prevalence of mental health, ADHD and autism, and the levels of co-occurrence between these conditions.

5.2 Overdiagnosis and under-diagnosis

Debates about diagnosis become polarised because they are framed as though only one explanation can be true: either conditions are overdiagnosed or under- recognised, either thresholds are too broad or too narrow, either services identify too many people or too few with the assumption that under-recognition and overdiagnosis are mutually exclusive. In practice they frequently coexist and both are to some extent inevitable for statistical reasons[75].

The Clinical Practice Research Datalink (CPRD) provides a clear picture of administrative prevalence in the UK. Historically, recorded ADHD administrative prevalence lay below estimates from epidemiological population studies, consistent with substantial under-recognition[59, 76]. An epidemiological principle is that as administrative prevalence starts to approach the expected ‘true’ population rate, the numbers of those still undiagnosed with ADHD will steadily decrease, so the rate of increase will first slow down, and then level off, albeit still below the accepted prevalence ceiling, since not everyone who fulfils the criteria for a given disorder will want assessment, diagnosis or treatment. This levelling off has yet to happen for ADHD (or autism), where the rate of increase shows no sign of slowing. We also know that current administrative figures still underestimate total demand because they do not yet capture most people on waiting lists or many assessed through the independent sector, but what is clear is that administrative prevalence has increased rapidly in this country and, in some age groups is approaching or may soon exceed the best estimates derived from population studies. The epidemiology suggests that underdiagnosis is no longer sufficient on its own to explain current trends.

Important under-recognition gaps nevertheless remain (for example, older adults and some minoritised ethnic groups). The same broad pattern is evident for autism. Historical under-recognition among women and girls, adults and people without intellectual disability is increasingly being addressed[10, 58]. At the same time, the diagnosed population has become more heterogeneous, raising legitimate questions

about where diagnostic boundaries should be drawn, examined further in Chapters 7, 8, and 9 and the next section.

5.3 Misdiagnosis: where and how it arises

Public discussion often treats misdiagnosis as though it were simply a matter of error: one condition mistaken for another, or a diagnosis that is either right or wrong. Such mistakes do occur. However, many of the important forms of misalignment arise because clinicians need to make binary decisions (and therefore categorise) problems that are complex and multifactorial. Diagnostic systems necessarily introduce thresholds for both measures and symptoms to guide decisions. These thresholds are essential, but remain conventions that support decision-making rather than natural divisions within human biology or experience. A challenge, shared across much of medicine, is deciding where to draw thresholds. This decision is not solely a scientific one, but one that involves societal processes that relate to prevailing norms and values[77].

Uncertainty is an inherent feature of clinical assessment. People whose difficulties lie close to current thresholds may be assessed differently depending on when they are seen (see 5.10 for a worked example), the information available, the settings in which difficulties occur, the range of problems they have, and the purpose of the assessment. In helping clinicians understand and respond to individual need, diagnosis is truly an art as well as a science.

Autism and ADHD illustrate these challenges. Diagnostic thresholds represent practical points along, for autism, a continuum of social interaction and communication skills, and of restrictive and repetitive patterns of interest and behaviour; and for ADHD, attentional, hyperactive and impulsive behaviours, rather than distinct biological boundaries[45]. Clinicians inevitably differ in how they interpret borderline presentations, particularly where symptoms are evident but impairment is less clear[78]. Where services are under pressure, there may also be less opportunity to explore these issues in sufficient depth, all contributing to underdiagnosis or misdiagnosis at the individual level.

Assessment of impairment presents a related challenge. Impairment is central to all diagnoses but is not a fixed property residing solely within the individual. It emerges through the interaction between a person’s characteristics and the environment in which they live, learn and work. Reports from individuals, parents, teachers and employers differ substantially because demands differ across settings. Many assessment tools were originally developed for school-aged children and may be less sensitive to presentations in adults, women and girls, or people whose

difficulties emerge primarily in occupational or social contexts. Judging impairment requires contextual understanding as well as symptom counts.

The growing diversity of people presenting for assessment adds further complexity. Individuals referred for autism and ADHD assessment now differ widely in symptom profile, functional impact, co-occurring conditions and life circumstances.

The challenge becomes greater once co-occurrence is recognised. As Chapter 9 demonstrates, neurodevelopmental and mental health conditions frequently overlap, and many people present with patterns of difficulty spanning several diagnostic domains[79-81]. The clinician’s task has shifted from identifying a single ‘correct’ diagnosis to assessing the relative contributions of several. Many characteristics, including attentional difficulties, emotional dysregulation, social communication differences and intolerance of uncertainty, occur across diagnostic categories rather than belonging uniquely to one condition. Some people therefore experience delayed recognition because one aspect of their presentation overshadows another, while others accumulate several diagnoses that each describe part of a broader pattern. Diagnostic uncertainty often reflects the complexity of what is being assessed rather than shortcomings in clinical practice alone[79, 81].

This also helps explain why diagnostic overshadowing remains so persistent. Once one diagnosis has been established, subsequent difficulties may be interpreted primarily through that existing framework. Autism may be overlooked because attention is focused on anxiety or depression; ADHD may be obscured by emotional or behavioural difficulties; equally, mental health conditions may receive insufficient attention once a neurodevelopmental diagnosis has been made[82-84]. In such circumstances, the diagnosis is not wrong but incomplete. The person’s needs and treatment options arise from the interaction of several difficulties, strengths and circumstances rather than from any single condition or guideline.

Thus the same system may fail to recognise some people whose needs are substantial while diagnosing others whose difficulties lie close to current thresholds. Margaret McCartney described this as the ‘patient paradox’: overtreatment of some occurring alongside undertreatment of others, influenced by wider factors including unequal access to care and, at times, commercial pressures. Such pressures may stem, for example, from the pharmaceutical industry[85, 86], but may also be the consequence of private healthcare. An example is fee-for-service healthcare, in which physicians have been found to carry out more procedures, and to use more severe diagnostic codes, where there are financial incentives for the institution they work for[87, 88].

5.4 Medicalisation, professionalisation and the impact of a diagnosis

Discussions of mental health and neurodevelopmental conditions are often accompanied by concerns about medicalisation, a term usually used critically, suggesting that ordinary experiences are increasingly being understood as medical problems. It is unfortunately rarely used to describe the opposite – how medicine can be ‘The Greatest Benefit to Mankind’[89].

In many ways the fact that people are becoming more interested in using the language of mental health in everyday discourse is a good thing. But difficulties are more frequently framed now in terms formerly more restricted to professional discourse and diagnostic labels are viewed as part of personal identity[90]. An expansion of mental health descriptors is used to describe phenomena well beyond the original meaning of a term, trauma being a characteristic example[91]. Thus diagnostic concepts are encountered not only in healthcare but also in schools, universities, workplaces, public discourse and social media.

While some of the above changes have brought substantial benefits, debates about medicalisation can at times unhelpfully obscure those benefits, particularly where pharmacological treatments are concerned. For example, despite widespread concerns about ‘medicalising childhood’ implying negative impact, children diagnosed with ADHD actually reported experiencing increased self-understanding and agency when taking stimulant medication[92]. Qualitative research consistently shows that many people experience diagnosis as validating, relieving and explanatory[93-96]. For many people it provides an explanation for experiences that previously felt confusing or isolating, reduces self-blame and offers a language through which difficulties can be understood and communicated. This Review’s Lived Experience Partners described correct diagnosis as providing explanation, treatment direction, prevention, personal understanding and access well beyond clinical services.

“I felt as though I was firefighting a never-ending storm of mental health issues, without being able to see what was fuelling them or having the tools to put them out. Diagnosis helped me recognise that I had been attempting to manage an extremely dysregulated nervous system in environments that neither understood nor accommodated my needs.

Now that I am diagnosed, I am better able to identify and apply adjustments and solutions that are effective and appropriate. Understanding the difference between burnout and a depressive episode was huge for me, as was

distinguishing anxiety from overstimulation: each requires a very different response.

Diagnosis has also given me the language to separate out my experiences and communicate them more effectively to others. Members of my family can now recognise the warning signs of a meltdown and help prevent it from escalating, whereas previously it seemed to them to appear from nowhere.”

Lived Experience Partner Greater professional understanding has transformed many such experiences. People who might previously have internalised blame are now more likely to receive explanations that are compassionate, evidence-informed and capable of guiding effective support. The improved recognition of autism in women and girls[58] and ADHD across the lifespan[97] has enabled many people to access help that was previously unavailable.

Much attention has rightly been given to the placebo effect of diagnosis and treatment: the improvements associated with hope, expectation and confidence in treatment. Less attention has been paid to its counterpart, the nocebo effect, whereby expectations of harm or limitation may themselves contribute to poorer outcomes for those diagnosed. Across medicine, beliefs about illness and treatment are known to influence symptoms, behaviour and recovery. Similar mechanisms may sometimes influence how people understand the implications of diagnostic labels. These issues are considered further in Chapter 11 in relation to changing public awareness of mental health.

Diagnosis may alter expectations – both a person’s own expectations and those of others. Some people describe feeling understood and empowered; others feel that they come to be seen primarily through the lens of a diagnosis. Diagnostic labels may still carry stigma, and some studies have reported associations between self- labelling and lower perceived agency, more avoidant coping and reduced self- esteem, although the direction of these relationships remains uncertain and causal conclusions cannot be drawn[98-100]. The evidence therefore suggests not that diagnosis is inherently beneficial or harmful, but that its effects depend greatly on how it is understood and what follows from it.

One focus-group participant captured this balance well: “Diagnosis can help an individual to understand their symptoms, care options, and their experiences. However, diagnosis can still come with a huge amount of stigma, misunderstanding, and being funnelled into services that might not be applicable for you at that time. For example, the only mental health care I have ever gotten for bipolar through the NHS was medication – no talking therapies, no psychiatric nurse, no group therapies, no consideration of me being autistic – just medication the entire time I was under the care of a psychiatrist.”

Mind focus-group participant, 26–35 age group Reported harms are associated more often with misdiagnosis, stereotyping or absent follow-up. Diagnosis may have unintended psychological consequences. Explanations can themselves influence experience. Stigma remains. People who have benefited from taking prescribed medication for depression or ADHD often report stigmatising responses, such as ‘GPs hand out pills like sweeties’, ‘psychiatry is a pseudoscience’ and ‘pathologising unhappiness’.

History reminds us that misdiagnosis linked with professional confidence can lead to tragedy, as shown by the cautionary story of Sir William Arbuthnot Lane, surgeon at Guy’s Hospital between 1888 and 1920. Lane believed that he possessed ‘greater professional understanding’. He was convinced that young women presenting with what would now be recognised as common mental disorders such as depression and anxiety were experiencing ‘autointoxication’ – the products of digestion leaking through the walls of the colon. The solution was to remove the colon surgically, and his files recorded many grateful patients. But one in ten died of surgical complications. Lane was proud that his operative mortality was ‘only’ 10 per cent, against a standard mortality at the time of some 20 per cent; he was one of the best technical surgeons of his day. So, misdiagnosis combined with charismatic certainty can be lethal even in the ‘best’ hands[101].

Closer to our time, a major debate in developmental psychiatry at the beginning of the 21st century was the so-called ‘paediatric bipolar’ controversy[102], triggered by the view promoted by certain US physicians at Harvard that children as young as 5 years of age with irritability actually had a manic illness. This co-occurred with a dramatic increase in diagnosis and prescription rates of antipsychotics for bipolar disorder in prepubertal children. The debate reached a climax following the death of prepubertal children on antipsychotic medication, and a congressional investigation into undisclosed pharmaceutical industry payments to one of the main proponents of the ‘paediatric bipolar’ view.

Medical understanding changes, sometimes profoundly, and diagnoses that appear self-evident in one generation may later be reconsidered or abandoned. Likewise, as the neurologist Suzanne O’Sullivan has argued, not all diagnostic stories end well, albeit rarely as catastrophically as for those who encountered Lane[103].

5.5 Aligning diagnosis, need and support

Why has diagnosis become so important? Because it validates self-recognition and experiences and because it gives access to treatment, educational support, workplace adjustments, welfare benefits, social care and specialist services. In the latter, diagnosis has acquired functions it was never originally designed to perform. The more responsibilities diagnosis is asked to carry, the greater the pressure placed upon diagnostic systems. Rising demand for assessment may therefore reflect several processes operating together: widening diagnostic criteria, improved recognition, greater awareness, the search for explanation, the need to demonstrate legitimacy of distress, flawed systems, and the practical requirements of accessing support.

The important question for the individual is how diagnosis, need and support are aligned. Throughout this Review we have seen different forms of misalignment: some people remain unrecognised despite substantial need; others receive recognition but little effective support; some enter pathways that address only part of their difficulties; some obtain a diagnosis more quickly than others because of differences in awareness, advocacy, local provision or access to independent assessment.

A young person waiting for assessment described the practical consequences of waiting:

"I do struggle in school a lot so I stay in a place called The Bridge. It just helps me to be in a different environment and no one bothers me there. I was given a Bridge pass so I don’t have to go to class and I have a uniform pass which allows me to wear things I can cope with. I don’t get much help with school work because I don’t have the diagnosis yet."

Tellmi user, aged 14 This account illustrates that people are often seeking more than a diagnostic label. They are seeking understanding, practical support and environments in which they can participate successfully.

The significance of diagnosis therefore extends beyond identity alone. People are rarely seeking a diagnosis simply to acquire a label, though some do, out of curiosity. More often they are seeking understanding, recognition, validation and a coherent account of experiences that have previously been confusing or difficult to explain. Some require treatment or practical support; others are managing well and value

primarily the explanation and validation that diagnosis provides. General practitioners responding to the Review’s survey similarly reported that some adults asking whether they might have ADHD were seeking to understand longstanding patterns in their lives rather than requesting medication or specialist treatment. Qualitative studies across autism, ADHD and common mental health conditions consistently show that diagnosis may bring relief, reduce self-blame and help people make sense of longstanding experiences even where no further intervention follows[90, 96, 104].

Understanding why diagnosis matters requires attention not only to its clinical validity but also to the psychological and social work it now performs. At its best, diagnosis accurately identifies a clinically valid condition, helps people understand themselves, guides treatment and opens routes to support and participation. Yet even an accurate diagnosis may become associated with reduced expectations, stigma or assumptions about permanent limitations. The task for health, education and wider public systems is therefore not only to diagnose accurately, but to ensure that diagnosis is used in ways that expand understanding and opportunity rather than narrowing them.

Giving an individual a diagnosis is not a neutral act. As discussed throughout this Review, diagnosis has become the gateway to many forms of educational support, workplace adjustment and specialist provision. The tensions involved in deciding whether to make a diagnosis or not for an individual are often experienced directly in clinical practice. Clinicians recognise the importance that diagnosis may have for the person or family seeking help and may feel pressured but may still feel that a diagnosis is not clinically warranted or in the best interests of the patient. Patients and families, in turn, are usually aware of the practical consequences of receiving – or not receiving – a diagnosis. This can be a source of confrontation and/or complaints. Such situations are rarely the result of bad faith on either side; they arise because diagnosis is being asked to perform too many clinical and administrative functions simultaneously.

5.6 The risks of under-recognition and not diagnosing

A long-standing principle of medicine is to balance benefit against harm, although societal processes may shift that balance in either direction, often in unintended ways[70]. Every intervention has the potential to do both, and diagnosis is no exception. Throughout this chapter the Review has argued that diagnosis should neither be withheld because it is imperfect nor pursued without regard to its wider consequences. The challenge is to maximise its benefits while minimising its harms. That requires recognising both the risks of diagnosing unnecessarily or inaccurately,

and the equally important risks of failing to diagnose when diagnosis is clinically valid and could make a meaningful difference to a person’s life.

The preceding sections have examined the limits of diagnostic classification, the uncertainty inherent in diagnosis and the imperfect relationship between diagnosis and need. It is equally important to consider the opposite risk: failing to recognise clinically significant difficulties when they are present and the consequences. Under- recognition is the precursor to underdiagnosis. Recognition is about awareness. Diagnosis is about the official medical descriptor. You cannot diagnose a condition if the medical community or the patient does not first recognise that it is a possibility.

Untreated ADHD is associated with increased risks of accidental injury, substance misuse, depression, contact with the criminal justice system and premature mortality, with recent evidence suggesting reductions in life expectancy largely attributable to preventable causes[105]. These outcomes are not inevitable. A large population study from Sweden suggests that timely recognition and treatment are associated with reductions in suicidal behaviour, substance misuse, transport accidents and criminality[106]. Observational evidence must always be interpreted cautiously and the balance of risk to benefit continues to be debated, even in Sweden.

These risks are not unique to ADHD. Similar patterns are observed across many mental health conditions, often with equally strong or stronger associations with adverse educational, occupational and health outcomes.

Delay is particularly consequential where difficulties accumulate or where one condition obscures another. As discussed earlier, neurodevelopmental conditions may remain unrecognised because attention is focused on anxiety, depression, eating disorders or behavioural difficulties – a pattern of diagnostic overshadowing with significant consequences for treatment and support[107]. Some people never reach services because of language barriers, complex referral systems or pathways that are poorly suited to the way their difficulties present. People with multiple interacting conditions consistently experience poorer outcomes in education, employment, adaptive functioning, wellbeing and quality of life than those with a single condition[108, 109].

5.7 The risks of diagnosing too broadly: A Norwegian natural experiment

Direct evidence about the consequences of setting diagnostic thresholds more broadly or more narrowly is surprisingly limited. One of the few studies able to examine this question comes from Norway, where national registry data and marked

variation in diagnostic practice between child and adolescent mental health services created what researchers described as a natural experiment[110].

The study followed more than 26,000 children and young people seen by CYPMHS between 2009 and 2011 and linked their records to national administrative data over the following 11 years. Some clinics adopted relatively liberal approaches to ADHD diagnosis, while others applied more conservative thresholds[111]. The researchers found little evidence that these differences reflected the characteristics of the populations served[78, 112], but arose largely from differences in clinical practice between services. This created an opportunity to examine outcomes for young people whose likelihood of receiving an ADHD diagnosis depended primarily on which clinic they happened to attend.

Among young people whose medication depended on these differences in clinical practice, medication was associated with few measurable long-term benefits across most of the outcomes studied. The authors reported time-limited reductions in some outcomes, such as contact with child welfare services[113] and certain offending behaviours[114], but found little evidence of improvements in educational attainment[115], later psychiatric disorder[116], injuries[117] or several other long-term outcomes.

The study also examined the consequences of diagnosis itself in a manuscript currently under review[110, 118]. For those whose diagnostic status depended on which clinic they attended, receiving an ADHD diagnosis was associated with lower rates of completing upper secondary school or being in work (a combined measure) and higher receipt of disability benefits. These findings are consistent with earlier studies suggesting that, among people with relatively mild behavioural difficulties, diagnosis may sometimes be associated with less favourable educational or psychosocial outcomes than would otherwise have been expected[119, 120].

These findings require careful interpretation. The study was observational rather than a randomised trial, and although the investigators took considerable steps to minimise confounding, residual differences between services cannot be excluded entirely. The findings relate to Norway between 2009 and 2011 and may not generalise directly to contemporary England, with different diagnostic pathways, service pressures and patterns of referral. Most importantly, the results apply only to those whose diagnosis depended on differences in clinical practice. The authors do not claim this is evidence against diagnosing or treating ADHD more generally, and the associations with poorer outcomes should not be read as the direct effects of the label itself.

Nevertheless, the study provides an important insight. The balance between the benefits and potential disadvantages of diagnosis is unlikely to be constant across

the whole spectrum of severity. Among those ‘on the margins’ – those whose diagnosis depended on differences in clinical practice – medication showed little measurable benefit and the diagnosis itself appeared to carry drawbacks. The design cannot say whether the same holds for those with clearer and more severe presentations, who would have been diagnosed in any clinic. People with clear and substantial impairment are likely to benefit from timely recognition and treatment, although long-term outcome evidence across the whole diagnosed population remains limited[59, 121]. We have already pointed out that the balance of risk versus benefit can vary over time, but now we can also add the context in which it is made.

So the Norwegian findings strengthen the case for careful, high-quality assessment not just of symptoms but of functioning and co-occurring difficulties, for clear communication of uncertainty, and for meaningful support after diagnosis.

But debates about diagnosis, which are the theme of this chapter, must not obscure those whose needs remain unmet altogether. As William Davies[122] has observed, 314,000 young people in this country are neither in education, employment, training nor receiving benefits and are thus largely invisible to public systems.

5.8 The view from elsewhere

Very little discussed in this chapter is unique to mental health or neurodevelopment. Across medicine, clinicians debate where the boundaries between health and illness should be drawn, how thresholds should be set and how to balance the benefits of diagnosis against the risks of unnecessary investigation or treatment. The British Medical Journal (BMJ) ran an editorial, ‘New hypertension guideline risks overdiagnosis and over treatment’, saying that this would bring little or no benefit, turn large sections of the population into patients, create unnecessary medicalisation, an increase in adverse drug effects, a massive strain on resources and unnecessary anxiety[123]. In the event, little of this came to pass; the episode illustrates how strongly opinions are held on either side of such debates.

Patients with dementia frequently have mental health needs and mental health conditions increase dementia risk[124]. Dementia is also an example of how the two prevalences, population and administrative, can appear to be moving in opposite directions. The Cognitive Function and Ageing Studies assessed dementia in people aged 65 and over in the same three areas of England twice, two decades apart, using consistent methodology. Prevalence fell from 8.3% in 1991–93 to 6.5% in 2008–11, a relative reduction of around 22%, and incidence fell by approximately 20%[125]. In 2014 a national ambition was set to raise the proportion of people with dementia who had a formal diagnosis from around half to two-thirds. In October of that year NHS England introduced a Dementia Identification Scheme paying general

practices £55 for each new diagnosis recorded, described at the time as the first national initiative to reward doctors for diagnosing a condition. The professional response was immediate; the number of people recorded with dementia rose. However, the best available evidence was indicating that the number of people in the population with dementia was falling. The administrative and population prevalence measures were moving in opposite directions. In April 2015 it was announced that the Dementia Identification Scheme would not continue.

Similar discussions surround osteoporosis, prostate cancer and many other conditions. Campaigns such as Choosing Wisely and the BMJ’s Too Much Medicine initiative reflect a wider commitment within medicine to question whether diagnosis and treatment are always being used appropriately. These debates are signs of a self-critical profession rather than of failure. What is distinctive about mental health and neurodevelopment is not the uncertainty of its boundaries but the combination of that uncertainty with unusually wide consequences beyond healthcare, for education, employment, welfare and identity.

5.9 A short guide to the benefits and risks of diagnosis

‘Diagnosis can open doors, but also close them.’

Making a diagnosis is a skilled clinical task. It requires more than familiarity with diagnostic criteria. Good diagnosis depends on understanding the individual, their developmental history, current circumstances, strengths, difficulties, functioning and the contexts in which problems arise. It cannot be reduced to a checklist or a single assessment. Done well, diagnosis provides an important part of a broader clinical formulation; done poorly, it may misdirect support or fail to capture what matters most.

Throughout this Review we have argued that diagnosis should be understood alongside functioning, participation, context, strengths, risks and co-occurring conditions. This broader perspective moves the discussion beyond asking whether rising diagnosis is inherently good or bad towards more practical questions. Are people receiving the support they need? Are interventions improving functioning and participation? Are systems responding effectively to complexity and uncertainty?

To inform these questions, the Review commissioned a rapid evidence review from the Policy Research Unit at King’s College London and University College London examining the benefits and potential harms of diagnosis in common mental disorders, supplemented here by evidence relating to neurodevelopmental conditions and the wider literature. Overall, the evidence is reassuring, though it

does not support complacency. It suggests that diagnosis is associated with substantially more documented benefit than harm, while recognising that harms do occur and deserve careful attention.

5.9.1 Potential Benefits of Diagnosis

Understanding and validation

The most consistently reported benefit is understanding. Diagnosis can help people make sense of longstanding experiences, reduce self-blame and provide a coherent explanation for difficulties that previously seemed confusing or inexplicable. For many adults diagnosed later in life, particularly those with ADHD or autism, diagnosis reframes earlier experiences of failure or difference and provides a more compassionate understanding of themselves[95, 96]. Across autism, ADHD and common mental health conditions, qualitative studies consistently describe diagnosis as relieving, validating and explanatory[90, 96, 104]. Diagnosis therefore offers not only a description of difficulty but also a framework through which experience becomes meaningful.

Access to support

Diagnosis may also open access to support. It can facilitate treatment, educational adjustments, workplace accommodations, occupational therapy, peer support and communities organised around shared experience[95, 96, 126]. Delayed diagnosis may postpone these opportunities and prolong uncertainty, leaving people without the support that could have improved their participation and wellbeing[127].

Cost of not being diagnosed

The evidence reviewed throughout this chapter also demonstrates the costs of remaining undiagnosed where clinically significant conditions are present. Undiagnosed ADHD and autism are associated with higher rates of anxiety, depression, reduced quality of life and suicidal behaviour[107]. People with undiagnosed anxiety disorders similarly experience poorer quality of life, greater impairment and increased healthcare use than those whose difficulties have been recognised[128, 129]. These findings reinforce the importance of timely recognition where diagnosis can guide effective intervention.

5.9.2 Potential harms of a diagnosis

Diagnosis may also have unintended consequences. Stigma, both external and internalised, remains one of the most frequently described concerns. Some people report that diagnosis changes how others respond to them or alters how they think about themselves[95]. Others worry that diagnostic labels may become understood as

permanent and defining, particularly when insufficient attention is given to individual strengths, recovery or changing circumstances[127].

Diagnosis does not invariably bring reassurance. Where information about prognosis, treatment or likely outcomes is limited, uncertainty itself may become distressing. If a diagnosis is communicated in ways that imply permanence or inevitability, it may unintentionally reduce hope or confidence in the future[127]. This reinforces one of the Review’s central themes: diagnosis should be used to inform understanding and support rather than to define the limits of what someone may achieve.

Related to this is the possibility that diagnosis may sometimes contribute to fatalism or reduced agency. The available evidence is mixed. Some studies suggest that biomedical explanations reduce self-blame, whereas others report greater self- stigma or lower expectations[130, 131]. What emerges consistently is that the effects depend less on diagnosis itself than on how it is explained, understood and responded to. The Review therefore recommends that professional practice, public services, the voluntary sector and wider public narratives avoid presenting diagnosis as implying permanent incapacity or limited potential. Instead, diagnosis should support realistic hope, effective intervention and opportunities for participation wherever these are possible.

5.10 Towards a more useful framing

Throughout this chapter, the Review has argued that many of the fiercest debates about diagnosis arise from expecting diagnostic systems to answer questions they were never designed to answer. What is important is whether diagnosis is being used in ways that are proportionate to people’s needs and likely to improve their lives. That requires attention both to systems, which shape access to diagnosis and support, and to the experiences of individuals, whose lives are ultimately affected by those decisions.

This is not a new way of thinking. Every clinical decision involves weighing potential benefits against potential harms. Doctors do this routinely when prescribing medication, balancing expected improvements against possible adverse effects. Diagnosis deserves similar consideration. Like treatment, it can have important benefits but also unintended harms. Used well, it offers an explanation for difficulties, guides treatment and opens routes to support. Used poorly, or without appropriate follow-up, it may create misunderstanding, unnecessary pessimism or interventions that do not address the person’s needs. Diagnosis can open doors, but it may also, unintentionally, close some.

Population-based studies remain essential for picking up both ends of the diagnostic spectrum. They can identify groups who remain underserved; they can also, aided by randomised trials, identify medical screening programmes that have been found to do more harm than good – the classic cases of mis- or overdiagnosis arising from poorly performing screening tests. The same applies to mental health. These are legitimate and important tasks for public health and policy, even when the findings challenge established assumptions.

Diagnosis remains one of the most valuable tools available to clinicians, researchers and services. A world without diagnosis would create suffering, not reduce it. The Review proposes a system in which diagnosis is no longer the sole gateway to help, and is set within a broader assessment of functioning, context, risk, strengths and individual goals. The challenge is not whether diagnosis should be used, but how it can be used more carefully and more effectively.

Section Two: What is changing in population prevalence, need and service demand?

Chapter 6. Population mental health: what is changing, for whom, and with what impact?

Distress has increased, especially among young people. Common mental disorders rose in prevalence from one in six 16–24-year-olds in 1993 to one in four in 2024. Significant emotional symptoms climbed from one in five 10–15- year-olds in 2009 to one in three in 2023. Self-harm has risen sharply.

Different indicators point the same way. Prevalence, diagnosis and demand are not the same measure, but all three show common mental disorders rising steadily over the past three decades.

Some conditions are stable or have fallen. Childhood conduct problems and psychotic, bipolar and personality disorders in adults remain steady; irritability has fallen and hazardous drinking fell from 26% in 2000 to 19% in 2024.

The symptoms people report matter: they remain linked as strongly as ever to disrupted education, reduced participation and poorer outcomes into adulthood. Among children with a disorder, the proportion of parents reporting that their child experienced significant emotional impact doubled between 1999 and 2017.

Development is central. Most mental health conditions start early in life, with median age of onset about 14.5 years. Difficulties often persist; about half of children meeting criteria still do so three years later.

Distress has social patterns. The prevalence of common mental disorders is higher in young women (36.1%) than young men (16.3%); and among people in the most deprived areas (26.5%) compared with the least (15.3%). Sexual- minority adolescents and adults report roughly double the rate of depression and anxiety.

Some support need remains hidden: a minority of highly distressed adolescents reach specialist mental health services. Among adults with a common mental disorder, treatment receipt has risen from one in four (2000) to one in two (2024). However, among people with a similar level of symptoms, those least likely to get support are younger (16-24), older (65+), men, or of Black or Asian ethnicity.

Overall, the picture is one of cautious but clear concern: genuine increases in distress and impairment coexist with changes in recognition and diagnosis. The task for policy is to understand the changing relationship between distress, diagnosis, need and support – not simply to respond to rising demand.

This chapter examines how symptoms and distress have changed in the population, and how these trends relate to diagnosis, functioning, participation and service demand. The evidence shows both that awareness and recognition have increased, and that emotional symptoms and associated functional impairment have increased, particularly among young people.

6.1 Symptoms and distress

Increases are evident across symptoms: anxiety, worry, panic, concentration difficulties and fatigue have all increased, while irritability has fallen.

Self-harm has risen substantially, particularly among young women. Although self-harm has many causes, it provides an important behavioural indicator that psychological distress has increased.

Distress is not equally distributed. The greatest increases are seen among young people, women, sexual-minority groups and those experiencing socioeconomic disadvantage.

6.1.1 Understanding what is changing

Symptoms, diagnosis, service demand and need are not interchangeable. Symptoms describe people’s experiences and behaviours. Diagnosis determines whether those experiences meet recognised criteria for a particular condition and can shape access to treatment and support. Service demand reflects whether people seek or are referred for help. Need concerns the support required to improve functioning, wellbeing and participation (see Chapter 4).

These measures can follow different trends. Symptoms may increase without a corresponding rise in diagnosis. Diagnoses may increase as recognition improves. Demand may grow because awareness rises, barriers to help-seeking fall, services become more accessible, or diagnosis increasingly becomes a route to support.

This chapter draws together evidence from population surveys, cohort studies and health service data to examine how symptoms, diagnosis, functioning, participation and service demand have changed over time. Surveys use both brief screening questionnaires (such as the SDQ and GHQ-12) and structured diagnostic interviews (such as the DAWBA and CIS-R). Diagnostic interviews provide more robust estimates of disorder but are used less frequently because they are considerably more resource-intensive[132, 133].

Among children aged 5–15, England’s Mental Health of Children and Young People (MHCYP) survey found that the prevalence of emotional disorders increased from 4.3% in 1999 to 5.8% in 2017 using structured diagnostic assessments[134]. This remains the most recent national survey using diagnostic measures. More recent surveys have relied mainly on brief screening questionnaires. These also suggest that emotional symptoms have continued to rise. In UKHLS, the proportion of 10–15- year-olds scoring above threshold on the emotional symptoms scale increased from 21% in 2009 to 36% in 2023[32, 135]. Other cohort studies report similar increases[136,

137], while comparisons across successive British birth cohorts suggest that adolescent emotional distress has risen steadily since the 1980s[138].

Figure 6.1 Emotional disorders and emotional symptoms among children Source: MHCYP, UKHLS and related studies[134, 135]. (Table 2.1, Annex) The increase is even more pronounced among adolescents and young adults. England’s Adult Psychiatric Morbidity Survey (APMS), using a structured diagnostic interview[133], found that the prevalence of common mental disorders among 16–24- year-olds increased from 15.4% in 1993 to 25.8% in 2024[2] (Table 1.A.5) – a rise from around one in six to one in four. UKHLS similarly shows psychological distress increasing sharply after the mid-2010s, reaching 31% in 2024[32, 139]. In the Next Steps cohort, 23% of 16–17-year-olds exceeded the distress threshold in 2007, compared with 44% of similarly aged young people in the COSMO cohort in 2021/22[140]. Because these are different cohorts rather than the same individuals followed over time, they indicate generational differences rather than change within individuals.

Figure 6.2 Common mental disorders and distress among young people Source: APMS, UKHLS, COSMO and related studies[2, 140] (Table 1.A.5) Among adults aged 25 and over, increases are smaller but still evident. APMS shows that common mental disorders rose across ages 25–64, with much of the increase occurring after 2014[2] (Table 1.B.1). British birth-cohort analyses also find higher levels of distress at comparable ages in later-born generations, including increases in the proportions experiencing moderate and high levels of distress[141].

Figure 6.3 Common mental disorders and distress among adults (25+) Source: APMS, UKHLS and cohort analyses. (Table 1.B.1) Although awareness, recognition, and willingness to disclose difficulties have changed over time, reporting effects alone are unlikely to account for the consistency of the trend.

Taken together, diagnostic assessments, screening questionnaires, and – as Section 6.6 shows – self-identification and reported professional diagnosis all indicate that emotional difficulties and psychological distress have increased. The increase is clearest among adolescents and young adults, but also evident among older adults.

6.1.3 Which symptoms are increasing?

Changes are not uniform. APMS shows increases between 1993 and 2024 in worry, anxiety, depressive ideas, panic symptoms, obsessions and compulsions. Problems with concentration and forgetfulness approximately doubled among working-age adults, from around 8% to 16%, while fatigue is now reported by around one-third of adults (Table 1.A.6).

These changes extend beyond anxiety and depression to cognitive difficulties and fatigue. Possible explanations – including work, education, digital environments and wider social pressures – are considered in later chapters. Similar patterns are seen among younger people: emotional symptoms have increased, whereas hyperactivity and conduct-related difficulties have remained broadly stable[142-144].

Figure 6.4 Individual CMD symptoms among adults, by age Source: APMS 1993–2024 (Table 1.A.6) Irritability provides a useful counterexample: it declined, particularly among younger and working-age adults, even as several other symptoms increased (Table 1.A.6).

This selective pattern strengthens the conclusion that greater willingness to disclose cannot by itself explain the trends. If reporting alone were responsible, one would expect all symptoms to increase rather than some increasing, others remaining stable, and irritability declining.

6.1.4 Self-harm as an indicator of distress

Self-harm provides a behavioural indicator of distress that complements symptom questionnaires and diagnostic assessments. Population surveys and healthcare records show that self-harm has increased substantially over recent decades,

particularly among adolescents and young adults, especially young women[3] (Table 1.A.7).

Figure 6.5 Non-suicidal self-harm and suicidal thoughts among adults, by age Source: APMS 2000–2024 (Table 1.A.7) Self-harm is associated with emotional distress, difficulties in emotion regulation, psychiatric disorder, social adversity, and increased suicide risk. However, non- suicidal self-harm and suicidal thoughts have not followed the same trajectory: self- harm has risen, while suicidal thoughts have remained more stable. APMS data suggest that much of the increase involves self-harm used to cope with overwhelming emotional states, consistent with evidence that emotional distress and difficulties in coping have become more common[145].

6.1.5 Who is most affected?

Increases in distress have not been evenly distributed. In terms of age, adolescents and young adults show the greatest deterioration across psychological distress, common mental disorder and self-harm. In 2024, the prevalence of common mental disorders among 16–24-year-olds (25.8%) was more than twice that among adults aged 75 and over (10.2%)[2] (Table 1.B.1). Cohort and repeated cross-sectional studies show that more recent generations enter adolescence and early adulthood with higher levels of distress. This concentration matters because these are formative periods for education, employment, relationships, and wider participation.

The gender gap has also widened, particularly during adolescence. Girls report higher levels of emotional symptoms than boys[143], and longitudinal studies show steeper increases among adolescent girls[146]. Among young women, common

mental disorders increased from 20.5% in 1993 to 36.1% in 2024; among young men they increased from 10.3% to 16.3%[2] (Table 1.A.5).

Marked inequalities are seen by sexual orientation and gender identity and may be widening. Sexual-minority young people consistently report higher levels of depression, anxiety, self-harm, and psychological distress than their heterosexual peers[147-149]. The evidence base for transgender and gender-diverse young people is smaller but indicates similar inequalities. These differences emerge early, may widen during adolescence, and remain evident into adulthood.

Patterns by ethnicity are complex. Some studies find lower levels of emotional symptoms among young people in certain minoritised ethnic groups than among White young people[144, 149]. Recent increases have also often been greater among White young people than among Black and South Asian groups. These comparisons require care to interpret because cultural stigma may affect disclosure and help- seeking, as Section 6.5 discusses.

Socioeconomic inequalities remain deeply entrenched, and associations with social adversity are pronounced. Overall, recent increases in distress are concentrated among younger people, particularly young women, while longstanding inequalities associated with deprivation and minoritised sexual orientations and gender identities remain pronounced. Patterns by ethnicity are more complex. Mental health inequalities in both prevalence and treatment are considered further in Section 6.5.

6.1.6 What can we conclude?

Distress has increased, and more in some groups than others. The next section examines whether functioning and participation have changed alongside symptoms.

6.2 Functioning, participation, and consequences

Rising distress cannot be explained simply by a lower reporting threshold. Emotional symptoms remain strongly associated with impairment, and this relationship has not weakened over time.

Parents of children with a disorder are increasingly likely to report that their child experiences significant emotional impact, rising from one in four in 1999 to one in two in 2017. Around 10% of children with a disorder have significant attendance problems, compared with under 3% of those without a disorder.

Among adults, nearly 60% of those with a common mental disorder reported loneliness, compared with 16% of those without. Need is also becoming more complex: by 2024, around four in ten working-age adults with a common mental disorder also screened positive for ADHD symptoms, and four in ten also screened positive for autistic traits.

6.2.1 Why functioning matters

Prevalence alone cannot show how mental health difficulties matter in people’s lives. Similar levels of distress may have different consequences depending on developmental stage, personal strengths, social circumstances, and available support. Need emerges from the interaction between a person’s difficulties, strengths, and context. A needs-led approach requires personalised assessment of the whole person. Standardised measures should inform professional judgement, rather than replace personalised assessment or become gatekeepers to support.

The Review places particular emphasis on functioning and participation. Outcomes that matter include the ability to learn, work, form relationships, maintain independence, and participate in everyday life. A mental health system should be judged by whether it helps people function, participate, and flourish.

6.2.2 The consequences of symptoms in childhood

Emotional symptoms continue to be strongly associated with impairment and poorer outcomes. Research has found poorer social functioning and educational attainment among children with mental health difficulties in more recent British cohorts than among comparable children in earlier cohorts[150] and adolescent mental health

difficulties continued to predict poorer adult health, employment, and relationship outcomes[151].

Parent reports also indicate increasing impact. Among children with a disorder, the proportion whose parents reported that their child’s difficulties had a significant impact on their life rose from around a quarter in 1999 to a half in 2017. Among children who did not meet diagnostic criteria, parents also became slightly more likely to report emotional difficulties (Tables 2.4–2.7).

Together, the evidence does not support the view that rising emotional symptoms simply reflect greater reporting of milder difficulties.

Figure 6.6 Parent-reported emotional difficulties (SDQ) among children with and without mental disorder (DAWBA)

Source: MHCYP 1999–2017 (Tables 2.3–2.5)

6.2.3 Education and participation in childhood

Mental health difficulties in childhood are associated with reduced participation in education and school life, with consequences that can extend across the life course.

MHCYP shows that around 10% of children meeting criteria for a mental disorder had significant attendance problems, compared with fewer than 3% of children without a disorder. Depressive disorders were particularly strongly associated with

poor attendance at school, while conduct disorders were associated with the highest rates of exclusion (Tables 2.8, 2.9). MHCYP also found young people experiencing multiple difficulties were substantially less likely to take part in clubs and organised activities than their peers (Table 2.15).

Longitudinal studies associate childhood and adolescent emotional difficulties with subsequent lower educational attainment, reduced employment participation, lower income, relationship difficulties, and poorer physical health in adulthood[151-153]. Mental health difficulties are not the only explanation for disengagement or economic inactivity, but they make mental health an important consideration when examining transitions into adulthood, as discussed further in Chapters 10 and 16.

Figure 6.7 Educational disruption among children and young people with and without mental disorder.

Source: MHCYP 2017 (Table 2.8)

6.2.4 Functional impairment and complexity in adulthood

The consequences of mental health difficulties extend well beyond symptoms into everyday functioning, relationships, community participation and physical health. Adults with common mental disorders are substantially more likely than those without to report problems with activities of daily living. In 2024, more than a quarter of adults with a common mental disorder reported difficulties with personal care and a third reported difficulties with paperwork (Table 1.E.5). This greater level of impairment has shown no sign of diminishing over time.

Common mental disorders are also associated with poorer social functioning. Almost 60% of adults with a common mental disorder reported loneliness or isolation in 2024, compared with around 16% of those without a common mental disorder. A third reported problems with close relationships (Table 1.E.5). These differences have remained remarkably stable over time, suggesting enduring consequences for relationships and social participation as well as individual wellbeing. Other forms of vulnerability and impairment may have become more common over time among adults with a common mental disorder, including trauma-related difficulties and drug use problems (Table 1.D.4).

“From a young age, my trust in others around mental illness was damaged by stigma and a lack of understanding. At that age, I could not hide my struggles so easily. I learned that concealing them kept me safe and protected my relationships.

When I became unwell later in life, people often distanced themselves. I was told I was a ‘burden’ and to ‘leave my baggage at the door’ – so I stopped entering the door. When I trusted people enough to disclose what was really happening, they often stepped back or did not know how to respond. My ability to trust was eroded further.

The times I have reached out for help – from both my community and professionals – I have often been harmed rather than helped. Each experience has made me more reluctant to engage with services, more wary of being honest, and less able to establish trusting relationships.

It is often said that people retreat or withdraw from participation, as though this response is irrational or unfounded. But withdrawal can be learned self-protection. When someone’s history is marked by rejection, alienation and harm, isolation may feel like the only safe avenue left.”

Lived Experience Partner Figure 6.8 Difficulties with activities of daily living among adults (16+) with and without common mental disorders Source: APMS 2007–2024 (Table 1.E.5) The overlap between mental health and screening positive for neurodevelopmental traits has strengthened over time. By 2024, around four in ten working-age adults with a common mental disorder screened positive for ADHD symptoms, and a similar proportion screened positive for autistic traits, compared with around a quarter in 2007. The rise has been even steeper for young adults with a common mental disorder (Table 1.D.20). Chapter 9 considers this overlap in detail.

The growing overlap between common mental disorders, neurodevelopmental traits, trauma-related symptoms and social adversity suggests that contemporary mental health needs may be becoming more complex as well as more prevalent.

Figure 6.9 Co-occurring difficulties among adults (25–64) with common mental disorders

Source: APMS 2007–2024 (Table 1.D.20)

6.2.5 Symptoms, service use and unmet need

The difference between self-reported distress and primary-care-recorded common mental disorder is not necessarily unmet clinical need. Some people recover with support from family, friends, schools, workplaces, or communities; some choose not to seek formal treatment; and others face barriers, waits, or unsuitable provision. Among adults in England with a common mental disorder, the proportion receiving mental health treatment rose from 23.1% in 2000 and 24.4% in 2007 to 39.4% in 2014 and 47.7% in 2024 (Table 1.C.1).

Primary-care records alone may understate changes in population distress because they depend on help-seeking, clinical contact, recognition, and recording. At the same time, symptom prevalence cannot be translated directly into treatment rates. Differences between symptom prevalence and treatment receipt reflect varying levels of need, personal choice, and unmet need.

Self-reported distress and primary-care-recorded common mental disorders both rose among young adults between 2009 and 2019, but distress increased almost twice as much as recorded disorder[154]. Primary-care recording did not increase at the same rate as self-reported distress.

Figure 6.10 Current receipt of mental health treatment among adults (16–74) with Cmd (CIS-R)

Source: APMS 2000–2024 (Table 1.C.1) Rising demand reflects the combined effects of increasing distress, improved recognition, greater help-seeking and changes in service provision. As distress becomes more common, it becomes more visible and more discussed, which in turn may encourage recognition and help-seeking among others (Chapter 11). The evidence also highlights persistent inequalities in access to professional support and a continuing gap between need and service use[155, 156]. As detailed in Section 6.5, among people with a similar level of symptoms some groups are less likely than others to receive support, in particular men, people who are Black or Asian, and those who are younger (16–24) or older (65+) (Table 1.C.13).

6.2.6 What can we conclude?

Rising symptoms and distress are associated with meaningful consequences for education, employment, relationships, daily living, and community participation. Difficulties increasingly co-occur, suggesting that current need may also be more complex, requiring responses that recognise co-occurrence rather than treating conditions in isolation. The next section examines whether recent changes are found in other mental health conditions or concentrated within particular diagnostic groups.

6.3 Specific diagnostic conditions

Patterns vary: some conditions have risen, some have fallen, and others have remained stable.

Depression and anxiety disorders are rising in prevalence and healthcare recognition: increases are also clear for OCD, PTSD and drug dependence. Increased recorded diagnoses suggest rises for eating disorders.

Hazardous and dependent alcohol use has declined in prevalence, from 26% in 2000 to 19% in 2024. The fall is steepest among young people.

Psychotic disorder, bipolar disorder, and personality disorder are stable and remain associated with substantial and enduring disability and support needs.

Diagnostic prevalence estimates for ADHD and autism suggested broad stability up to the most recent national surveys, although recent evidence is lacking. Self-identification, recorded diagnoses, referrals, and demand have risen sharply. Chapters 7 and 8 examine this evidence.

6.3.1 Why diagnostic conditions matter

The preceding sections focused on emotional symptoms and their consequences. This is important because change in symptoms can provide an early indication of population changes and because many symptoms cut across diagnostic categories. This section considers trends across different diagnostic conditions. Diagnostic categories shape clinical pathways, access to support, treatment decisions, educational provision, and public understanding of mental health and neurodevelopmental difficulties.

Patterns differ by condition. The strongest evidence for increasing population prevalence relates to common mental disorders, particularly anxiety and depression. Eating disorders may also be rising. Some severe mental illnesses appear relatively stable in prevalence, despite changes in healthcare recording, recognition, diagnostic concepts, demand, and the organisation of support.

Chapters 7 and 8 consider ADHD and autism separately.

6.3.2 Common mental disorders: anxiety and depression

As set out in Section 6.1, across structured diagnostic interviews and self-reported diagnoses, common mental disorders show a sustained increase over time. On the MHCYP, the diagnostic DAWBA interview showed increased emotional disorders among children (5–15), from 4.3% in 1999 to 5.8% in 2017[134] (Tables 2.1, 2.2).

The APMS uses another structured diagnostic interview (CIS-R) to assess anxiety and depressive disorders[133]: prevalence rose from 15.5% in 1993 to 22.6% in 2024 among adults (16–64)[2] (Tables 1.B.1, 1.A.5). Self-recognition and reported professional diagnosis also increased. By 2024, almost half of adults reported that they had experienced a common mental disorder at some point in their lives, and around one-third reported having received a professional diagnosis. The proportion reporting the presence of a diagnosed CMD in the past year rose from 13.7% in 2014 to 18.9% in 2024 (Table 1.A.4).

Administrative datasets and clinical records corroborate these findings. Recorded prevalence of anxiety and depression has increased steadily, particularly among younger adults and women. Although administrative data are influenced by help- seeking, recording practices and access to care, they show the same broad pattern as the population surveys.

Figure 6.11a Recorded prevalence of depression by age band Recorded prevalence (%) Age group (years) 10-17 18-24 25-34 35-44 45-54 55-64 65+ 2005 2010 2015 2020 Calendar year Source: CPRD Figure 6.11b Recorded prevalence of anxiety by age band Recorded prevalence (%) Age group (years) 10-17 18-24 25-34 35-44 45-54 55-64 65+ 2005 2010 2015 2020 Calendar year Source: CPRD Administrative health records support the conclusion that common mental disorders have become more prevalent, particularly among younger people and young women.

6.3.3 Eating disorders

Eating disorders show evidence of increasing symptoms, recognition, recorded diagnoses, and service demand, particularly among adolescents and young adults.

Estimating population prevalence of eating disorders is challenging because many surveys rely on screening measures rather than detailed diagnostic assessment. Multiple sources do suggest that eating-related difficulties have become more common. Symptom measures show rises in concerns about weight, eating, body image, and related behaviours, particularly among adolescent girls and young women.

Primary care and specialist datasets show substantial increases in recorded prevalence and incidence over the past two decades, especially among adolescents and young adults, with an acceleration during and immediately after the COVID-19 pandemic.

Figure 6.12a Recorded diagnostic prevalence of eating disorders

Source: CPRD

Figure 6.12b Recorded diagnostic incidence of eating disorders

Source: CPRD Subthreshold eating-related difficulties may also involve substantial distress and impairment and increase future risk. More than three-quarters of young people screening positive for eating difficulties screened positive again one year later, although only a minority met full diagnostic criteria at follow-up[4].

While evidence indicates an increase in eating-related difficulties and associated need, the magnitude of change in underlying population prevalence remains uncertain. Much of the trend evidence comes from symptom measures, administrative records, and service demand. Robust diagnostic estimates remain

limited, particularly for adults, for whom the first national prevalence estimate using a diagnostic interview became available only in 2024[157].

There is considerable heterogeneity across eating disorders. Conditions such as anorexia nervosa, bulimia nervosa, binge eating disorder and ARFID differ in their age of onset, co-occurring conditions, clinical course and outcomes, while disordered eating behaviours also occur in many young people who do not meet diagnostic criteria for an eating disorder. These behaviours may nevertheless cause substantial distress and impairment and are associated with increased future risk of eating disorder and other mental health conditions[158].

Eating disorders have become more visible and more prominent within mental health services and are likely to have become more common. However, the scale of the underlying population change, and the extent to which trends differ between individual eating disorders, remain uncertain.

6.3.4 PTSD, OCD and substance dependence

Some conditions have increased, others have remained stable, and some have declined. The accompanying figure summarises some of these differing trends for adults in England between 2014 and 2024.

Some of the biggest shifts were for the proportion of adults screening positive for ADHD and autistic traits this decade, which are considered in Chapters 7 and 8.

Figure 6.13 Changes in prevalence of selected diagnostic conditions and symptom screens

Source: APMS 2014–2024 (Table 1.A.7)

Post-traumatic stress disorder (PTSD)

Probable PTSD, measured in APMS using the PCL-C screening instrument rather than a diagnostic assessment, rose from 4.4% in 2014 to 5.7% in 2024[159, 160] (Table 1.A.7).

Over the same period, the proportion of adults reporting that they believed they had experienced PTSD increased from 3.3% to 4.9%, while reported professional diagnosis increased from 1.9% to 3.1% (Tables 1.A.1 and 1.A.2)[2]. These measures suggest increases in both symptom prevalence and recognition.

Obsessive-compulsive disorder (OCD)

The proportion of adults identified with OCD using a structured diagnostic interview increased from 1.6% in 2014 to 2.6% in 2024, with a notable rise among 16–24-year- olds.

A third of adults identified with OCD in 2024 believed that they had experienced OCD, and about one in ten reported a professional diagnosis[2]. These differences highlight that prevalence, self-recognition, and reported diagnosis are related but distinct.

Figure 6.14 Prevalence of Ocd (CIS-R) by age group Source: APMS 1993–2024 (Table 1.A.8) Several factors may contribute to this gap. Recognition of OCD has broadened beyond contamination and checking presentations, with increased awareness of the ways in which the condition can present. At the same time, OCD frequently co- occurs with other mental health and neurodevelopmental conditions, meaning that obsessive-compulsive symptoms may be overlooked or attributed to another primary diagnosis. The findings suggest that increases in symptom prevalence, recognition and diagnosis are all occurring, but not at the same pace[2].

Substance use and dependence

Trends in hazardous, harmful and dependent alcohol use, measured using the brief Audit screening tool, show that prevalence is falling, especially among young people (Table 1.A.7). This likely reflects a combination of changing social and cultural norms, together with wider changes in patterns of alcohol consumption. Alcohol use provides an important reminder that different conditions and behaviours respond differently to broader social and policy changes.

Figure 6.15 Hazardous, harmful or dependent alcohol use among adults (16+) by age

Source: APMS 2000–2024 The trend for drug dependence is different. Using a brief screening tool, APMS shows increases in drug dependence, especially among people also experiencing common mental disorders[161] (Table 1.D.4). These findings point towards growing complexity among those experiencing distress. For some, substance use may represent an attempt to alleviate or cope with distress. The close relationship between common mental disorders and drug dependence reinforces the need to recognise and address co-occurring conditions together, not in isolation.

6.3.5 Psychotic, bipolar and personality disorders

Psychotic disorders, bipolar disorder and personality disorders are conditions where population prevalence appears comparatively stable. They remain associated with substantial disability, premature mortality, and social exclusion, as well as sustained support needs across health, social care, housing, and employment services.

Psychotic disorders

Psychotic disorders include psychosis and schizophrenia. APMS uses structured screening followed by detailed clinical assessment using the Schedules for Clinical Assessment in Neuropsychiatry (SCAN). Prevalence was 0.4% in 2007, 0.7% in

2014, and 0.4% in 2024, indicating stability, with a pooled estimate of around 0.5%[5, 162, 163] (Table 1.A.7). Recorded diagnostic prevalence also indicates broad stability.

Bipolar disorder

Using the 17-item Mood Disorder Questionnaire, APMS found little evidence of change in the proportion of adults screening positive for bipolar disorder between 2014 and 2024[6] (Table 1.A.7). Administrative datasets, however, show increasing recorded diagnoses and numbers receiving care. This may reflect improved recognition, changing service use, and the duration of diagnoses rather than an increase in underlying population prevalence. Diagnosis is also frequently delayed.

Figure 6.16 Recorded diagnostic prevalence of bipolar disorder

Source: CPRD

Personality disorders

Survey data indicate that the proportion of adults meeting criteria for borderline or antisocial personality disorder or broader personality disorder traits has remained relatively stable between 2014 and 2024[164, 165] (Table 1.A.7). UKHLS also found no evidence of increases in antisocial behaviour among adolescents[135]. Despite stable prevalence, personality disorders remain associated with high levels of distress, self- harm, suicidal behaviour, adversity, and service use. Recognition remains limited (Tables 1.A.1, 1.A.2).

Severe mental illness, inequality and burden

People living with psychotic disorders, bipolar disorder and personality disorders experience some of the poorest health and social outcomes of any group considered in this Review[166-169]. Rates of unemployment, economic inactivity, housing instability and social exclusion remain exceptionally high. Schizophrenia shows one of the steepest socioeconomic gradients observed, with prevalence substantially higher in the most deprived communities than the least deprived.

Figure 6.17 Deprivation gradients in schizophrenia

Source: CPRD Physical health inequalities are profound, contributing to reduced life expectancy through elevated rates of cardiovascular disease, metabolic disorders and other physical illnesses. Although these conditions affect a relatively small proportion of the population, they account for a disproportionate share of disability, inpatient care and long-term support needs and a large proportion of specialist mental health expenditure and intensive community support. As explored in Chapter 9, these frequently occur alongside substance use disorders, trauma histories, personality difficulties and social adversity, contributing substantially to impairment and complicating treatment and recovery. Consequently, even small changes in prevalence or severity can have substantial implications for health service planning.

6.3.6 What can we conclude?

The epidemiological picture for psychotic disorders, bipolar disorder and personality disorder is one of relative stability combined with persistently high burden. Their importance lies not in rising prevalence but in the severity, inequality, and persistence of associated need. Their contribution to disability, mortality, social exclusion and healthcare expenditure remains profound. Understanding

contemporary mental health requires attention not only to where prevalence is increasing, but also to where burden remains persistently high despite stable prevalence.

6.4 Trajectories, persistence and age of onset

Many mental disorders first emerge early in life: the median age of onset is about 14.5 years, and nearly two-thirds begin before age 25[170]. Difficulties often persist, with around half of children meeting criteria for a disorder still doing so three years later.

Trajectories are shifting across generations. Successive birth cohorts (1946 → 1958 → 1970) report higher levels of distress at the same age, suggesting that more recent generations experience earlier onset and, for some conditions, greater persistence.

Early difficulties cast long shadows. Mental health problems beginning in childhood and adolescence are associated with poorer educational attainment, employment, relationships, physical health and economic participation across adult life.

6.4.1 Understanding age, development and cohort effects

This section considers evidence from longitudinal studies, which follow the same people over time and help distinguish developmental change from differences between generations. Together, these studies suggest that while mental health follows recognisable developmental pathways, later-born cohorts experience greater distress at comparable ages, earlier emergence of emotional difficulties, and, for some conditions, less favourable long-term outcomes. They also show that difficulties arising in childhood, adolescence and early adulthood often persist and influence later education, employment, relationships and participation.

6.4.2 Emotional difficulties emerge earlier

Comparisons between the Avon Longitudinal Study of Parents and Children (ALSPAC), whose participants were born in the early 1990s, and the Millennium Cohort Study (MCS), whose participants were born around a decade later, found that emotional difficulties increased more rapidly during adolescence and remained higher in the later cohort[146].

The differences were particularly marked among girls. Overall, the findings suggest that developmental trajectories are changing, with more recent generations experiencing emotional difficulties that emerge earlier, increase more rapidly during adolescence and, for some, persist for longer than in earlier cohorts.

6.4.3 Symptoms persist across childhood

Follow-up studies from the MHCYP found that around half of children meeting criteria for a psychiatric disorder continued to meet criteria within the same diagnostic grouping three years later[171]. Persistent anxiety was associated particularly with peer difficulties, while persistent conduct-related difficulties were associated with peer problems, poorer family functioning, housing difficulties and wider disadvantage[171].

Continuity most often occurs within the same broad diagnostic domain, but developmental pathways are not fixed. One difficulty may increase vulnerability to another. Hyperactivity at age 11, for example, has been associated with a greater risk of depression at ages 14–15[172]. Difficulties with attention or activity may affect schooling, relationships and self-esteem, potentially contributing to later depression. Timely and appropriate support may influence whether these difficulties resolve or become more entrenched.

6.4.4 Symptoms persist into and across adulthood

Psychological distress often rises from early adulthood into midlife before declining in later life, although patterns vary by sex and across major life transitions, including parenthood and menopause. These age-related patterns cannot, by themselves, distinguish the effects of ageing from generational change.

Evidence from three nationally representative British birth cohorts shows that, at comparable ages, later-born cohorts experienced consistently higher levels of distress than earlier cohorts, from early adulthood through to midlife[141]. Although distress changes with age within each cohort, later-born cohorts generally follow a consistently higher trajectory. The greater distress observed among younger people therefore cannot be explained simply by adolescence and early adulthood being more vulnerable periods.

6.4.5 Cohort effects and changing developmental contexts

While age effects reflect changes associated with development and ageing, cohort effects occur when people born in different periods experience different levels of distress at the same age. Period effects arise when historical events or conditions

affect everyone living through them, although their impact may differ according to age and circumstance.

The COVID-19 pandemic is an example of a period effect: it affected the whole population, but not equally across age groups. Period effects may become cohort effects when exposure occurs during a sensitive developmental period and has lasting consequences. Chapter 10 considers a range of possible contextual drivers, including the pandemic, alongside changes in education, employment, housing, family life, technology and wider social conditions. These influences may accumulate and interact over time, affecting different groups in different ways. The cohort evidence therefore suggests that contemporary trends must be understood not only as changes within individuals, but also in relation to the changing social, economic and cultural contexts in which development occurs.

6.4.6 Early years, from infancy, matter

Babies, children, young people and young adults are central to this Review because many mental health conditions begin early in life, developmental foundations are being established, and later-born cohorts report higher levels of distress.

Difficulties in emotional regulation, attachment, communication and social engagement may emerge during infancy, while neurodevelopmental differences and, more rarely, developmental regression may become apparent in the earliest years of life[173, 174]. Attention to mental health should therefore begin in infancy, rather than only when children enter school or become able to report their own distress. The Review’s emphasis on the period from infancy to early adulthood reflects the early onset of many conditions, their persistence, their cumulative consequences across the life course, and the importance of changing developmental environments. Understanding contemporary mental health therefore requires attention to babies as well as to children, adolescents and young adults.

6.4.7 Most conditions have onset in childhood

Understanding when difficulties begin helps distinguish conditions with different developmental profiles and identify when recognition and timely support may be most effective. Across mental disorders, the median age of onset is approximately 14.5 years, and almost two-thirds first emerge before the age of 25[170]. Different conditions, however, follow distinct developmental patterns.

Separation anxiety and specific phobias commonly begin before puberty, while social anxiety frequently emerges during later childhood or adolescence. Depression and many eating disorders become more common during adolescence, particularly among girls and young women. Psychotic disorders and bipolar disorder most often

first present in late adolescence or early adulthood, although diagnosis may follow a prolonged period of emerging or unrecognised difficulty. Difficulties associated with personality disorder also commonly become apparent during adolescence, even where formal diagnosis occurs late[170].

For ADHD and autism, early developmental origin must be distinguished from the age at which features become visible or are recognised. Across conditions, age of onset, age at which difficulties become apparent, age of recognition and age of diagnosis are related but not synonymous. Services must therefore be able to respond across the life course, and the age at which someone is first diagnosed should not be assumed to represent when their difficulties began.

6.4.8 Later life outcomes are wide ranging and interact

Longitudinal studies associate childhood and adolescent mental health difficulties with poorer educational attainment, reduced employment participation, lower income, relationship difficulties, poorer physical health and premature mortality[152, 153, 175]. These associations have been observed across countries and generations and remain evident after accounting for many aspects of family and social background. British cohort studies also suggest enduring associations with economic participation, social mobility and relationship stability[152].

These outcomes often interact, creating cascades of disadvantage across multiple areas of life. Educational disruption may reduce later employment opportunities; insecure employment may undermine housing and financial security; and both may restrict social participation and wellbeing. Such patterns are neither inevitable nor explained by the child alone. Rather, they emerge through the interaction of individual difficulties, family and social circumstances, and the responsiveness of the systems surrounding them.

Their effects may also extend across generations. Difficulties affecting mental health, relationships, participation and economic security in adulthood can influence the circumstances in which people become parents and the relational environments in which babies develop. This is not a deterministic transmission of disadvantage. Families possess important strengths, and timely support can interrupt these pathways. It does, however, reinforce the importance of attending to mental health from infancy onwards and of supporting parents and carers as well as children.

Family circumstances and the responsiveness of surrounding systems may either compound or mitigate these effects. Early experiences of support may also shape later trust in services and willingness to seek help. Chapter 8 considers barriers to healthcare, including those experienced by autistic people, while Chapter 9 examines co-occurring mental health and neurodevelopmental conditions.

6.4.9 What can we conclude?

Contemporary increases in distress should not be understood as short-lived fluctuations confined to young people. They reflect developmental trajectories whose consequences may extend across adulthood and, in some circumstances, into the next generation. Understanding mental health therefore requires attention not only to current symptoms but also to the timing of onset, the persistence of difficulties, and their consequences across the life course. Chapters 19 and 20 consider the implications for prevention, early support and service design.

6.5 Inequalities in mental health and treatment

Mental health inequalities change over time and differ across both the prevalence of conditions and access to treatment and support.

While women are more likely to experience anxiety and depression, men with comparable symptoms are less likely to receive support.

Sexual-minority and gender-diverse groups experience substantial inequalities: rates of depression and anxiety are around twice those of heterosexual and cisgender peers, and around two-thirds of non-binary young people report clinically significant distress.

People from minoritised ethnic groups are consistently less likely than white people to receive treatment when experiencing similar levels of symptoms.

Socioeconomic inequalities in mental health are substantial: in 2024, the prevalence of common mental disorder was 26.5% in the most deprived areas, compared with 15.3% in the least deprived. Socioeconomic inequalities in access to treatment, however, are less clear.

6.5.1 Why inequalities matter

Mental health difficulties are not distributed evenly across the population. Symptoms, disorders, service use and outcomes vary according to people’s characteristics and circumstances.

Population averages can conceal both concentrated need and unequal responses. Interpretation therefore requires care. Sex is a biological classification, whereas gender, ethnicity, sexual orientation and socioeconomic position are social and demographic characteristics whose meanings and consequences vary across

contexts. They may be associated with different exposures, opportunities, experiences of discrimination, sources of support and pathways into care.

Section 6.1.5 summarised inequalities in emotional distress; this section considers inequalities in common mental disorders and in access to treatment and services.

6.5.2 Access to support and treatment in mental health

Not everyone experiencing distress requires or seeks specialist treatment. Many people recover with support from family, communities or primary care, while others choose not to seek formal help. Gaps between distress and service use should not automatically be interpreted as unmet need. However, persistent differences in treatment receipt between groups with comparable levels of need indicate that equitable access has not yet been achieved. MHCYP analyses found that children from ethnic minority and socioeconomically disadvantaged households were less likely to receive professional support than White and more advantaged children with comparable levels of difficulty[156].

Figure 6.18 Likelihood of mental health treatment receipt among adults (16+) Source: APMS 2024 (Table 1.C.13) APMS data also reveal marked inequalities in treatment receipt among adults in England with similar levels of need[176]. Figure 6.18 shows particularly pronounced differences by sex, age and ethnicity, indicating that variations in treatment cannot be explained by differences in need alone. Possible explanations include service availability, previous experiences of care, stigma, discrimination, cultural understandings of mental health, health literacy, and the confidence or resources needed to navigate services. Families with greater knowledge, time or financial resources may be better able to advocate for assessment and treatment, while others encounter barriers before reaching specialist services[177].

Mental health inequalities are therefore inequalities of recognition, access and treatment, as well as prevalence. They reflect differences not only in the occurrence of mental health conditions but also in how people enter, experience and benefit from health systems. Addressing these inequalities is central to achieving parity of esteem and ensuring that support is matched to need rather than to a person’s ability to navigate the system.

This section concerns inequalities in mental healthcare. Chapters 7 and 8 consider inequalities in the assessment, diagnosis and support of neurodevelopmental conditions. Chapters 14 and 15 examine waiting times, referral pathways and models of care in greater detail.

6.5.3 Sex

Women have a higher prevalence of common mental disorders than men across a range of symptoms. This disparity has persisted, and widened over recent decades, particularly among young adults. Men, meanwhile, are more likely to experience substance-related and some other mental health conditions. Inequalities in the recognition and diagnosis of neurodevelopmental conditions are considered in Chapters 7 and 8.

Beyond differences in prevalence, it is also important to consider who does and does not receive treatment. Figure 6.18 shows that, among people with similar levels of symptoms, men were less likely than women to receive either mental health medication or psychological therapy. This pattern has remained consistent over time, being evident in data from 2007, 2014 and 2024 (Table 1.C.13).

Figure 6.19 CMD symptoms among young adults (16–24) by sex Source: APMS 1993–2024 (Table 1.A.6)

6.5.4 Sexual orientation and gender identity

Sexual orientation and gender identity are distinct, but both are associated with marked mental health inequalities. Across UK cohorts, sexual-minority young people report higher levels of depression, anxiety, self-harm and psychological distress than their heterosexual peers[147-149]. These inequalities emerge during adolescence, often widen during the transition to adulthood, and remain evident throughout adult life.

Evidence from APMS[178] and a meta-analysis of UK population studies[179] similarly shows that sexual-minority adults are around twice as likely to experience anxiety and depression, with comparable inequalities in self-harm, suicidal ideation and psychological distress. APMS found little evidence that these inequalities narrowed between 2007 and 2014[180], while UKHLS suggests that they may have widened between 2010 and 2021, particularly among bisexual women[181].

Figure 6.20 CMD prevalence by sexual minority status among adults (16+) Source: APMS 2007–2024 (Table 1.B.1) Evidence concerning gender-diverse populations is more limited but points in the same direction. In the COSMO study, around two-thirds of non-binary young people reported clinically significant psychological distress, while self-harm was substantially more common than among male and female participants[140]. GP Patient Survey analyses also found higher rates of self-reported mental health conditions, autism and learning disability among people reporting minority gender identities than among those identifying as male or female[182] (Table 3.2).

The overlap between gender diversity, gender-related distress and autism is increasingly recognised in both clinical and population studies[183]. This association requires careful interpretation. It should neither be ignored nor treated as invalidating gender-diverse identities. Rather, it raises important questions about recognition, communication, co-occurring needs and developmentally appropriate support. Chapters 8 and 9 consider autism and co-occurrence in greater detail.

Explaining the persistence of these inequalities is likely to require consideration of minority stress, discrimination, social relationships and access to support, as discussed in Chapter 10.

6.5.5 Ethnicity

The relationship between ethnicity and mental health is complex. Different studies, age groups, outcomes and service settings reveal different aspects of that relationship, and interpretation requires particular care because prevalence, help- seeking, diagnosis and service use may all be influenced by cultural as well as clinical factors. Several surveys report lower levels of self-reported emotional symptoms among some minoritised ethnic groups than among White and mixed- ethnicity young people[143, 149]. Recent analyses similarly suggest that increases have been greatest among White young people, while Black and South Asian groups have shown more stable trajectories[144, 184].

These findings require careful interpretation. Distress may be understood, expressed and disclosed differently across cultures, families and communities. Standard survey measures may underestimate need where distress is not expressed in expected ways, while service data may reflect differences in access and recognition as much as differences in underlying prevalence. Experts by Experience and Lived Experience Partners consistently emphasised the importance of cultural context when interpreting epidemiological findings. Family relationships, community networks, religious participation and cultural identity may be protective in some contexts. However, these factors should not be generalised to all members of an ethnic group, treated as fixed characteristics, or assumed to outweigh discrimination, deprivation and other adversities. Their contribution is considered further in Chapter 10.

Recent GPPS analyses found higher rates of self-reported autism, learning disability and mental health conditions among adults identifying with mixed or multiple ethnic backgrounds than among other ethnic groups[182] (Table 3.1). These findings reinforce the conclusion that mental health, neurodevelopmental conditions and social inequalities frequently intersect.

APMS found a higher prevalence of common mental disorders among some Black ethnic groups than among White British adults, although survey samples are often too small for robust subgroup analyses[176] (Tables 1.B.1, 1.C.13).

Figure 6.21 Common mental disorder (CIS-R) among adults (16+) by ethnicity Source: APMS 2007–2024 (Table 1.B.1) Across both children and adults, people from Black and some Asian ethnic groups are less likely than White peers to receive treatment after differences in need are taken into account[156, 176]. Structural barriers, racial stereotyping, discrimination and differences in pathways into care may all contribute. MBRRACE reports also continue to demonstrate marked ethnic inequalities in maternal deaths associated with mental illness[185].

Among adults aged 16 years and over in England, APMS 2024 found that, after adjustment for symptoms and other characteristics, Black/Black British adults had around one-third the odds of receiving treatment compared with White British adults (adjusted odds ratio 0.34, 95% CI 0.18–0.67). Asian/Asian British adults had around two-fifths the odds (0.40, 95% CI 0.20–0.80) (Table 1.C.13).

The broader lesson is that patterns differ according to what is measured. Symptoms, diagnosis, recognition, access and outcomes do not always coincide. Interpretation therefore requires attention both to cultural context and to unequal pathways into care.

6.5.6 Socioeconomic disadvantage

Socioeconomic disadvantage is one of the strongest and most consistent correlates of poor mental health. In APMS 2024, the prevalence of common mental disorder was 26.5% among adults living in the most deprived neighbourhoods, compared with 15.3% among those in the least deprived. Prevalence was also higher among adults living in rented accommodation, those who were unemployed or economically inactive, those experiencing serious debt, and those receiving out-of-work disability benefits. More than two-thirds of adults receiving these benefits met criteria for a common mental disorder, compared with around one-fifth of those not receiving them (Table 1.B.1).

Figure 6.22 Common mental disorder among adults by socioeconomic factors Source: APMS 2007–2024 (Table 1.B.1) The same pattern is evident in childhood. Across MHCYP surveys, children living in lower-income households or rented accommodation were more likely to meet criteria for a mental disorder[186]. Whether these inequalities are widening is less certain. Some cross-cohort studies suggest that disadvantage has become a stronger predictor of emotional difficulties in more recent generations[138, 187], while UKHLS analyses provide less consistent evidence that recent increases have been concentrated in more deprived areas[144, 184]. One interpretation is that disadvantage acts as a persistent amplifier of vulnerability rather than the principal driver of changing prevalence.

Violence and abuse, homelessness, insecure employment, financial insecurity and caring responsibilities are also strongly associated with common mental disorders (Tables 1.B.3, 1.F.1). People experiencing multiple adversities have particularly high levels of distress, suggesting that cumulative disadvantage may matter more than any single exposure.

Figure 6.23 CMD prevalence by social adversity indicators among adults (16+) Source: APMS 2007- 2024 (Tables 1.B.3 and 1.F.1) The relationship between mental health and socioeconomic disadvantage is unlikely to operate in only one direction. Economic hardship may increase distress through insecurity, reduced opportunities and chronic stress, while mental health difficulties may restrict education, employment and financial security. These processes may reinforce one another over time. Socioeconomic inequalities therefore concern exposure to adversity, functioning, access to support and life chances, as well as symptoms. Chapters 10 and 16 examine the underlying mechanisms and wider implications.

6.5.7 Students, education and participation pathways

Education is a key pathway through which mental health inequalities influence later life. As discussed in Section 6.4, mental health difficulties are associated with poorer attendance, reduced educational participation, lower attainment and more disrupted transitions. Emotional difficulties may affect attendance, concentration, learning and participation, while educational disruption may in turn affect later employment,

income, social mobility and community participation. Mental health and educational participation can therefore reinforce one another over time.

Population studies generally find that students in higher education report similar or slightly better mental health than young people of the same age outside higher education, although the differences are modest[188, 189]. The broader issue is participation across school, further and higher education, apprenticeships and other routes into adult life. Mental health difficulties are closely intertwined with educational disengagement and being not in education, employment or training (NEET), although they are not the sole cause. The implications for participation, employment and productivity, together with the evidence concerning different educational environments, are considered in Chapters 10, 13 and 16.

6.5.8 What can we conclude?

Mental health inequalities are inequalities in life chances as well as symptoms. They influence who experiences distress, how difficulties are recognised, whether support is received, and how mental health affects education, employment, relationships and participation across the life course. Understanding these patterns is essential for developing effective and equitable responses.

6.6 Groups at risk of being missed

No single data source fully captures those with the greatest need: the most excluded and disadvantaged are also the most likely to be missing from both survey data and health service records.

People who are care-experienced, homeless, living in insecure housing, in prison or other institutional settings experience high levels of need, have poorer access to support, and are often under-represented in available data.

People with learning disabilities are often under-represented in research and face barriers to recognition in both data and services.

Many people go without support. Three-quarters of children and young people with a mental disorder have no contact with specialist mental health services and many are therefore absent from administrative datasets.

Complexity can hide need. Co-occurring conditions are increasingly common, while services organised around single conditions may fail to recognise overlapping needs.

6.6.1 Why some groups are missed

Understanding population mental health requires attention not only to those who appear in services and datasets, but also to those who remain partially or wholly invisible. This is partly a problem of measurement. Administrative data depend on recognition, diagnosis and contact with services, while even population surveys under-represent people who are homeless, living in secure settings or otherwise socially marginalised. Recognition and reporting may also differ according to culture, stigma, language, age and previous experiences of services.

People may be missed because symptoms are overlooked, help is not sought, services are difficult to access, or care pathways respond poorly to complexity and co-occurrence. Estimates based solely on diagnosis or service use will therefore underestimate need in some groups. This section considers who is least likely to be recognised, counted or supported. Chapters 10 and 11 explore these issues further.

6.6.2 Children and young people with unmet need

Across several studies, many children and young people meeting criteria for a mental disorder report no contact with specialist mental health services. MHCYP analyses found that three-quarters of children meeting DSM-5 criteria for a mental disorder had not been in contact with a mental health specialist[155]. Although service use increased with severity and complexity, many children with clinically significant difficulties remained outside formal care and are therefore absent from administrative datasets.

Figure 6.24 Type of professional service contact among children (5–19) with a mental health condition (DAWBA)

Source: MHCYP 2017;[155, 156].

Recognition by adults is an important part of this picture. Parent- and teacher- reported difficulties are more strongly associated with specialist contact than adolescents’ self-reported difficulties, suggesting that young people often depend on adults to recognise concerns and facilitate access to support. Internalising presentations, including anxiety, depression and self-harm, are often less visible than disruptive behaviour. Their recognition may depend particularly on disclosure, trusting relationships and active enquiry[190, 191]. Chapters 7 to 9 consider the recognition of neurodevelopmental conditions.

Growing demand has widened this gap. Although mental health provision has expanded, demand has frequently increased faster than capacity, contributing to delays in assessment and treatment[192]. Chapter 14 considers waiting times and service capacity in greater detail. Prevalence estimates based only on diagnosis or specialist service use therefore underestimate need.

6.6.3 Socially excluded and highly vulnerable populations

Some groups with the greatest mental health needs are also among the least visible in conventional data. Household surveys and routine healthcare records depend on stable housing, contact with services and participation in research, and therefore

under-represent people who are homeless, living in secure settings, or otherwise socially excluded.

Young people in contact with the youth justice system and adults in prison[193] have high rates of mental disorder, trauma, substance misuse and learning difficulties, with many experiencing several simultaneously[194]. Care-experienced children also have substantially elevated rates of emotional difficulties, self-harm and psychiatric disorder[195]. Chapters 15, 17 and 19 consider these groups and the implications for services.

Adults experiencing homelessness have high rates of common mental disorders, trauma-related difficulties, substance misuse and severe mental illness (Table 1.B.3). Mental ill health may increase vulnerability to homelessness, while homelessness may, in turn, worsen mental health. Comparable patterns are seen among people experiencing domestic abuse, insecure housing and multiple forms of social exclusion[196-199].

People with learning disabilities also experience elevated rates of psychiatric disorder[200] and higher self-reported rates of mental health conditions and autism[201], while remaining at risk of under-recognition. Diagnostic overshadowing – in which symptoms of mental illness are attributed to learning disability rather than assessed in their own right – remains an important concern[202]. Chapters 5 and 9 consider these issues further.

The common feature linking these groups is not any particular diagnosis but their limited visibility within conventional data sources. As people move further from mainstream educational, healthcare and social systems, they become progressively harder to capture through standard epidemiological methods. Population surveys remain indispensable, but they cannot fully describe the mental health needs of those living at the margins of society.

6.6.4 People with multiple and overlapping needs

Many people experience interacting combinations of mental health and neurodevelopmental conditions, physical health problems and social difficulties, rather than a single isolated condition. Although diagnostic categories remain essential for communication, research and service planning, they do not always capture this complexity.

Anxiety and depression frequently occur together and are often accompanied by trauma-related difficulties, self-harm, substance misuse or physical illness. Severe mental illnesses commonly coexist with substance misuse, chronic physical illness and social adversity. Together, these difficulties may create needs greater than

would be predicted by any single condition alone. Systems organised around single diagnoses may struggle when needs span several domains. People may move repeatedly between services, fail to meet individual service thresholds, or receive care that addresses one need while overlooking others.

Complexity extends beyond diagnosis. Many people experience overlapping combinations of psychological distress, social adversity, educational disruption, physical illness and economic disadvantage. These interacting difficulties often shape need more powerfully than any single diagnosis considered in isolation. The mechanisms linking mental health with education, employment, housing and social disadvantage are explored further in Chapters 10 and 16. The key point is that many people at risk of being missed are not defined by a particular diagnosis, but by needs that span multiple conditions, services and areas of life.

6.6.5 What can we conclude?

Prevalence, diagnosis and service use are incomplete measures of need. Some people remain unrecognised or under-represented in available data, while others cannot access support or are poorly served because their needs cross conventional service boundaries. Understanding who is missed, and why, is essential for designing systems that respond equitably to need.

6.7 Predicting future need

Substantial numbers of children and young people are currently experiencing mental health-related need.

If current trends continue, these numbers are likely to rise. By 2030, an additional half a million 5- to 24-year-olds may have self-harmed, bringing the total to 2.48 million.

These projections are approximate, but they reinforce the need for action to address both the causes of distress and the systems that respond to it.

The Review has set out why children and young people are a particular focus: levels of need in this age group are already substantial and have continued to increase.

To inform future service planning, estimates were made of the number of children and young people that may be affected in the future. Projections were based on populations aged 2- to 24-years or 5- to 24-years in 2030, depending on data availability, in two scenarios: one scenario based on estimated 2024 prevalences (i.e. prevalence in 2030 remains similar to prevalence in 2024) and a second scenario in which prevalence in 2030 is based on recent trends in different

conditions continuing over the coming years. The focus was on five indicators of mental health need: high psychological distress, self-harm, emotional disorders, conduct disorder and eating disorders. The methods are described in the Tables Annex (Annex B, Appendix 11).

Figure 6.25 Estimated number of children and young people with different types of mental health need in 2030 based on two scenarios (1) continued trends and (2) prevalence remains at 2024 levels

Source: Estimates based on APMS data presented in this report. Data available from Tables Annex (Appendix 11).

  • High psychological distress was experienced by almost 30% of young people

in 2024. If current trends continue, this could increase to around 32% by 2030, which is approximately 4.47 million children and young people (Figure 6.25).

  • The projected increase in self-harm may be even greater. By 2030, at current

prevalence rates almost 2 million 5-24 year olds may have self-harmed. However, if increasing trends continue at similar rates, this might be higher at 2.48 million. Because the population aged 2 to 24 years is projected to decline over the same period, these increases reflect rising prevalence rather than population growth.

  • These projections underline the importance of action both to prevent distress and to strengthen the systems that respond to it.

6.8 Synthesis and implications

Has mental ill health increased? The evidence supports a qualified yes – but also shows why the question must be framed carefully.

There is persuasive evidence that emotional distress has increased over recent decades. Across screening questionnaires, diagnostic interviews, cohort studies and administrative datasets, anxiety, depression, psychological distress and self-harm have all increased, particularly among adolescents, young adults and more recent generations. While increasing awareness and willingness to report difficulties have undoubtedly contributed, they do not fully explain the observed patterns.

Contemporary mental health cannot be understood through prevalence alone. Symptoms, diagnosis, recognition, service demand and need describe related but distinct aspects of the phenomenon. Their relationship varies across conditions and across different stages of the pathway from distress to support.

The evidence consistently points to functioning and participation as critical outcomes. Emotional difficulties remain strongly associated with educational disruption, social relationships, employment and wider participation. Their importance lies not only in how many people experience them, but in the extent to which they shape opportunities across the life course.

Development matters. Childhood, adolescence and early adulthood are the periods during which many mental health conditions first emerge, persist and influence later outcomes. Cohort studies suggest that more recent generations are following different developmental trajectories, indicating that developmental environments themselves have changed.

Mental health inequalities remain substantial. Distress, access to care and long-term outcomes are socially patterned, with some groups consistently experiencing greater burden and greater barriers to support. These inequalities mirror wider inequalities in opportunity and participation.

Finally, some of the greatest needs remain only partially visible. Diagnoses, administrative records and service use all underestimate the burden of mental ill health because some people are never recognised, others cannot access support, and others experience complex needs that are not easily captured within conventional systems.

These findings point towards a more nuanced understanding of contemporary mental health than is often reflected in public debate. The evidence supports neither a simple narrative that all mental health conditions are increasing uniformly nor the view that observed changes can be explained solely by greater awareness or changing diagnostic practice. Instead, contemporary trends reflect the interaction of changing distress, changing recognition, changing help-seeking and changing social contexts.

The central contribution of this chapter is conceptual. Prevalence, diagnosis, recognition, demand and need are related but distinct. Much of the current debate arises from treating these concepts as interchangeable. Understanding contemporary mental health requires attention both to each of these concepts individually and to the relationships between them.

Chapters 7 and 8 apply this framework to ADHD and autism. Chapter 10 considers wider social, educational and cultural influences, while later chapters examine prevention, service design and policy. The conclusion is cautious but clear: emotional distress has increased, especially among younger generations, with important consequences for functioning and participation. Effective responses therefore require systems that distinguish distress, diagnosis, demand and need, while addressing the social conditions that shape them.

Chapter 7. ADHD: population prevalence, symptoms, diagnosis and demand

ADHD population prevalence appears broadly stable or to have increased only modestly, while recorded diagnoses, referrals and waiting lists have risen rapidly, particularly since 2020. An English population survey (MHCYP) estimated ADHD prevalence of 5- to 15-year-olds as 3% and the narrower diagnosis of hyperkinetic disorder prevalence as 1.9% in 2017.

The increase in recorded ADHD reflects several interacting processes, including changes in recognition, self-identification, help-seeking, diagnostic practice, wider societal and cultural change, recognition of inattentive presentations, co-occurring mental health difficulties, changing environmental demands, and access routes such as Right to Choose (RTC) and independent provision. Growth has been greatest among adolescents, young adults and women.

Diagnosis remains important for understanding difficulties and guiding treatment but, on its own, does not determine need. Services should combine timely, high- quality diagnosis with support based on functioning, co-occurrence and context through more integrated, equitable neurodevelopmental and mental health pathways.

This chapter examines why prevalence, symptoms, self-identification, diagnosis and demand do not move in parallel, and what that divergence means for ADHD care in England. It builds on work undertaken by the ADHD Taskforce[203, 204].

7.1 What is known about ADHD prevalence (in England and internationally)

ADHD is one of the most common neurodevelopmental conditions but estimating how many people are affected is more complex than is often assumed. Administrative prevalence has increased more than three-fold since the mid-2010s, with further acceleration following the COVID-19 pandemic. Rising demand for NHS assessment and treatment, together with increasing self-identification of ADHD symptoms[205], has prompted considerable public and professional debate about the relative contributions of persistent under-recognition and possible overdiagnosis.

Prevalence estimates vary according to how ADHD is defined, measured and diagnosed, the populations studied, and the diagnostic framework applied. Consequently, no single figure can be regarded as the definitive prevalence of ADHD. Throughout this chapter it is important to distinguish between epidemiological prevalence, which estimates the proportion of the population meeting diagnostic criteria under research conditions, and administrative prevalence, which reflects the proportion with a recorded diagnosis within healthcare systems.

The best available population evidence in an English population sample is the MHCYP survey which estimated prevalence of ADHD according to DSM criteria as 3% of 5–15-year-olds in 2017 (Table 2.2, Annex). We estimate by projecting from existing MHCYP data points that the current prevalence among 5 to 15 year olds might be around 3.8%, but this is at best a rough estimate in the absence of contemporary data based on the modest rise between 2004 and 2017 continuing unchanged. The same survey estimated the prevalence of hyperkinetic disorder, a narrower diagnosis in ICD-10, at around 1.5% in 1999 and 1.9% in 2017[143]. More recent international reviews using the DSM-5 criteria which encompass a broader presentation, estimate ADHD prevalence at approximately 5–6% in children, 3–4% in adolescents and 3% in adults although estimates vary greatly according to how criteria are applied and on methodologies used (see Section 7.2)[206-208]. This is consistent with NICE guidance[209], which cites epidemiological prevalence of around 5% in children and young people and 3% in adults.

ADHD has no biological test. Diagnosis depends on identifying characteristic patterns of inattention, hyperactivity and impulsivity together with clinically significant functional impairment. Prevalence estimates therefore depend not only on the distribution of symptoms within the population, but also on where diagnostic thresholds are set and how impairment is assessed. Methodological differences (questionnaires, interviews and administrative records) and diagnostic frameworks (ICD versus DSM) produce differing estimates[210]. Across studies, prevalence is generally higher in children than in adolescents and adults[207].

Increasing evidence indicates that ADHD-related traits are continuously distributed across the population rather than forming a discrete category[211, 212]. Diagnostic thresholds therefore reflect professional judgements about severity, persistence and functional impact rather than natural boundaries. Successive revisions of diagnostic criteria, particularly DSM-5, have broadened eligibility by relaxing age-of-onset requirements, allowing co-occurring diagnoses and lowering adult symptom thresholds[213, 214]. Concepts such as impairment and pervasiveness require clinical judgement and may vary across clinicians and services[78].

The population meeting diagnostic criteria is therefore heterogeneous, ranging from individuals with severe, pervasive impairment to those whose difficulties are more context-dependent or less impairing. This reality motivates debate about the appropriate clinical threshold for diagnosis, but it does not imply that any particular diagnosis is inappropriate. Rather, it highlights that prevalence estimates reflect the interaction between underlying traits, diagnostic criteria and clinical judgement, rather than providing fixed measures of a discrete disorder.

Epidemiological prevalence estimates should therefore be regarded as approximate ranges and as benchmarks against which administrative prevalence can be interpreted. On their own, they do not explain patterns of diagnosis, referral or service demand.

7.2 Changes over time and methodological uncertainties

One of the most frequently asked questions about ADHD is whether it has become more common. The answer is not straightforward. Epidemiological trend data, including the MHCYP surveys described in the previous section and Global Burden of Disease analyses[215], suggest only small increases in ADHD prevalence over recent decades, in contrast to the much steeper increases in administrative diagnoses and referrals described in Section 7.3.

There are also important limitations in the available evidence. Although epidemiological prevalence estimates have remained broadly stable from the 1990s to the late 2010s, there are no contemporary UK population-based diagnostic prevalence estimates for children and young people after 2017, and no equivalent diagnostic prevalence estimates for adults. Consequently, estimates or projections based on earlier trends may not accurately reflect the current epidemiological prevalence of ADHD. Whether prevalence has remained stable over the past decade, increased modestly, or changed within particular population groups therefore remains uncertain.

Changes in diagnostic frameworks also influence prevalence estimates over time. DSM-5 criteria produce higher estimates than earlier DSM editions[208, 216]: child prevalence increases from around 8.3% under DSM-IV-TR to 11.3% under DSM-5, and adult prevalence from 3.3% to 5.1%. These broader criteria reflect the evolution of clinical understanding and allow a wider range of presentations to be recognised. They also mean that estimates produced using different diagnostic frameworks are not directly comparable, and that part of the apparent increase reflects broader diagnostic criteria rather than a change in underlying prevalence.

7.3 Administrative prevalence, referrals and waiting lists

Administrative prevalence captures the proportion of people assessed and diagnosed within healthcare systems. It reflects not only underlying prevalence, but also awareness, help-seeking, access, diagnostic practice and recording. Consequently, administrative prevalence can change much more rapidly than epidemiological prevalence.

Recent years show a striking divergence between relatively stable population estimates and rapidly rising administrative prevalence and demand. OpenSAFELY primary care data show recorded ADHD prevalence in the UK increasing from around 0.4% in 2016/17 to 1.3% in 2024/25, with particularly marked increases among adolescents and young adults (ages 10–17: 1.6% to 3.7%; ages 18–24: 1.6% to 3.1%). Notably, recorded prevalence increased sharply following the COVID-19 pandemic and has continued to rise rather than plateau.

Analysis of CPRD data (Clinical Practice Research Datalink) covering 2000–2025 shows the increase in administrative prevalence of ADHD from 0.8% to 1.4%, while incidence rose from 107 to almost 395 new diagnoses per 100,000 person-years over the same period[7, 217]. But this figure hides considerable heterogeneity with much of the increase in diagnosis being in children and young people. In boys aged 6-11 years, recorded prevalence rose from 0.7% in 2000 to 1.2% in 2010 to 3.8% in 2025 and for boys aged 12 to 17 years from 0.4% in 2000 to 1.7% in 2010 to 5.1% in 2025 respectively. Updated CPRD analyses of prevalence (Figure 7.1) and incidence (Figure 7.2) similarly demonstrate sharp post-pandemic increases across both sexes and all age groups.

Figure 7.1 Administrative ADHD prevalence by age band and sex, 2000–2025 Source: CPRD. Each panel has its own y-axis scale.

Figure 7.2 Administrative ADHD incidence by age band and sex, 2000–2025 Incidence per 100,000 at-risk person-years 6–11 years Children 12–17 years Adolescents 18–34 years Young adults 2000 2005 2010 2015 2020 2025 0 2000 2005 2010 2015 2020 2025 0 2000 2005 2010 2015 2020 2025 0 Incidence per 100,000 at-risk person-years 35–54 years Mid-adults 55+ years Older years Sex Male Female 2000 2005 2010 2015 2020 2025 0 2000 2005 2010 2015 2020 2025 0 Calendar year Each panel has its own y-axis scale Source: CPRD. Each panel has its own y-axis scale.

Although the number of new ADHD diagnoses has increased sharply since 2020, the proportion of people subsequently receiving medication has declined. This is unlikely to have a single explanation. Growing numbers of people experience delays in

accessing treatment following diagnosis and are unable to access timely support and intervention: the independent ADHD Taskforce found that only around 15 to 25% of people with ADHD receive medication, against the 70 to 90% who show benefit in trials[203]. However, the low prescription rate could be due to changes in case mix, prescribing practice or patient preference, and treatment discontinuation may also contribute[218, 219].

Referral rates have increased even more rapidly than administrative prevalence. Mental health services in England received around 5,000 referrals per month for suspected ADHD in 2019. By 2024 this had increased to approximately 17,000 referrals per month. Open referrals exceeded 560,000 by the end of 2025, more than seventeen times the number recorded in 2019[8]. Community paediatric waiting lists, the majority relating to ADHD assessments, reached approximately 173,000 individuals[8]. ADHD referrals nevertheless account for a relatively stable proportion of overall secondary mental health referrals, although this does not reflect the wider pattern seen in administrative prevalence.

Figure 7.3 Referrals received by mental health services per month for ADHD Source: NHS England, MHSDS Administrative indicators reflect not only underlying prevalence but also the way systems respond to demand. They are influenced by public awareness, referral pathways, diagnostic capacity, commissioning decisions, waiting times, perceived impairment and access to support. The substantial increase in referrals therefore reflects changing patterns of help-seeking as well as the organisation and capacity of services, and should not be interpreted as evidence of an equivalent increase in underlying prevalence. Nevertheless, demand continues to increase and has accelerated further during the most recent year of available data.

Figure 7.4 Open referrals (NHS and non-NHS) awaiting or in assessment, England, all ages, April 2019–December 2025

600k Autism ADHD 500k Open referrals (assessment waiting list) 400k 300k 200k 100k 2020 2021 2022 2023 2024 2025 0k Source: NHS England (KSM891) Geographical variation illustrates this clearly. Waiting times vary markedly across England, ranging from a matter of weeks in some areas to several years in others. Rates of diagnosis and treatment also differ substantially between regions, reflecting variation in local commissioning, service capacity and recording practices as well as underlying prevalence; the independent ADHD Taskforce found treatment rates varying up to twelve-fold across the UK[203] The expansion of Right to Choose (RTC) pathways and independent-sector provision has increased assessment capacity and provided alternative routes into care for many people facing long waits within NHS services. NHS spending on independent-sector ADHD services rose by around 252% over three years, with a projected overspend of £164 million across the ICBs analysed[18]. At the same time, these developments have altered referral patterns and highlighted the absence of consistent national standards for assessment, data collection, follow-up arrangements and shared-care prescribing.

The gap between epidemiological and administrative prevalence is often interpreted as evidence of under-recognition and unmet need, and there is strong evidence that many adults with ADHD remain undiagnosed. Administrative prevalence, however, should not be expected to converge completely with epidemiological prevalence. Not everyone meeting diagnostic criteria will seek, require or benefit from a formal diagnosis. If administrative prevalence approaches the underlying population prevalence, its rate of increase would be expected to slow and eventually plateau below the true prevalence. The absence of such a plateau suggests either that estimates of underlying prevalence are too low, or that administrative prevalence in

some groups is approaching, or possibly exceeding, those estimates – as may already be occurring among boys aged 12–17 years, where recorded administrative prevalence is around 5% (Figure 7.1).

In summary, administrative prevalence, referrals and waiting lists have increased far more rapidly than the best available estimates of underlying population prevalence. Understanding current pressures on ADHD services therefore requires attention not only to prevalence, but also to awareness, help-seeking, cultural change, diagnostic practice, co-occurrence, environmental demands and the organisation of services.

7.4 Inequalities by age, sex, socioeconomic status and ethnicity

ADHD diagnosis and service use vary by age, sex, socioeconomic disadvantage and ethnicity.

Age remains a strong predictor of diagnosis. Historically, rates were highest among school-aged children, particularly those aged 6-11 years[220]. More recent data show the greatest growth among adolescents and young adults. Administrative prevalence now peaks at ages 12 to 17 years (Figure 7.1), where CPRD data suggest recorded prevalence of around 5.1% in males and 2.7% in females by 2025, indicating increased recognition beyond mid-childhood.

ADHD diagnoses have typically been more common in males than females. Earlier studies reported male-to-female ratios of around 4:1 during childhood[220]. Recent evidence suggests that this gap is narrowing. Between 2017 and 2025, recorded prevalence in the CPRD analysis increased more than ten-fold among females compared with around three-fold among males. By 2025, recorded prevalence was 1.3% in females and 1.5% in males. This pattern is consistent with improved recognition in groups that have historically been under-recognised, particularly females and adults, and with increasing recognition of inattentive and internalising presentations that were previously overshadowed by the more overtly hyperactive and disruptive presentations, including conduct disorder, more commonly seen in boys[221] (see Section 7.5.1). Catch-up appears largely complete among adults, while Figures 7.1 and 7.2 suggest that a similar process is now occurring among children and adolescents. Socioeconomic disadvantage has long been associated with higher ADHD incidence and prevalence[220]. More recent CPRD data, however, show that although recorded prevalence remains higher in more deprived populations, the incidence ratio between the most and least deprived quintiles has narrowed from around 2.3 to approximate parity (0.93 in 2023–24 and 1.01 in 2025).

Several factors may contribute to this change. Historically, ADHD referrals were strongly associated with oppositional and conduct disorders in boys, conditions that are themselves closely associated with socioeconomic deprivation. More recently, increasing co-occurrence with autism and common mental disorders, together with presentations emerging later in development alongside educational and social impairment, may have reduced the strength of this deprivation gradient. Broader social, educational and occupational pressures may also be increasing help-seeking across all socioeconomic groups. These explanations are not mutually exclusive. More recently still, differential access to independent-sector services may also have contributed. The finding that ADHD medication prescribing appears more common in affluent areas is consistent with this interpretation[222].

Figure 7.5 ADHD deprivation gradient: Q5:Q1 ratio (administrative incidence and prevalence)

Source: CPRD Marked differences are also seen across ethnic groups. CPRD analyses show substantially higher recorded prevalence among White and Mixed ethnic groups than among South Asian, Black and Other groups – for example, around 2.0% among people of Mixed ethnicity and 1.7% among White people compared with 0.4% among South Asian groups and 0.7% among Black groups in 2025. These findings should be interpreted cautiously. Current evidence cannot determine the extent to which these differences reflect underlying prevalence, differences in help-seeking or access to services, cultural factors, or variation in pathways to diagnosis.

Figure 7.6. Administrative prevalence and incidence of ADHD by ethnic group over time

White Mixed Black South Asian Other Not recorded

Active five-year prevalence Recorded prevalence

Crude rate per 100,000 person-years Recorded incidence

2.00 Incidence per 100,000 at-risk person-years 1.75 Recorded prevalence, active 5-year (%) 1.50 1.25 1.00 0.75 0.50 0.25 2000 2005 2010 2015 2020 2025 0.00 2000 2005 2010 2015 2020 2025 0 Calendar year Source: CPRD

7.5 Changes in symptom prevalence

Evidence from both adult and child population surveys suggests that some ADHD symptoms have become more commonly reported over time, while others have changed little.

7.5.1 Adult inattention and hyperactivity symptoms (ASRS, APMS 2007, 2014 and 2024)

Using the six-item Adult ADHD Self-Report Scale (ASRS), symptom patterns can be compared across 2007, 2014 and 2024 in the APMS. The proportion screening positive increased from 8.2% in 2007 to 9.7% in 2014 and 13.9% in 2024. Among young adults aged 16–24 years, the increase was from around 14% to approximately 25% over the same period[205].

As shown in Figures 7.7 and 7.8, the increase in adult ADHD symptom reporting is concentrated almost entirely within the domain of inattention. Difficulties relating to concentration, organisation, planning and task completion increased across successive survey waves, whereas hyperactivity and impulsivity changed relatively little. Adults increasingly reported avoiding tasks requiring sustained mental effort, struggling to organise activities, having difficulty completing complex tasks and forgetting appointments or obligations. For example, the proportion reporting that they often avoided tasks requiring sustained mental effort increased from 12.1% in 2007 to 18.8% in 2024.

Figure 7.7. Item-level endorsement on the six-item Adult ADHD Self-Report Scale (ASRS-6), in adults aged 16+

Source: APMS. The four left-hand bars represent inattention items; the two right- hand bars represent hyperactivity items.

Figure 7.8. Mean ASRS-6 domain scores in adults aged 16+ Source: APMS. The mean inattention score has risen steadily, whereas the hyperactivity/impulsivity score has remained unchanged. Error bars show 95% confidence intervals.

Measurement-invariance analyses suggest that the scale functioned consistently across survey waves (Tables Annex, Section 1.1). Independent increases in concentration difficulties and forgetfulness strengthen the evidence for a genuine increase in self-reported attentional difficulties, although the magnitude of this increase remains much smaller than the rise in diagnosis and service demand.

Figure 7.9 shows the proportion of young adults screening positive for common mental disorders, inattention, autistic traits, and drug or alcohol dependence. All increased except drug and alcohol dependence. Among young adults with a common mental disorder, around 60% screened positive for inattention.

Figure 7.9 Trends in conditions screened or assessed for among young adults (16–24 years), APMS 2007–2024

Source: APMS. Shaded areas show 95% confidence intervals. AQ8+ = autistic traits screen (AQ-17 8+); ADHD = ASRS4+ screen.

7.5.2 Recalled age at onset of ADHD symptoms

Among people identifying with ADHD, the mean recalled age at onset of difficulties increased from 9.8 years in 2014 to 21.7 years in 2024. This may reflect later recognition or later emergence of clinically significant impairment when family, educational or occupational scaffolding falls away, particularly among women and people with co-occurring internalising or autistic difficulties.

Recall bias and changing self-understanding limit interpretation, but these findings indicate that the recent growth in demand is not explained solely by people with long- recognised childhood difficulties seeking delayed assessment.

One possible explanation is that symptom onset and the onset of clinically significant impairment do not necessarily coincide. ADHD-related traits may be present from childhood but remain accommodated by coping strategies or environmental support, becoming impairing only when educational, occupational or social demands exceed a critical threshold. In this interpretation, later reported onset reflects the point at which difficulties became noticeable or impairing, rather than the first emergence of underlying traits.

An alternative explanation is that many adults who identify with ADHD and do not recall childhood difficulties are remembering accurately because those difficulties were not present. Longitudinal cohort studies provide one way of examining this question because information is collected contemporaneously across the life course, reducing the influence of recall bias and retrospective reinterpretation. One of the best-known examples is the New Zealand Dunedin birth cohort, which followed 1,037 individuals born in 1972 through to 2010, achieving a 95% retention rate. Participants underwent repeated interviews, cognitive assessments, informant reports from parents and teachers, and genetic analyses. The poor outcomes among those with childhood-onset ADHD were consistent with the wider literature. However, those identified with ADHD in adulthood differed markedly. As Moffitt and colleagues[223] observed: ‘Unexpectedly, the childhood ADHD and adult ADHD groups comprised virtually non-overlapping sets; 90% of adult ADHD cases lacked a history of childhood ADHD. Also unexpectedly, the adult ADHD group did not show tested neuropsychological deficits in childhood or adulthood, nor did they show polygenic risk for childhood ADHD.’

These findings reinforce the conclusion that the recent growth in adult demand cannot be explained solely by people with long-recognised childhood ADHD seeking delayed assessment, and the authors do not attribute the lack of prospectively documented childhood ADHD to later overdiagnosis[223].

7.5.3 Children: parental perception of hyperactivity

MHCYP asked parents whether their child experienced significant difficulties with concentration, activity or sitting still. Reports of significant hyperactivity increased modestly between 1999 and 2017, with the increase concentrated among 11–15- year-olds and little change in younger children. The increase in parental reports of hyperactivity exceeded the increase in diagnosed ADHD over the same period, illustrating that symptom recognition and concern can increase without equivalent changes in administrative prevalence.

Figure 7.10 Parental perception of significant hyperactivity difficulties among children aged 5–15 years in England

Source: MHCYP

7.5.4 Self-identified ADHD in school-age young people

Self-identification has increased dramatically. In APMS 2024, 5.8% of adults reported having experienced ADHD, while 1.8% reported a professional diagnosis. Both figures represent substantial increases compared with previous survey waves. In the OxWell 2025 survey, approximately one in five participating secondary-school and further-education students reported some form of neurodivergence[14]. Across the full analytical sample, 17.6% reported having ADHD and 11.4% reported being autistic, including 5.7% who reported both. These categories included professional diagnosis (in approximately 28% with ADHD), being on a waiting list, self- identification and identification by others, and should not be interpreted as estimates of clinically confirmed prevalence[224].

Self-identification is not equivalent to diagnosis, but this increase shows that ADHD has become an important framework for understanding experience, with self- identification rising much faster than population prevalence or symptom reporting.

The relationships between traits, symptoms, self-identification, diagnosis and treatment provide important insights, but these should not be regarded as interchangeable measures. Each captures a different stage in a longer pathway. Traits describe experiences within the population. Symptoms and self-identification

reflect how those experiences are interpreted. Diagnosis reflects the application of clinical criteria. Treatment reflects access to support and intervention.

Understanding these relationships within an identification cascade (Figure 7.11) is essential for interpreting the marked increases in referrals, waiting lists and service demand.

Figure 7.11 The ADHD identification cascade in adults aged 16+ Source: APMS. 95% confidence intervals are shown.

APMS data show that the proportion of adults who believed they had ADHD increased approximately eight-fold between 2014 and 2024, while self-reported professional diagnosis increased around four-fold over the same period. In contrast, increases in symptom measures were much more modest.

It is important to note that screening positive on the Adult ASRS indicates that an individual exceeds the threshold on a screening questionnaire; it is not equivalent to receiving a clinical diagnosis. Nevertheless, impairment among those screening positive remains substantial. They report markedly greater difficulties in everyday functioning, relationships, loneliness, stress and mental health than those screening negative. Across several domains, the gap in impairment between those screening positive and those screening negative has remained stable or widened over time.

Co-occurrence and complexity have also become increasingly prominent (see Chapters 6 and 9). Among adults screening positive on the six-item ASRS, receipt of any mental health treatment rose from 20.2% (2007) to 37.8% (2024); autistic traits increased from 27.2% to 45.6%; and rates of self-harm, suicide attempts and drug dependence remained high. Among 16–24-year-olds with a common mental disorder

in the APMS, 57% screened positive for ADHD traits, 61% for autistic traits, 36% for probable PTSD and 52% reported self-harm. Rising ADHD-related demand may therefore reflect attentional difficulties becoming more visible and more impairing in the context of increasing co-occurrence. Other possibilities are that people are increasingly interpreting their distress through several available explanatory frameworks, or that current clinical differentiation does not fully capture the underlying overlap between conditions.

Figure 7.12. Mental-health treatment receipt among adults screening positive for ADHD (ASRS-4+)

Source: APMS

One possible explanation for rising ADHD identification and diagnosis is that increasing awareness has encouraged growing numbers of people with relatively mild difficulties to recognise themselves in descriptions of ADHD and seek assessment. If this were the case, average levels of impairment associated with ADHD symptoms might be expected to decline over time.

However, despite increasing heterogeneity and broader diagnostic criteria, this has not been observed among adults screening positive for ADHD symptoms (but not diagnosed) in the most recent APMS (2023/24), who continued to report substantially greater difficulties across almost every domain of functioning than those screening negative[15]. Difficulties were particularly evident in managing everyday activities, responsibilities, close relationships, stress, loneliness and social isolation. In 2024, adults screening positive for ADHD were several times more likely to report

difficulties managing paperwork, handling finances, maintaining relationships and coping with the demands of daily life. More than half (56.1%) reported loneliness and social isolation, compared with 19.3% of those screening negative.

Another important indicator of ADHD-related impairment is the increased risk of being not in education, employment or training (NEET), discussed further in Section 7.5.7. These differences have not diminished over time. Indeed, there is some evidence that disparities relating to independence, self-management, social connection and psychological wellbeing have widened.

It is possible that the severity of impairment is changing in some groups. For example, the association between ADHD diagnosis and poorer educational outcomes among children and young people appears to be weakening in South London[218, 219] and in Norwegian data, although the South London findings derive almost entirely from NHS clinical populations and capture relatively little of the assessment now undertaken in the independent sector.

At the same time, interpretation requires caution. The period during which ADHD- related impairment has remained elevated coincides with broader increases in common mental health difficulties across the population. Some of the observed impairment may therefore reflect the cumulative and interacting effects of co- occurring conditions rather than ADHD symptoms alone. This interpretation is consistent with evidence presented elsewhere in this chapter and in Chapter 9 showing increasing rates of anxiety, depression, self-harm, substance misuse and autistic traits among people screening positive for ADHD symptoms.

A Danish registry study compared 71,317 children diagnosed with ADHD or autism between 2012 and 2022 with 713,170 matched controls across nineteen pre- diagnostic characteristics. As annual diagnoses nearly doubled, the differences between cases and controls narrowed — the odds ratio for low birth weight fell from 1.54 to 1.17 — with attenuation greater for ADHD than autism and among those diagnosed at older ages. The authors conclude that this reflects changing diagnostic practice and capacity, and caution explicitly that it is not evidence that either condition has become less impairing[225].

7.5.6 Impairment, function, support needs and outcome in ADHD

ADHD diagnosis alone does not determine support needs because functional impairment varies widely within the diagnosed population. Some individuals who meet DSM-5 criteria function well with modest support, whereas others experience severe impairment across multiple domains requiring intensive, multi-agency intervention. Symptom counts, DSM subtypes and severity specifiers are relatively

poor predictors of functioning and long-term outcomes, and impairment may persist even when symptoms remit[226-229].

Two broad sets of factors consistently predict outcomes and treatment response more strongly than symptoms alone: functional impairment and co-occurring conditions. Functional impairment across key domains of life – including education or employment, family relationships, social functioning, emotional regulation, daily living, and risk or safety behaviours – is a stronger and more independent predictor of clinical outcomes, treatment response and persistence than symptom count alone[230, 231]. Impairment and symptoms are related but partially distinct constructs. They become less closely aligned during childhood and adolescence but converge more strongly in adulthood, and impairment may persist even after symptom remission[232]. Diagnosis therefore captures only part of what determines need. Different domains of impairment are associated with distinct developmental trajectories: social impairment predicts later depression, academic impairment predicts educational non-completion, and emotional dysregulation independently predicts later psychiatric impairment beyond ADHD symptoms[233, 234].

The second major determinant is the pattern and burden of co-occurring conditions described in Chapter 9. ADHD commonly co-occurs with autism, anxiety, depression, substance misuse and learning difficulties[235, 236]. Cumulative co- occurrence predicts severe multi-domain impairment in a dose-dependent manner[237, 238]. Externalising conditions, including conduct disorder and oppositional defiant disorder, predict delinquency, substance misuse, educational failure and persistence of symptoms[239, 240], whereas internalising conditions predict emotional outcomes, quality of life and occupational disability[234]. Data-driven patterns of co- occurrence are themselves associated with distinct profiles of impairment and service use[237]. For many individuals, it is the accumulation and interaction of these difficulties that determines the level of need.

Persistence of ADHD across the life course is the strongest single predictor of adverse long-term outcomes. Important baseline predictors include clinician-rated functional impairment, co-occurring conduct difficulties, lower intelligence quotient (IQ), and the absence of compensatory support, all of which can be assessed at referral and should inform needs assessment[230, 240, 241]. Assessment should therefore incorporate environmental context as an active modifier of need rather than treating it as background information (see Chapter 18).

Understanding ADHD therefore requires attention not only to symptoms and diagnosis, but also to co-occurring conditions, environmental pressures and functional need. This supports a stratified model of care in which diagnosis is complemented by structured assessment of functional need. Relevant dimensions

include symptom severity and presentation, affective and anxiety burden, cognitive profile, educational and occupational functioning, social and family circumstances, and engagement with previous treatment[242]. Each contributes information that a diagnostic label alone cannot provide, and each has implications for the type and intensity of support required.

7.5.7 The relationship between ADHD and not being in education, employment or training (NEET)

ADHD can have important consequences for functioning and participation in education and employment.

MHCYP data show that children diagnosed with ADHD are far more likely to be excluded from school than those without a mental disorder (15% versus <1%), with permanent exclusions around sixteen times higher[143]. After adjustment for co- occurring conditions, children with ADHD remain around four times more likely to experience permanent exclusion. Children with attention difficulties are also less likely to participate in the wider social life of school, including extracurricular activities and school clubs. Educational disengagement may therefore occur through multiple pathways, not only through disciplinary processes but also through gradual withdrawal from opportunities for social participation and belonging. Millennium Cohort Study follow-up data show that young people diagnosed with ADHD before the age of 17 are around four times more likely to be NEET in early adulthood than those without ADHD – a risk similar to autism and greater than that associated with many other long-term conditions.

The Taskforce read the same relationship in the other direction, estimating that between 20% and 34% of young people not in education, employment or training are likely to have ADHD, most of them undiagnosed[203]. In adults, ADHD symptoms are much more common among people who are unemployed, economically inactive or receiving out-of-work disability benefits than among those in employment. By 2024, over one-third of unemployed adults and around 40.5% of adults receiving out-of- work disability benefits screened positive for ADHD traits[15]. These are screening- positive individuals, however; no clinical assessment was undertaken to confirm a diagnosis.

ADHD is also over-represented within the criminal justice system. One meta-analysis estimated ADHD prevalence at 25.5% among prison populations[243]. This is discussed in greater detail in Chapter 9 (Section 9.8).

Given concerns about rising economic inactivity and increasing NEET rates, ADHD may be one contributing factor, although it is clearly not the sole explanation.

Whether earlier recognition and support can themselves reduce NEET risk remains an important question that current evidence cannot yet answer.

7.6 Self-identification and help-seeking

One of the most significant recent changes in the ADHD landscape has been the growth of self-identification. Increasing numbers of people now recognise ADHD as a possible explanation for their difficulties and, on that basis, seek information, assessment or support. Self-identification has become an important pathway into services and has likely contributed substantially to rising demand.

Self-identification is distinct from clinical diagnosis. Diagnosis involves structured clinical assessment of symptoms, developmental history, functional impairment, alternative explanations and co-occurring conditions. Self-identification reflects how individuals interpret their own experiences and decide whether ADHD provides a useful explanatory framework. Polling undertaken for this Review shows clear age- related differences between the proportion identifying themselves as having ADHD and the proportion reporting a professionally confirmed diagnosis (see Chapter 5). Among those identifying themselves as having ADHD, only around half (52.3%) reported a professional diagnosis. The proportion was highest among 16–29-year- olds (58.2%) and lowest among those aged 55 years and over (39.3%) (Figures 5.2 and 5.3).

Growth in self-identification appears to reflect several interacting developments. Public awareness of ADHD has increased markedly, and health and education systems may inadvertently encourage self-identification because diagnosis is often perceived as the route to support. ADHD is far more visible in schools, universities, workplaces, healthcare settings and public discourse than it was a decade ago. For many people, greater awareness helps make sense of longstanding difficulties, reduces self-blame, encourages help-seeking and facilitates access to support.

Digital environments also play an important role. Social media, online communities, podcasts, websites and video platforms are now major sources of ADHD information. Many people report that online content prompted them to seek further information or a formal assessment[244]. However, the quality and accuracy of online information vary. Studies of ADHD-related social media content identify substantial inaccuracy, oversimplification and blurring of the boundaries between ADHD and broader emotional or psychological difficulties[244, 245]. Common experiences such as anxiety, mood instability and stress-related concentration difficulties are often portrayed as uniquely characteristic of ADHD despite occurring across many conditions[245]. Individuals may therefore identify with ADHD without distinguishing it from anxiety, depression, stress, sleep disturbance or other causes of attentional difficulties.

Self-identification and diagnosis are different in kind. Self-identification can be a valuable first step towards understanding experiences and seeking help, whereas clinical assessment determines whether ADHD is the most appropriate explanation, identifies co-occurring conditions, evaluates impairment and guides intervention. Importantly, self-identification has increased far more rapidly than symptom prevalence.

Growth in self-identification also reflects a broader shift in how difficulties are understood, exemplified by changing ideas about neurodiversity (see Chapter 11). Rising demand for ADHD assessment cannot be understood through changes in prevalence alone, but also through changing patterns of recognition, help-seeking and access to services. Long waits for NHS assessment have further increased the use of independent-sector providers, reinforced by advertising.

7.7 Under-recognition, over-diagnosis and misdiagnosis

Public discussion of ADHD is often framed as a debate between underdiagnosis and overdiagnosis. Chapter 4 defines these terms, and Chapter 5 argues that this framing is too simplistic. The independent ADHD Taskforce proposed a more practical vocabulary for the stages between a trait and a clinical diagnosis, distinguishing the ADHD spectrum, suspected ADHD, possible ADHD (above a validated screening threshold) and clinical diagnosis, established through comprehensive assessment[203]. These distinctions matter because each group has different needs, only the last requires clinical diagnosis, and much public debate about over- and underdiagnosis fails to distinguish between them. Nevertheless, when diagnosis becomes a principal gateway to support, greater awareness may also increase the risk of over-identification or excessive pathologising in some contexts[246, 247]. This includes concern that normal variation in attention, activity, maturity or stress-related difficulties may sometimes be interpreted as ADHD. For example, studies have raised concerns about immaturity being diagnosed as ADHD in some school-age contexts[248], while evidence that many adults presenting with ADHD report no history of childhood ADHD raises questions about how consistently ADHD is being identified as a childhood-onset neurodevelopmental condition[223].

Chapter 5 discusses the concept of diagnostic overshadowing. Although recognised across medicine, it has particular relevance to ADHD because it can contribute to under-recognition.

Difficulties with attention, organisation, emotional regulation and executive functioning have often been attributed to depression, anxiety, substance misuse, bipolar disorder, personality disorder or age-related cognitive decline without ADHD

being considered[71, 249]. For many people, an ADHD diagnosis is reached only after years of treatment for anxiety or depression. This does not necessarily mean that the earlier diagnoses were incorrect. Rather, they may have been incomplete: the initial diagnosis overshadowed the presence of ADHD, which was recognised only later. Although difficult to quantify directly, diagnostic overshadowing may help explain why administrative prevalence has remained substantially below epidemiological estimates, particularly among adults. The Taskforce reached a comparable conclusion for children, finding that fewer than one third of those meeting diagnostic criteria for ADHD reach mental health services at all[203].

“I had been living with undiagnosed ADHD reasonably successfully until the birth of my first daughter. All my own coping mechanisms collapsed, and emotional deregulation took over, which made me behave in a way I didn’t want to, which fed my distress in a vicious circle. It has taken diagnosis and medication to recover from this, and I still live with a feeling of guilt and upset about how I behaved.

I approached various GPs at various points throughout my adult life about emotional deregulation, and latterly depression. I have been offered (and completed) a lot of Cognitive Behavioural Therapy courses which made no difference, but no one was curious about why, or considered the wider possibilities of how I was presenting. The opportunity for earlier diagnosis was there, but I was met with a total lack of curiosity by primary care.

After many years of not receiving help and becoming distressed by the side effects of anti-depressants I was taking to deal with the stress of living with my symptoms, I was shown a list of ADHD symptoms that aligned perfectly with the issues I had taken to the GP for the last 20 years. This time I went to the GP and asked directly for an ADHD assessment, which ultimately led to the life-changing diagnosis and medication. I have never had to undertake such self-diagnosis for physical symptoms.”

Lived Experience Partner Evidence for under-recognition does not imply that every increase in diagnosis should automatically be regarded as beneficial. A key question is whether diagnosis improves outcomes through understanding, support, treatment and participation – or whether, in some circumstances, it may not (see Chapter 5 for a broader discussion). Some commentators have argued that broader diagnostic criteria and variable thresholds may increase diagnoses among people with milder ADHD symptoms, particularly where diagnosis is the principal route to support[250]. On the other hand, the Norwegian studies reviewed in Chapter 5 raise questions about whether this necessarily leads to better outcomes. The Kazda review highlighted both benefits (empowerment and improved understanding) and potential harms (disempowerment and stigma), concluding that there is insufficient evidence to determine the overall balance of benefit and harm across the full spectrum of severity. Comparisons of prevalence can indicate potential unmet need, but they cannot determine whether individual diagnoses are helpful, unnecessary or harmful. That requires evidence on functioning, wellbeing, participation and quality of life, rather than diagnosis alone.

Misdiagnosis adds a further layer of complexity. Difficulties with attention, concentration, organisation and impulse control are not unique to ADHD and may arise in anxiety, depression, trauma, sleep disturbance, substance misuse and many other conditions. ADHD also frequently co-occurs with these difficulties, making differential diagnosis challenging.

The clinically important question is therefore not simply whether ADHD is present or absent, but how it interacts with other conditions. Individuals may be under- recognised, over-diagnosed or misdiagnosed because the complexity of their presentation is not fully understood. They may receive no diagnosis when one is warranted, a partial explanation that overlooks co-occurring conditions, or a diagnosis that captures only part of a broader pattern of need because assessment has been too limited or too narrowly focused.

“Two years after my autism diagnosis, I still felt there was a missing piece. As the demands of university increased, I was struggling to cope, and even simple daily tasks felt increasingly impossible. I was referred for an ADHD assessment, but after waiting more than a year while my difficulties continued to worsen, I eventually paid for a private diagnosis.”

Lived Experience Partner Evidence from highly vulnerable populations illustrates these issues further (see Chapter 9). Children in secure settings have elevated rates of ADHD (12–30%), and around 35% of children in the Children and Young People’s Secure Estate identify as having ADHD[251]. Many also have communication difficulties, learning disabilities, traumatic brain injury and extensive adverse childhood experiences (ACEs)[252]. Delayed diagnosis and limited access to support, including medication, are common[251].

Reviews of girls in secure settings show that ADHD often remains unrecognised until difficulties are severe and entrenched, with presentations interpreted primarily through trauma, emotional distress, self-harm, exploitation or behavioural concerns[253, 254]. Although children in secure settings are not representative of the wider population, these findings reinforce broader evidence that ADHD presentations

in girls and young women may be less readily recognised within pathways historically shaped around more overtly hyperactive and disruptive presentations in boys. For these girls, delayed recognition may contribute to the accumulation of educational, social and mental health difficulties.

7.8 Interventions and outcomes: benefits, risks and variation

Diagnosis can have important benefits. For many people, it explains longstanding difficulties, makes sense of past experiences, reduces self-blame, enables access to support and reasonable adjustments, and opens opportunities for evidence-based interventions. Diagnosis, however, is neither an endpoint nor an intervention in itself. Its value depends on what follows: timely, appropriate support, treatment and environmental adaptation.

Evidence reviewed here and by the ADHD Taskforce indicates that both psychosocial interventions and medication can improve outcomes for many people with ADHD, although effects vary substantially between individuals and long-term outcomes remain uncertain. The key questions are which interventions benefit which people, under what circumstances, at what stage of life, and how service organisation influences their effectiveness.

Evidence suggests that around 60% of those diagnosed during childhood continue to experience clinically significant symptoms in adult life. A scoping review also suggests that individuals with similar clinical ADHD profiles have better long-term outcomes in supportive, well-resourced environments than in more adversarial or under-resourced contexts[255].

Low socioeconomic status, parental psychopathology, family instability and the absence of compensatory support each independently predict poorer outcomes beyond the clinical presentation itself[72, 240, 256]. Need therefore cannot be inferred from diagnosis alone but must be interpreted in context. Social adversity predicts outcomes independently, meaning that diagnosis and treatment cannot substitute for wider social change. Improving outcomes for people with ADHD, and for those with ADHD-like difficulties, therefore also requires attention to the socioeconomic and environmental conditions that shape symptom expression, impairment and cumulative disadvantage.

7.8.1 Psychosocial interventions

Psychosocial interventions include parent training, behavioural approaches, psychoeducation, cognitive behavioural therapy (CBT), educational support and

broader psychosocial programmes. Their aims range from symptom reduction and emotional regulation to improvements in relationships, educational participation and everyday functioning.

Overall, the evidence base is less robust than for medication[257], but several approaches demonstrate meaningful benefit. Parent training for preschool children has the strongest evidence among non-pharmacological interventions. Across multiple trials, these programmes improve parenting practices, reduce disruptive behaviour and produce modest improvements in ADHD symptoms[258, 259]. NICE therefore recommends parent training as the first-line intervention for preschool children before medication is considered[260].

Among adults, CBT has moderate evidence of benefit, particularly for symptom management, coping strategies and functional difficulties[257]. Some studies also suggest benefits from mindfulness-based approaches, although the certainty of this evidence remains low.

A recurring and important finding is that successful treatment extends beyond reducing the core symptoms of hyperactivity, impulsivity and inattention. Improvements in parenting, emotional regulation, educational participation, relationships and quality of life may be just as important as changes in ADHD symptoms themselves. However, many psychosocial studies remain small, follow-up periods are often short, and high-quality long-term evidence is limited. Psychosocial interventions should therefore be regarded as important components of care whose effectiveness varies according to age, context, co-occurring conditions and individual goals.

7.8.2 Medication

Medication has the strongest evidence base and is recommended by NICE as first- line pharmacological treatment for most children over five years of age, adolescents and adults[260]. Stimulant medications include methylphenidate, lisdexamfetamine and dexamphetamine, while non-stimulants include atomoxetine and guanfacine.

Umbrella reviews and meta-analyses consistently show that medication reduces core ADHD symptoms, often with larger and more reliable effects than psychosocial interventions alone[257]. Estimated stimulant effect sizes for core symptoms (around 0.7–0.9) are among the largest reported for any psychiatric intervention[203]. Benefits have been demonstrated across all age groups, although response varies between individuals and between medications. For some people, medication enables engagement with psychosocial interventions that would otherwise be difficult.

Potential benefits extend beyond symptom reduction. ADHD is associated with educational underachievement, accidents, substance use disorders, self-harm, suicide and premature mortality[261]. Longitudinal studies also associate ADHD medication with reductions in accidents, substance misuse, suicidal behaviour, criminality and premature mortality, although the strength of these associations and their causal interpretation vary[106, 262]. These broader outcomes are important because many of the long-term consequences of ADHD arise through education, relationships, employment and safety rather than through symptoms alone.

“At 43 I was diagnosed with ADHD and prescribed medication… The medication has helped me in multiple ways, with the most profound being able to gain control of the emotional deregulation I have suffered with my entire life. This has allowed me to rebuild my family relationships and continue living with my children.”

Lived Experience Partner Medication is not universally effective and has recognised adverse effects. Common side effects include reduced appetite, sleep disturbance, anxiety, headaches, gastrointestinal symptoms and mood changes[222]. Rare but more serious adverse events include cardiovascular complications, psychotic symptoms and suicidal behaviour[222]. However, a large cohort study found no increased rate of psychotic events following initiation of methylphenidate among adolescents and young adults, including those with a previous history of psychosis[263]. NICE recommends regular monitoring and review throughout treatment[264].

Because stimulant medications act on dopaminergic systems, concerns have been raised about dependence, misuse and diversion[265]. National surveillance data from England, however, provide little evidence of widespread harm arising from prescribed ADHD medication[222]. Mortality data, pharmacovigilance systems, poison information services and drug treatment monitoring have not identified large-scale dependence or rapidly increasing medication-related harm despite substantial growth in prescribing. Police seizures of methylphenidate in England averaged around 230 tablets or capsules a year between 2020 and 2025, which is a fraction of the annual dispensed volume. Seizure data capture only part of any diversion, most of which occurs informally between individuals, so this evidence should be read alongside the mortality, pharmacovigilance and treatment-monitoring data rather than on its own.

The harms review undertaken for this Review found that psychiatric symptoms – including anxiety, depression and mood disturbance – were the most frequently reported adverse effects, rather than dependence[222]. This reinforces the view that medication should be considered one component of a broader package of care,

including monitoring, psychoeducation, psychosocial support and attention to co- occurring conditions.

There is also evidence that medication may reduce some substance-related harms. ADHD is over-represented among people with substance use disorders and is frequently under-recognised[222]. Effective treatment of ADHD may reduce subsequent substance misuse and improve outcomes within these populations[266], highlighting the need to weigh the harms of untreated ADHD alongside the potential risks of medication.

Despite the substantial evidence supporting medication, important uncertainties remain. Most evidence concerns short- and medium-term symptom reduction. High- certainty evidence regarding long-term effects on functioning, participation, education, employment and quality of life remains limited[257], largely because of the practical and ethical challenges of conducting long-duration randomised controlled trials. Discontinuation trials suggest that symptomatic benefit can persist for at least two years[267], although they do not address longer-term functional outcomes. The current evidence therefore reinforces the importance of ongoing outcome monitoring and regular review in routine practice.

Overall, medication is an important and often highly effective intervention for many people with ADHD. It carries recognised risks and side effects, but current evidence does not support the view that increased prescribing has produced widespread medication-related harm at the population level. Treatment decisions should be guided by careful consideration of benefits and risks, individual preferences, co- occurring conditions, functional impairment and wider life circumstances. Shared decision-making, long-term follow-up and effective shared care between GPs and specialists remain essential components of high-quality ADHD care.

7.9 What remains uncertain

The major uncertainties concern contemporary prevalence, the consistency of measuring impairment and applying diagnostic thresholds, patterns of co- occurrence, long-term outcomes, the benefits and harms of diagnosis, and the future trajectory of demand.

Current evidence suggests that ADHD is common and that underlying prevalence has probably remained broadly stable over recent decades. However, England lacks population-based diagnostic prevalence estimates for children after 2017 and for adults, and national education datasets do not include a dedicated ADHD code. As a result, it is not possible to determine confidently whether underlying prevalence has remained stable, increased modestly, or changed within particular population groups.

This lack of data has been a persistent constraint on the Review (see Chapter 20, Recommendation 8).

Self-identification without clinical assessment is also increasing. The proportion of adults screening positive on the ASRS rose from 8.2% in 2007 to 13.9% in 2024, with the increase concentrated in inattention. However, concurrent increases in anxiety, depression, loneliness and self-harm mean that changes specific to ADHD cannot yet be separated confidently from broader increases in cumulative distress.

As discussed throughout this chapter, ADHD-related traits are continuously distributed across the population rather than separated by a clear natural boundary. Diagnostic criteria therefore require thresholds to be applied to dimensions of behaviour and functioning. Changes in these thresholds, and in how they are interpreted, may influence both prevalence estimates and diagnostic rates.

The evidence base for ADHD interventions also remains limited in important respects. Evidence is strongest for short-term symptom reduction, whereas long- term effects on education, employment, social functioning, quality of life, and physical and mental health remain less certain[209, 257].

Diagnosis may provide explanation and access to treatment, but evidence regarding its overall balance of benefit and harm near diagnostic thresholds remains limited. Demand is likely to continue increasing, but the relative contributions of awareness, prevalence, symptom expression and co-occurrence remain uncertain and have different implications for workforce planning.

7.10 Drivers of the increase

The evidence reviewed suggests that no single explanation is sufficient to account for the rapid increase in ADHD-related diagnosis, referrals and service demand observed in recent years. Historical under-recognition remains important. However, recent accelerations in diagnosis have been particularly marked among females since the pandemic[221, 268, 269]. Diagnostic substitution may also operate in both directions: ADHD may previously have been attributed to anxiety, depression, substance misuse or conduct disorder, while contemporary attentional difficulties may instead reflect anxiety, low mood, trauma, sleep disturbance or stress[71, 249].

The most plausible explanation is that several interacting processes are operating simultaneously. These include: some increase in attentional difficulties; improved recognition of historically under-recognised groups; increasing public awareness and self-identification; broader cultural change; changing environmental demands, including digital environments and the effects of the pandemic on education and

work; evolving diagnostic concepts; increasing co-occurrence and complexity; interactions between underlying traits, mental health conditions and environmental factors; and systems in which diagnosis increasingly functions as the gateway to support; a conclusion the independent ADHD Taskforce reached separately, identifying the perceived requirement for a diagnosis in order to obtain support – particularly in education – as a principal driver of referral growth[203]. Longitudinal evidence will be needed to distinguish changes in ADHD itself from changes in co- occurrence and cumulative complexity.

7.11 Summary

First, ADHD is associated with substantial impairment and disadvantage across multiple domains of life. The evidence reviewed here demonstrates associations with poorer educational outcomes, school exclusion, social isolation, unemployment, economic inactivity, mental health difficulties and reduced participation, irrespective of debates about prevalence.

Second, diagnosis remains important for understanding, prognosis and treatment planning. For many people, diagnosis provides an explanation for longstanding difficulties, reduces self-blame, facilitates access to support and medication, and creates opportunities for effective intervention.

Third, diagnosis alone is an incomplete guide to need and support. Diagnostic categories provide valuable information, but they do not fully capture the complexity of real-world presentations. Functioning, context, co-occurrence and participation all influence the type and intensity of support required.

Fourth, there is no simple explanation for the rapid growth in ADHD-related demand. Neither a dramatic increase in underlying prevalence nor explanations based solely on either historical underdiagnosis or current misdiagnosis adequately account for the observed patterns. Rather, the evidence is most consistent with several interacting processes, including: some increase in attentional difficulties; improved recognition of previously under-recognised groups; changes in self-identification and help-seeking; evolving diagnostic concepts; increasing complexity and co- occurrence; changing educational and occupational environments; digital technologies; and changes in the organisation of services.

Accurate and timely diagnosis should therefore be combined with needs-led support focused on functioning, wellbeing and sustainable participation.

Chapter 8. Autism: population prevalence, recognition and system response

Earlier English population studies estimated prevalence of around 1.0–1.2%, although no recent study provides a definitive estimate. Recorded cumulative administrative prevalence in CPRD rose from 0.04% in 2000 to 1.72% in 2025, and now exceeds 4% in some age groups, with the sharpest increases among children, young people, girls and women, and people without intellectual disability.

The evidence points to changes in recognition, recording, diagnostic concepts, help-seeking and access, rather than a clear increase in underlying population prevalence. Autism is highly heterogeneous and commonly co-occurs with ADHD, mental and physical health conditions, and intellectual disability. The need for support depends on functioning, participation, environment, support and adjustment, as well as diagnosis.

Current systems remain too dependent on diagnosis. Rising demand has resulted in long waiting times, variable assessment quality, limited post-diagnostic support and unequal access, leaving many people without timely help. Earlier recognition and support, shaped around a person’s environment, education and everyday participation, may reduce the accumulation of mental health and other difficulties that autistic people too often experience while waiting. At the same time, the increasing number of people seeking to understand themselves and access support is, in many respects, a welcome development. The policy implication is that diagnosis should continue to play an important role in informing treatment and specialist intervention, while support should be organised around functioning, co-occurring conditions, environment and need.

8.1 Understanding autism across the life course

Autism is a neurodevelopmental condition characterised by differences in social communication and interaction, together with restricted and repetitive behaviours, interests or activities, and sensory differences[270]. Diagnosis requires these characteristics to be present from the early developmental period and to cause clinically significant impairment. Clinical diagnosis is not synonymous with autistic

characteristics or self-identification, nor does it by itself determine need or prognosis[271, 272].

Autism has no single developmental trajectory. Characteristics may be evident early in life or become clinically significant when social, educational or occupational demands exceed an individual’s coping strategies. Some children experience early developmental regression, most commonly affecting language or social responsiveness[273-276].

Autism is shaped by the interaction between the individual and their environment. Predictable, supportive settings may reduce difficulties, whereas uncertainty, sensory load and complex social demands may increase them. Functional impact therefore reflects person-environment fit, and support needs may change across developmental stages and life transitions[47].

Co-occurring intellectual disability, ADHD, anxiety, depression, epilepsy, sleep disorders and other physical health conditions are common and often shape need more than autism alone[277-279]. Chapter 9 considers co-occurrence in more detail.

Autistic characteristics are heritable, polygenic and continuously distributed across the population[44]. This does not mean that ‘everyone is a little autistic’. Rather, autism is a clinically meaningful condition defined by the pattern, persistence and functional impact of autistic characteristics. The diagnosed autistic population is also highly heterogeneous. Some groups – notably autistic people with severe or profound intellectual disability, minimal functional speech, complex epilepsy or very high support needs – have qualitatively different developmental profiles, co-occurring conditions and lifelong support requirements. Diagnostic thresholds exist to identify those whose characteristics are associated with clinically significant impairment and who are likely to benefit from support, intervention or adjustment.

Autistic people face increased risks of poor physical and mental health, lower educational participation, under-employment and premature mortality[277, 279-284]. These outcomes are not inevitable. Rather, they arise through the interaction between autistic characteristics, co-occurring conditions, access to support, and the extent to which families, educators, employers, services and wider society are able to accommodate difference.

Diagnosis alone is therefore insufficient for understanding need, which also requires attention to functioning, participation, co-occurring conditions, environment and support across the life course. Needs-led support should complement, rather than replace, timely diagnostic assessment.

8.2 What is known about autism prevalence and identification

There is little evidence of a substantial change in underlying epidemiological prevalence (see Chapters 3 and 4 for definitions). What has clearly increased is the number of people recognised, recorded and diagnosed as autistic within health and education systems.

8.2.1 Children and young people

Earlier English population studies estimated childhood prevalence at around 1.0– 1.2%[285]. The 2017 Mental Health of Children and Young People survey produced a similar estimate for 5–19-year-olds, although it was not a dedicated comprehensive autism ascertainment study and predated the recent acceleration in identification. Administrative data show a different pattern. Cumulative recorded administrative prevalence in CPRD analysis showed an increase in diagnosed autism in boys aged 6-11 years from 0.4% in 2000 to 1.6% in 2010 to 6.3% in 2025, and in boys aged 12- 17 years 0.2% in 2000 to 1.2% in 2010 and to 4.7% in 2025[9, 10]. These findings provide compelling evidence of increasing administrative prevalence, but not of an equivalent increase in underlying epidemiological prevalence. Figure 8.1 shows how recorded autism prevalence varies by age, with the highest recorded rates observed among children and young people. Autism-related SEND (Special Educational Needs and Disability) rose from 1.5% of pupils in 2016 to 3.3% in 2024. Figure 8.2 shows that recorded autism prevalence has increased substantially over time regardless of whether it is measured annually, over a five-year period, or cumulatively.

Figure 8.1 Administrative prevalence of autism by age group Recorded ASD prevalence, active 5-year (%) Age group (years) 0-5 6-11 12-17 18-24 25-34 35-44 45-54 55-64 65+ 2000 2005 2010 2015 2020 2025 Calendar year Source: CPRD Figure 8.2: Administrative prevalence of autism over time (annual, active 5- year, cumulative) Source: CPRD. Cumulative = ever recorded within the primary care record; Active 5- year = recorded within a five-year window; Annual (point) = coded within that year.

Identification has also changed by sex. Historically, autism diagnoses were concentrated among boys and among children with co-occurring intellectual disability. Although boys continue to be diagnosed more frequently than girls, the

gap has narrowed substantially. Between 2016 and 2024, autism-related SEND rose from 2.1% to 4.0% among boys and from 0.4% to 1.5% among girls, reducing the recorded ratio from 4.9:1 to 2.8:1[9]. This pattern is consistent with improved recognition of girls and women who were previously overlooked[286]. Figure 8.3 shows how recorded autism incidence varies by age and sex, illustrating changing patterns of identification across childhood and adolescence.

Figure 8.3 Recorded autism incidence by age band and sex Source: CPRD. Each panel has a different y-axis Much of the increase has occurred among people without recorded intellectual disability, while incidence among those with intellectual disability has remained comparatively stable. This pattern has also been reported internationally[287, 288], suggesting that recent growth reflects broader recognition among groups whose needs were previously less visible.

Figure 8.4 Recorded Autism Spectrum Disorder (ASD) incidence by intellectual disability (IDD) status – children and young people, 2000–2025

Source: CPRD. Common denominator (total at-risk person-years): the two lines sum to overall ASD incidence. Each panel has a different y-axis.

Figure 8.5 Recorded Autism Spectrum Disorder (ASD) incidence by IDD status – adults, 2000–2025

Source: CPRD. Common denominator (total at-risk person-years): the two lines sum to overall ASD incidence. Each panel has a different y-axis.

The growth in identification has also been uneven across childhood. The largest increases have occurred around the transition to secondary school, creating a pronounced peak in recorded prevalence in Year 7 that was not evident a decade

ago. Because younger cohorts now have higher rates of identification, these increases are likely to continue moving through the education system over the coming years. Planning for a growing and increasingly diverse population of autistic pupils is therefore a practical necessity[9].

8.2.2 Adults

Over the past two decades there has been growing recognition that many adults reached adulthood without their autism being identified. Despite this, administrative prevalence remains much lower among older adults than among children and young people. In 2025, recorded primary care prevalence ranged from around 2.5% among 18–24-year-olds to 0.07% among adults aged over 65[58]. This gradient is more plausibly explained by historic under-recognition, changing diagnostic concepts, patterns of help-seeking, access to assessment and recording practices than by true age differences in occurrence.

Earlier surveys estimated adult autism prevalence at around 1% using structured assessment[289]. More recent surveys show increasing reporting of autism-related traits, particularly among younger adults. Trait measures capture autism-related characteristics in the population but are not themselves diagnostic. Elevated scores may therefore reflect a range of experiences, including anxiety, psychological distress and social difficulties, as well as autism itself.

Importantly, later recognition of clinically significant impairment should not be confused with later onset of autism. Figures 8.6 and 8.7 show the distribution of adults screening positive for autistic traits by sex and age group. For some people, longstanding characteristics were accommodated by intelligence, family support, predictable primary school environments or other supports until social, educational or occupational demands increased.

Figure 8.6 Adults (16+) screening positive for autistic traits on a 17-item version of the autism quotient (AQ score of 8 or more) by sex

Source: APMS Figure 8.7 Adults (16+) screening positive for autistic traits on a 17-item version of the Autism Quotient (AQ score ≥8), by age group Source: APMS Some administrative datasets also show particularly rapid increases in autism diagnoses after around 2020. Clinician roundtables suggested that the COVID-19 pandemic may have contributed to increasing adult referrals, although further investigation is needed. Periods of isolation, poorer mental health, greater opportunity for self-reflection and the rapid expansion of online information about autism were all described as encouraging more people to consider whether autism might explain their experiences.

Whether there has also been an increase in the underlying distribution of autistic traits remains an open scientific question, although rapid changes in genetic liability or recognised risk factors appear unlikely. The most plausible explanations are improved recognition, broader diagnostic concepts, increasing identification of previously overlooked groups, stronger overlap with ADHD and common mental health conditions, and changing patterns of help-seeking[13, 44, 74, 290, 291].

8.2.3 Self-identification and recognition

In the 2025 OxWell survey, 11.4% of participating secondary and further education students reported being autistic. These included students reporting a professional diagnosis, being on an assessment waiting list, self-identification or identification by others. Of the full analytical sample, 5.5% reported a professional autism diagnosis[14]. Similarly, APMS data suggest that around 3.4% of adults identify themselves as autistic, compared with 1.2% who report a professional diagnosis, with many others waiting for assessment.

Self-identification is becoming increasingly common and remains distinct from clinical diagnosis, although emerging evidence suggests that self-identified adults often occupy an intermediate position between diagnosed autistic and non-autistic groups on measures of autistic traits, mental health and polygenic liability[292-294]. Self-identification should therefore neither be dismissed nor treated as equivalent to diagnosis.

Figure 8.8 Trends in self-reported autism by age group Source: GP Patient Survey. Shaded areas show 95% confidence intervals.

During long waits for assessment, self-identification may provide a framework for understanding experiences and accessing peer support. However, it may also strengthen expectations before assessment, making a non-diagnosis more difficult to receive.

“I think one of the biggest challenges is seeing people who have been

  • waiting a very long time for an assessment often their own

understanding has sort of solidified over that time in terms of what they

  • think is going on if you then reach a conclusion that they don’t have

autism, it can be devastating for a patient.”

Clinician Roundtable Polling commissioned for the Review (Figure 5.3, Chapter 5) found that around 18% of 16–29-year-olds self-identified as autistic, compared with 9% of those aged 30–45 years and 1% of those aged 55–99 years. Just over half of those aged 16–29 years, and just under half of those aged 30–54 years, who self-identified as autistic also reported having received a professional diagnosis.

These findings suggest that social and cultural trends in the population may be moving ahead of what is currently captured by administrative data sources such as CPRD. It is difficult to anticipate how neurodiversity as an identity, self-identification

and professional diagnosis will interact in future, although if current trends continue it is likely that identification with neurodiversity will increase.

8.2.4 What explains rising identification?

The evidence suggests that rising autism identification is best understood as the product of multiple interacting processes. Improved recognition, broader diagnostic concepts, reduced stigma, increasing identification of historically under-recognised groups, changing help-seeking behaviour, greater public awareness and systems that continue to link diagnosis closely to access to support, all appear to contribute.

One plausible explanation is that relatively stable neurodevelopmental vulnerabilities interact with increasing anxiety, distress and environmental demands, making autistic characteristics more visible, more impairing and more likely to lead to referral. This does not imply that autism begins in adolescence or is caused by distress. Rather, pre-existing characteristics may become clinically significant when social, educational or occupational demands exceed available supports.

Figure 8.9. Co-occurring positive screens for ADHD, autistic traits and substance dependence among young adults (16–24) with CMD (common mental disorders)

Source: APMS. 95% confidence intervals (wide on small bases)

8.3 Recognition, diagnosis and changing pathways to support

People reach assessment through parents, schools, primary care, mental health services, self-identification and wider public awareness. Pathways vary according to age, sex, co-occurring conditions, access to information and local service organisation. Recognition and diagnosis are therefore shaped not only by individual characteristics, but also by the systems through which people seek help[271, 295].

Schools play a central role in recognising autism during childhood because they observe communication, interaction, regulation, learning and participation. Some pupils cope well academically at school while struggling at home or in less structured settings, delaying recognition and support.

Many adults seek assessment after years of unexplained difficulties, often following contact with mental health services, exposure to information about autism, recognition of similar characteristics in a family member, or increasing awareness of autism across the life course.

Demand for assessment now far exceeds capacity. By December 2025, more than 254,000 people had an open autism referral; over 90% had waited beyond 13 weeks and only 3.6% had received a first appointment within the expected period[12]. Waiting lists were around thirteen times larger than in 2019, reflecting both improved data coverage and increasing recognition and help-seeking[296].

As a result, many people spend long periods between recognising possible autism and receiving a formal assessment, often with little or no support. Only around 55% of clinicians reported providing routine waiting-list support; 40% reported no ongoing post-diagnostic support, and only 12.6% described support extending beyond a single session[13]. Diagnosis too often marks the end, rather than the beginning, of a support pathway.

“There seems to be a heavy focus at the moment going on around the country where even when services have had post diagnostic support previously and they’ve had occupational therapists employed for that, now people are saying, well the waiting list is so large, we can only do

  • the diagnosis where actually they’re really important for people’s lives

and quality of life.”

Clinician Roundtable This was echoed by lived-experience partners: “When I got my diagnosis, I didn’t have any offer of any follow-up or anything. I think they only really gave me maybe one or two links.”

Ambitious About Autism focus group Diagnosis is widely perceived as a gateway to educational, workplace and service support, although reasonable adjustments under the Equality Act and Personal Independence Payment depend on functional impairment rather than diagnosis. In practice, however, a diagnosis often remains the most readily accepted evidence of need, particularly in pressured school environments.

“Before my diagnosis, I got zero support, despite it being very well known that I was on the pathway.”

Ambitious About Autism focus group Independent and Right to Choose providers completed more than one-third of NHS- funded autism assessments in 2025–26. Responding providers reported around 90% remote delivery and diagnostic conversion rates ranging from 66% to 93%, with averages of 79–86%[35]. These figures require careful interpretation, as comparable NHS data on case mix and outcomes are not routinely available.

The Review’s clinician roundtables and provider survey identified variation across NHS and independent pathways in assessment length, multidisciplinary input, formulation, consideration of co-occurring conditions and interpretation of impairment. These findings support the need for common, transparent quality standards across all providers, rather than assumptions about either sector[13].

Right to Choose has expanded partly because NHS capacity is constrained. In some areas, closed NHS waiting lists have made it the only practical route to a publicly funded assessment. Geographic variation also affects waiting times, assessment models, observation, differential diagnosis and access to post-diagnostic support[297-

299]. Common standards and comparable reporting are therefore essential.

Waiting should become an active period of support rather than a passive delay. Depending on need, this may include reasonable adjustments, psychoeducation, family support, communication and sensory adaptations, treatment of co-occurring mental health conditions, and clear routes for escalation if functioning or risk deteriorates.

At a minimum, diagnosis should be followed by an accessible explanation of the assessment, a practical profile of strengths and needs, identification of co-occurring mental and physical health conditions, advice on reasonable adjustments, and clear

routes back into care as needs change. A diagnostic report and signposting alone do not constitute an adequate post-diagnostic offer.

8.4 Heterogeneity, broadening and functional impact

8.4.1 Heterogeneity across communication, cognition, behaviour and functioning

Autism is best understood as a highly heterogeneous spectrum rather than a single, uniform condition. Autistic people differ substantially in their communication, intellectual functioning, language, adaptive skills, emotional regulation, sensory experiences and support needs. This diversity has become increasingly apparent as the diagnostic concept has broadened.

An autism diagnosis alone is therefore not a reliable guide to functioning. Two people with the same diagnosis may have very different strengths, challenges and day-to-day experiences.

8.4.2 Broadening of the autism construct over time

Diagnostic systems have broadened from viewing autism as a relatively rare condition associated with marked early difficulties to recognising a spectrum that includes people without intellectual disability and those whose differences are less immediately apparent. Current criteria recognise sensory differences, heterogeneity and difficulties that may emerge when environmental demands exceed an individual’s coping capacity, improving recognition of girls, women and adults[271].

The diagnosed autistic population is therefore substantially more heterogeneous than it was two or three decades ago. As Figures 8.4 and 8.5 illustrate, recorded incidence has increased predominantly among people without co-occurring intellectual disability.

ECHILD data show that pupils identified as autistic increasingly resembled pupils without SEND on Reception developmental measures between 2017/18 and 2021/22. This reflects broader recognition rather than an absence of clinically significant need[9].

Figure 8.10. Good level of development at the start of schooling by SEND group

Source: ECHILD (Education and Child Health Insights from Linked Data). Good level of development refers to assessment at the end of Reception (age 5) and is linked retrospectively to all pupils observed in each academic year.

8.4.3 Functional impact and variability across contexts

The functional impact of autism varies considerably and is strongly influenced by interaction with the environment. For example, an individual may experience significant difficulties only in unstructured or unpredictable contexts.

As one clinician participating in the Review observed: “An autistic trait for someone in a certain life situation may be very impairing. And if they weren’t in that life situation without those demands being placed upon them, they wouldn’t be impaired. And I think this makes it very difficult to standardise the assessment of impairment.”

Clinician Roundtable This distinction is central. Although many autistic people experience substantial difficulties in everyday living, healthcare access, employment and community participation[65, 300], these outcomes are neither universal nor inevitable. Many participate successfully in education, employment and relationships when appropriate support and accommodating environments are available.

Support should therefore be organised around functioning and participation, rather than diagnosis alone.

8.4.4 Diagnosis and identity: lived experience and meaning

For many people, an autism diagnosis serves purposes that extend well beyond clinical classification. Adults frequently seek diagnosis to understand longstanding experiences, make sense of earlier difficulties, obtain validation and develop a more coherent understanding of themselves[301]. This greater understanding may also help people identify strategies that support wellbeing and participation. For some, it reduces self-blame and supports self-acceptance. A growing body of evidence suggests that developing a positive autistic identity may also be associated with improved mental health and reduced social anxiety[302].

“If I hadn’t been diagnosed, I would have gone my whole life thinking there was something wrong with me.”

Ambitious About Autism Focus Group At the same time, diagnosis may evoke more complex reactions. Some people experience grief or frustration about years spent without recognition or support. Others feel uncertain about how diagnosis fits with their identity or future aspirations[303, 304].

The benefits of diagnosis often depend less on the diagnostic label itself than on what follows from it. Diagnosis is most likely to improve wellbeing when it leads to greater understanding, appropriate support, reasonable adjustments and meaningful changes in everyday life[299, 305]. Diagnosis is therefore neither uniformly positive nor uniformly negative. Its meaning, and its benefits, depend greatly on context, expectations and, above all, on the support that follows.

8.5 Social patterning and inequalities

Patterns of autism identification are not evenly distributed across the population. Recorded prevalence and diagnosis vary by sex, age, socioeconomic circumstances, ethnicity and intellectual disability.

8.5.1 Sex and gender

One of the most important changes in autism identification over recent decades has been the increasing recognition of autism among girls and women. Historically, autism was diagnosed much more frequently in males than females, contributing to the widespread perception that it was predominantly a male condition.

Active five-year recorded prevalence increased from 0.01% in 2000 to 0.95% in 2025 among females, and from 0.06% to 1.44% among males. Over the same period, autism incidence rose from 3.7 to 240.7 per 100,000 person-years among females and from 16.3 to 252.3 per 100,000 person-years among males. The much faster relative increase among females has therefore substantially narrowed the historical sex difference in recorded autism identification (Figure 8.11). From adolescence onwards, recorded incidence among females now exceeds that among males.

CPRD and School Census data record sex, rather than gender. Gender-diverse identities appear to be more common among autistic people but are not captured in these datasets[183].

Figure 8.11 Recorded incidence of autism by sex

Source: CPRD Growing evidence suggests that autism may present differently in many girls and women than in boys and men. Greater awareness of masking and less stereotypical presentations has improved recognition, and the narrowing male-to-female ratio most plausibly reflects improved identification of previously under-recognised groups rather than a recent change in underlying prevalence[286].

8.5.2 Socioeconomic patterning

Autism diagnosis is also socially patterned. CPRD analyses demonstrate a persistent socioeconomic gradient, with both recorded incidence and prevalence higher in more deprived communities. In 2025, administrative prevalence reached approximately 1.5% in the most deprived quintile, compared with 1.0% in the least deprived. The deprivation gradient strengthened during the 2000s before stabilising at around 1.5:1 between the most and least deprived groups (Figure 8.12).

Figure 8.12 Deprivation gradient: Q5:Q1 ratio (recorded incidence & prevalence)

Source: CPRD Figure 8.13 Recorded incidence of autism by socioeconomic (IMD) quintile Source: CPRD Figure 8.14 Recorded prevalence of autism by socioeconomic (IMD) quintile Source: CPRD The gradient is particularly marked in early childhood, where autism incidence among preschool children in the most deprived communities is approximately twice that observed in the least deprived areas. These findings should be interpreted cautiously, as administrative gradients reflect differences in recognition and recording as well as any underlying differences in occurrence.

Several explanations may contribute. Autism is strongly heritable, and unsupported autistic adults experiencing downward socioeconomic mobility may be more likely to have autistic children growing up in more deprived communities[290]. Pathways to identification may also differ. Children in more deprived areas may have greater contact with schools and services that recognise and record need, while privately funded diagnoses, which may be more common in more affluent groups, are incompletely captured in administrative datasets. At the same time, early adversity and disrupted early relationships may give rise to developmental presentations that overlap with autistic characteristics. Further research is needed to disentangle these possible explanations.

Figure 8.15 Autism incidence deprivation gradient (Q5:Q1) by age and sex Source: CPRD. Each panel has a different y-axis.

The evidence therefore suggests that socioeconomic inequalities influence not only outcomes but also the pathways through which autism is recognised and recorded.

8.5.3 Ethnicity and access to identification

CPRD analyses undertaken for this Review demonstrate substantial variation in administrative prevalence across ethnic groups. In 2025, recorded prevalence ranged from approximately 2.2% among people of mixed ethnicity and 1.9% among White groups to 1.1% among Black groups and 0.6% among South Asian groups. Similar patterns were observed for incidence (Figures 8.16 and 8.17).

Figure 8.16 Recorded autism prevalence by ethnic group and age Source: CPRD. Recorded rates by self-/clinician-assigned ethnic group (CPRD 5- group classification. Denominators are scaled to the registered ethnicity-coded population. Patients with no recorded ethnicity (‘Not recorded’) are excluded from group comparisons. Each panel has a different y-axis.

Figure 8.17 Autism recorded incidence by ethnic group and age Source: CPRD. Recorded rates by self-/clinician-assigned ethnic group (CPRD 5- group classification). Denominators are scaled to the registered ethnicity-coded population. Patients with no recorded ethnicity (‘Not recorded’) are excluded from group comparisons. Each panel has a different y-axis.

Access to assessment, help-seeking, cultural understandings of autism, language barriers, local recording practices and, potentially, stereotyping or bias within services may all contribute to these patterns of identification[306, 307].

Educational data also indicate that some ethnic minority groups are less likely to receive specific diagnostic labels despite identified SEND needs, suggesting that inequalities arise in recognition and classification as well as in access to services[308].

Overall, the evidence points more strongly towards inequalities in identification than to major differences in underlying autism prevalence.

8.5.4 Participation, SEND and longer-term outcomes

Across multiple datasets, autistic people experience poorer outcomes in education, employment and wider participation than the general population. These outcomes are strongly influenced by the severity of disability, co-occurring conditions, environmental barriers, access to appropriate support and wider social disadvantage.

Employment provides a clear example. Recent reviews estimate that around 37% of autistic adults are in paid employment[309]. UK national data suggest somewhat lower figures, and many autistic people remain under-employed or work without appropriate support[310, 311]. Where employment proves difficult to obtain or sustain, the evidence most consistently points to inaccessible systems – including a lack of reasonable adjustments, inadequate management support and poorly matched roles – rather than to autism itself.

It is equally important to recognise the diversity of outcomes within the autistic population. Some autistic people, particularly those with early-emerging autism and high support needs, experience lifelong difficulties affecting communication, learning, adaptive functioning, self-care and health. For this group, substantial support is often required throughout adulthood, and increasing demand may have reduced the availability of these resources (see Section 8.9).

Educational data provide a similar perspective. Analysis of 3.9 million pupils in England found that 22% of pupils with SEND became NEET (not in education, employment or training), compared with 7% of pupils without SEND[308]. Once measures such as absence, exclusion, social care contact and socioeconomic disadvantage are adjusted for the risk of being NEET falls from four times more likely to 1.7 times more likely for people with SEND. Several of these measures are themselves consequences of how the system responds to SEND, so the adjusted figure is a conservative estimate of the association, rather than a measure of what SEND alone contributes. The relationship between SEND, participation and transition to adulthood is considered further in Chapter 16.

These findings reinforce one of the Review’s central themes. The greatest inequalities associated with autism relate not simply to diagnosis, but to participation. A fair system therefore requires not only equitable recognition of autism, but also equitable access to the support and adjustments that enable participation throughout the life course.

8.6 Co-occurrence and layered needs

8.6.1 Co-occurrence as the norm rather than the exception

One of the most important developments in understanding autism has been the recognition that it rarely presents in isolation. Co-occurring neurodevelopmental, mental health and physical health conditions are common and often play a major role in shaping need, impairment and service use. Many autistic people experience combinations of neurodevelopmental, emotional, behavioural and physical health difficulties that cannot be understood through autism alone. Co-occurrence is considered in more detail in Chapter 9.

  • “ the complexity of the people we’re seeing is quite phenomenal compared to

what it would be like 6 years ago… they’ve often got lots of mental health issues as well.”

Clinician Roundtable People in the secure estate show particularly high levels of co-occurring autism, ADHD, language impairment, trauma, disrupted education and care experience[194, 253, 312-314] (Chapter 17). This pattern is not unique to secure settings. Where autism is associated with the greatest impairment and disadvantage, this frequently reflects the accumulation of co-occurring conditions and adversity rather than autism considered in isolation.

8.6.2 Autism, ADHD and neurodevelopmental overlap

Recognition of the substantial overlap between autism and ADHD has been one of the most significant developments in recent years. Under DSM-IV the two diagnoses could not be made together. DSM-5 recognised that they frequently co-occur, reflecting growing evidence that many people experience characteristics of both conditions.

Between 2016 and 2024, the proportion of autistic pupils also recorded with social, emotional and mental health needs rose from 9.2% to 14.3%, while the proportion with speech, language and communication needs increased from 12.5% to 16.8%. Throughout this period, almost half of autistic pupils had at least one additional need recorded alongside autism[9]. These trends reflect changes in recognition and recording as well as increasing complexity.

8.6.3 Autism, mental health, distress and suicidality

Anxiety, depression, eating disorders, self-harm and suicidality are substantially more common among autistic people than in the general population[79, 278].

UK Biobank analyses found elevated rates of depression, anxiety and social anxiety among both diagnosed and self-identified autistic adults, with the latter occupying an intermediate position. Autism-related difficulties and broader psychological distress frequently overlap.

Diagnostic overshadowing remains an important concern. Not everyone seeking an autism assessment is autistic; social anxiety, trauma and other conditions can resemble autism. Figure 8.18 shows a strengthening association between elevated autistic traits and both self-harm and suicide attempts. Changes in this association may reflect changing case mix, mental health overlap, self-perception, diagnostic substitution, help-seeking or reporting, rather than a strengthening causal relationship between autistic traits and these outcomes[81, 315].

Figure 8.18 Correlation between screening positive for autistic traits (AQ score ≥8) and non-suicidal self-harm and suicide attempt among adults aged 16+

Source: APMS Sleep disturbance, epilepsy, gastrointestinal conditions, chronic pain, joint hypermobility and multimorbidity contribute to disability and healthcare use[277, 280, 316]. Physical or mental illness may be misattributed to autism, while autism, particularly in women, may be missed beneath other diagnoses. Mental health difficulties reflect the interaction between autistic characteristics, exclusion, inaccessible environments, masking and access to care. Earlier recognition, educational adjustment and matched support may prevent some distress from becoming established.

NHS England inpatient data show substantial mental health crisis among autistic children and adults, including self-harm, eating disorders and suicidality, increasingly among people without intellectual disability. This broadens, rather than diminishes, the need for specialist intellectual disability services. Services therefore require earlier autism-informed community and mental health support, including for eating disorders. Where there is active anxiety, depression, eating disorder, self-harm, trauma-related distress or psychosis, appropriate mental-health assessment and treatment should not be delayed until autism assessment is complete. Care may need to be autism-informed and adapted, but it should not be withheld while diagnostic clarification is pending.

“My daughter was diagnosed as autistic just before her 15th birthday.

Before then, school became increasingly difficult and draining due to her learning difficulties. She used to hide and cry in lower school and in middle school she started to be bullied for being different. From middle school age, she started to not leave the house apart from going to school. When she got home to her safe space, she used to unleash all her anger on us. Her anxiety became through the roof, she stopped sleeping properly, and developed rigid patterns around food and routines such as having to make sure the door was locked a certain amount of times.

No one mentioned autism until she was at upper school, and then no adjustments were made until she had a diagnosis.

When she was 12, our daughter was put on ADHD medication, which lowered her mood, leading her to self-harm. She was referred to CAMHS who turned her down. Within the next month we ended up in A&E due to taking an overdose. She was then finally accepted by CAMHS, who suggested an autism assessment.

2 years later, after receiving an autism diagnosis, little changed. School would not make any adjustments until an EHCP (Education, Health and Care Plan) was accepted, and social care said she did not meet their threshold for support because she did not have a learning disability.

Our daughter was broken and unsupported and was admitted to inpatients at 15 years old.”

Lived Experience Partner Adult services need both lifelong specialist capacity and autism-capable mental health provision. Where inpatient care is necessary, it should meet NHS England Culture of Care standards and avoid distant placement where possible.

8.6.4 Why layered needs challenge diagnosis-led systems

Autism pathways increasingly need to respond to layered and overlapping needs. Single-condition pathways can fragment care across autism, ADHD, mental health, education and primary care. More integrated assessment and support have the potential to reduce repeated referrals, inappropriate intervention and unnecessary delay.

A high-quality pathway should do more than determine whether the criteria for autism are met. It should provide a coherent formulation of the factors contributing to distress and impairment, identify differential diagnoses and co-occurring conditions, and specify the treatment, adjustments and support required, regardless of whether autism is ultimately diagnosed.

8.7 Assessment practice and diagnostic uncertainty

There is no biological test for autism. Diagnosis is therefore based on observable behaviour and reported experience, interpreted in relation to established diagnostic criteria. Autism assessment remains a skilled clinical process requiring developmental history, direct observation, self-report, careful assessment of current functioning, and explicit consideration of both differential diagnoses and co-occurring conditions. This is consistent with NICE guidance, which recommends comprehensive assessment that considers coexisting conditions rather than treating autism as an isolated presentation[317]. Standardised assessment tools are valuable, but they support rather than replace multidisciplinary clinical judgement[271].

Several mental health conditions, including social anxiety disorder, obsessive- compulsive disorder, anorexia nervosa and trauma-related difficulties, may present with features that overlap with autism. Careful differential diagnosis is therefore essential and requires multidisciplinary teams with expertise in both autism and mental health.

Evidence from adults referred for assessment suggests that many people seek diagnosis not only to access services but also to gain understanding, validation and a coherent account of longstanding experiences. Diagnostic pathways should therefore include clear discussion of expectations and provide a formulation and summary of needs regardless of whether an autism diagnosis is ultimately made. This helps ensure that people whose assessment concludes ‘not autism’ are nevertheless offered an explanation and an appropriate route to support.

The threshold for receiving practical support should not be identical to the threshold for an autism diagnosis. People may have communication, attention, emotional or participation needs that warrant support without meeting full diagnostic criteria. Conversely, diagnosis alone does not determine the type or intensity of support required.

National surveys indicate that many services struggle to implement NICE guidance consistently. Teams report substantial increases in referral volume, greater clinical complexity and limited multidisciplinary capacity. These pressures contribute to variation in assessment models, approaches to differential diagnosis, diagnostic conversion rates, and the support available while people are waiting or following diagnosis. Clinicians also describe pressure to diagnose in order to enable access to support, alongside concerns that, in some settings, an autism diagnosis may overshadow co-occurring mental health conditions and impede access to appropriate treatment.

Diagnostic uncertainty has therefore become an increasingly common and legitimate feature of clinical practice. Support needs vary in ways that diagnostic criteria do not fully capture, and referrals often involve multiple overlapping conditions. In addition, infant studies suggest that developmental trajectories are not fixed, and elevated autistic characteristics may emerge at different stages of childhood development[318]. Diagnosis-led systems manage these uncertainties poorly.

The evidence therefore supports minimum national standards for autism assessment. These should include explicit consideration of clinically significant impairment, systematic assessment of co-occurring conditions, clearer guidance on when remote and in-person assessment are appropriate, and more consistent recording of what was assessed, the conclusions reached and the support recommended. Better assessment of functioning is likely to improve both diagnostic transparency and care planning[319].

8.8 Masking, impairment and environment

8.8.1 What is masking?

Masking, or camouflaging, describes strategies used to conceal or compensate for autistic characteristics, including rehearsing social interactions, suppressing behaviours, copying others or relying on routines.

Masking has become an increasingly prominent concept in both autism research and clinical practice. It is reported particularly frequently by girls and women, by people diagnosed later in life, and by those whose autistic characteristics are less immediately apparent to others[320, 321]. Many autistic people describe masking as an attempt to avoid stigma, reduce social exclusion or meet expectations in educational, occupational and social environments.

8.8.2 Why masking matters for recognition and diagnosis

Masking may reduce the visibility of autism to parents, teachers and clinicians, concealing the effort, distress and exhaustion that may underlie outwardly successful performance.

“A lifetime of masking, led to a lack of identity, and ultimately an identity crisis. I had no idea who I was, what I actually liked, who my real friends were – it was terrifying and extremely distressing.”

Lived Experience Partner Such compensatory social strategies may contribute to delayed recognition, particularly among autistic girls and women[79, 274]. At the same time, masking should not be regarded as a complete explanation for all diagnostic uncertainty. Similar compensatory strategies occur in anxiety and other conditions, while invisible impairment is also recognised in chronic pain, fatigue, anxiety and depression. People may appear to function well while sustaining considerable effort, exhaustion or distress. This may become more difficult during periods of increased demand, life transitions or hormonal change, contributing to autistic burnout. Support should not be withdrawn solely because someone appears to be coping[322].

The key implication is that impairment cannot always be inferred from outward appearance alone. This distinction is important because an autism diagnosis requires clinically significant impairment. Assessing impairment is therefore central to diagnosis, but is often more challenging than identifying autistic characteristics themselves. Although masking is a clinically useful concept, it should be interpreted

within the context of a broader developmental and diagnostic assessment[323]. Measures such as the CAT-Q should not be used diagnostically in isolation[323]. Assessment should integrate internal experience, observed functioning and information from people who know the individual across different settings[323].

The issue of masking was also highlighted by clinicians participating in the Review’s roundtables and especially in relation to women. Examples were provided of client complaints after not receiving a diagnosis, and there being a trend towards patients attributing this to their autistic features not being evident due to masking. Clinicians reported that this was particularly evident in patients who had already built a strong identity based on autism, who then felt let down if the diagnostic criteria were not met. Clinicians had a balanced discussion of masking as a deliberative, effortful process engaged in by some autistic people, but also considered that masking did not hide diagnostic features entirely during comprehensive diagnostic assessments often involving multiple clinicians and settings. There was discussion around some public discourse on this topic conflating typical social skills development with autistic masking:

“… there’s no question I’ve seen some people that have deliberately worked at camouflaging and masking strategies over years and it’s very effortful. And often when you do the ADOS [Autism Diagnostic Observation Schedule assessment], the mask sort of breaks down and… the impairment becomes more obvious. I think where I probably do have a slight question mark around it is the idea of unconscious masking and people will say, well, I’ve masked since childhood and now I just automatically do it. I’m not sure if that’s any different from normal social skills development that every child experiences”

Clinician Roundtable This reflects the broader challenge of autism assessment which requires clinicians to balance developmental history, observable behaviour, self-report and functional impact.

8.8.3 Environment, participation and functioning

Functioning emerges through the interaction between autistic characteristics and the environment, and should be considered across multiple areas of life.

In APMS, adults with elevated autistic traits reported greater difficulty with paperwork, managing money, household activities and transport, together with higher levels of loneliness, isolation and stress[13]. Functioning is therefore relational rather than a fixed property of the individual[47, 300]. Apparent coping may depend on

sustained effort, followed by later exhaustion or dysregulation. Figures 8.19 to 8.21 illustrate the wider social, functional and physical health differences reported by adults screening positive for autistic traits. Reasonable adjustments and supportive environments can reduce these difficulties, but they do not remove all disability. Some autistic people will continue to require substantial or intensive support even in highly accommodating environments. Family capacity also forms an important part of this context, particularly during childhood and where support needs are lifelong. The consequences of similar difficulties may vary according to parental health, caring responsibilities, housing, income, transport, access to community opportunities, and the availability of informal and statutory support.

Figure 8.19 Social functioning difficulties among adults (16+) by whether screening positive for autistic traits (AQ 8+)

Source: APMS. Each panel has a different y-axis.

Figure 8.20 Difficulties with activities of daily living among adults (16+) by whether screening positive for autistic traits (AQ 8+)

Source: APMS. Each panel has a different y-axis.

Figure 8.21 Physical conditions and predicted verbal IQ among adults (16+) by whether screening positive for autistic traits (AQ 8+)

Source: APMS. Each panel has a different y-axis.

8.9 Autism with severe or profound intellectual disability

This group comprises autistic people whose difficulties emerge early in development and who require substantial, often lifelong, support. Their needs commonly include severe or profound intellectual disability, very limited functional speech, marked adaptive difficulties, high levels of dependence in everyday living, complex physical health conditions and, for some, self-injurious behaviour or behaviour that challenges. Their support needs differ qualitatively from those of many other autistic people and require coordinated provision across health, education and social care.

This does not imply that autistic people without intellectual disability do not also require intensive support. Many experience severe mental health difficulties and other needs that may be less visible or less well captured by traditional indicators of disability. As a result, the absence of intellectual disability may itself become a barrier to accessing appropriate support. Systems therefore need to respond to need across the full range of autistic presentations.

8.9.1 What is known about prevalence

Estimating the size of this population is challenging because definitions vary between studies and services. Recent work has used the term profound autism to describe autistic people requiring very high levels of support, often defined by severe intellectual disability, minimal or absent functional language, or both, although the term itself remains contested[324]. A recent review estimated that between 11% and 48% of autistic people meet criteria for profound autism, depending on the definition used and the populations studied[325]. More recent surveillance studies generally lie towards the middle of this range, with US data suggesting that around one quarter of autistic eight-year-olds require very substantial lifelong support[324]. Although estimates vary, the evidence consistently indicates that a substantial minority of autistic people have enduring, high support needs.

England currently lacks the longitudinal data needed to determine whether the absolute number of autistic people with severe or profound intellectual disability has remained stable or increased over time. Indicators such as the sustained growth in demand for special school places are consistent with either a stable or a growing population and therefore cannot resolve this question. A consistent administrative marker for autism with severe or profound intellectual disability, or more broadly for very high support needs, would substantially improve planning and monitoring.

Because much of the recent increase in autism identification has occurred among people without intellectual disability, the proportion of autistic people with severe or

profound intellectual disability is likely to have declined, even if their absolute numbers have remained stable or increased more slowly. In England’s school records, the proportion of autistic pupils with recorded co-occurring intellectual disability fell from 15.1% in 2016 to 8.7% in 2024, although this partly reflects under- identification of intellectual disability in administrative data[9]. There is therefore a risk that this group becomes less visible within the wider autistic population just as recognition broadens elsewhere.

Families consistently report that recognition and support fall well short of the scale of need, including the challenges associated with lifelong dependency, co-occurring conditions and caring responsibilities. Planning therefore needs not only to protect existing specialist provision but also to expand it (see Chapter 20), particularly in anticipation of the point at which ageing family carers are no longer able to continue providing care.

8.9.2 Lifelong support needs

For many autistic people with early-emerging, high support needs, support extends across the lifespan. Difficulties commonly affect communication, learning, adaptive functioning, self-care, mobility, social participation and health. Co-occurring epilepsy, sleep disorders, gastrointestinal conditions, sensory differences and other medical needs are common and contribute substantially to disability and service use[277, 280]. Behaviours that challenge, including aggression, self-harm and stereotyped behaviours, are also common, affecting around 10–30% of people, depending on the level of intellectual disability.

Many people require support with activities of daily living throughout adulthood and may continue to depend on family carers or residential care. This places substantial emotional, practical and financial demands on families[325].

These needs create distinctive challenges for services. Effective support requires sustained multidisciplinary input across health, education and social care, together with careful planning at major life transitions. Particular attention is required when family carers become unable to continue providing care because of age, illness or other circumstances. Continuity of care is critical because interruptions in support may have profound consequences for both individuals and their families.

“My son is now 22 and has a diagnosis of what used to be called severe autism but is now usually referred to as profound autism.

He had a normal birth and delivery and was developing well but his health regressed during his second year of life. He suffered from eczema, chronic constipation, insomnia, acute anxiety, odd behaviours

and outbursts of aggression. His development slowed down and then went off a cliff.

He was diagnosed with severe autism when he was four and a half (in 2008).

At that time, autism was still relatively rare. He was one of just two boys with the condition in his mainstream primary school of 300 kids.

Even then, we struggled to get our son the support he needed from the school and local authority; the expectation was to stay in mainstream with extra support. He attended two mainstream and then three special schools (one private and two state-provided) where he made very little progress. He left the system at the age of 19. He has very limited language, can’t read or write, and is unable to manage any aspects of personal care. Since leaving school and following assessment by adult social care, he is now supported by a care company. He lives with us and goes out for walks with a carer five days a week.

In the two decades since his diagnosis, the situation has gotten dramatically worse. There are more children like my son and those with related conditions like ADHD and OCD, anxiety and depression. It feels like a crisis. The community support groups are awash with desperate calls for help. The special schools are overflowing – even in Cornwall where we are living now.”

Lived Experience Partner For people with lifelong high support needs, successful support should not be measured only by recovery, independence or entry into employment. Maintaining communication and daily functioning, preventing deterioration, protecting health and safety, sustaining family care and improving quality of life are important outcomes.

8.10 Evidence gaps and uncertainty

One of the most important unresolved questions concerns the relationship between underlying prevalence and changing rates of identification. Administrative prevalence has increased substantially, and self-identification appears to be increasing among both young people and adults. It remains uncertain, however, whether the underlying prevalence of autistic traits, or of clinically significant autism, has changed over time.

England lacks repeated population-based prevalence studies using consistent methods across the life course. Current discussions therefore rely on comparisons between administrative data, surveys and prevalence studies that address different questions and are not directly comparable. International data show similar increases in diagnosis but likewise lack repeated population-based studies capable of determining whether underlying prevalence has changed[326]. Differences in culture, language, recognition, diagnostic definitions, resources and service provision mean that international comparisons should be interpreted with caution.

The proportion of adults screening positive on the AQ-17 (APMS) increased from 10.9% in 2007 to 20.4% in 2024, with the largest increase among younger adults[13]. Before firm conclusions can be drawn about whether this reflects a genuine increase in autistic characteristics within the population, these findings require replication, assessment of measurement stability and extension to children and young people.

8.10.2 Changing case-mix

The evidence strongly suggests that the profile of people receiving an autism diagnosis has changed, with proportional increases among girls and women, adults, and people without intellectual disability. It remains unclear, however, how much of this reflects improved recognition of previously overlooked groups, broader diagnostic concepts, changing help-seeking behaviour, or the possibility that autism may genuinely be more common than previously recognised. These explanations are not mutually exclusive and may all contribute.

There is also uncertainty about the contribution of diagnostic substitution to current trends. The decline in the recording of general and moderate learning difficulties is consistent with such a process, but current datasets do not permit person-level analysis of diagnostic transitions over time and therefore cannot establish its extent[9,

327].

8.10.3 Co-occurrence and complexity

Considerable uncertainty remains about how patterns of co-occurring conditions may be changing over time, and how this contributes to rising demand. The relative contributions of changing autism prevalence, increasing co-occurrence, rising psychological distress and changing help-seeking remain uncertain.

8.10.4 Functioning and impairment

A key message of this chapter is that diagnosis is distinct from functioning and support needs. Yet functioning remains one of the least consistently measured

aspects of autism, particularly across education and social care. This makes it difficult to evaluate the effects of diagnosis or support.

Diagnostic criteria also require clinically significant impairment, but there remains considerable variation between clinicians in how impairment is understood and applied in practice. More consistent assessment of impairment is therefore warranted, given its importance in determining support needs.

8.10.5 Outcomes following diagnosis

Although many people report that diagnosis provides understanding, validation and access to support, relatively little is known about longer-term outcomes following contemporary autism diagnosis. Longitudinal evidence suggests that early presentation is often an unreliable predictor of later outcomes, except among those with the most profound difficulties. There is also evidence that some mental health problems common in early adulthood become less frequent later in life, with some older autistic adults reporting greater life satisfaction than when they were younger[328]. Nevertheless, the overall evidence base remains limited.

Relatively little is known about which groups benefit most from diagnosis, which forms of post-diagnostic support are most effective, how outcomes differ between those who do and do not receive a diagnosis, or the experiences of people who undergo assessment but are not diagnosed.

A related uncertainty concerns the stability of support needs across the life course. Autism is generally lifelong, but associated impairment is not necessarily constant. As with ADHD, support needs may fluctuate according to life stage, environmental demands and major transitions.

8.10.6 Under-represented populations and missing evidence

Important gaps remain for several groups whose experiences are under-represented in current research. These include: autistic people with early-emerging high support needs; those with severe or profound intellectual disability; minimally verbal autistic people; older autistic adults; people from minority ethnic communities; and those experiencing multiple forms of social disadvantage. As a result, understanding remains limited regarding how autism interacts with poverty, trauma, racism, gender diversity and different cultural contexts.

8.10.7 Summary

Despite these uncertainties, the evidence reviewed is sufficient to conclude that autism diagnosis and identification have increased substantially, that the diagnosed

population has become more heterogeneous, and that co-occurrence, functioning and environment are central to understanding need. The evidence supports a cautious interpretation: simple explanations are not supported, and the most plausible account remains that several interacting processes are operating simultaneously.

8.11 Synthesis: why autism challenges diagnosis-led systems

Autism illustrates the limitations of systems that rely on diagnosis as the principal route to support. Diagnosis captures only part of need, functioning and participation.

Autism is dimensional, heterogeneous, context-sensitive and commonly co-occurs with ADHD, mental and physical health conditions, and intellectual disability. Diagnostic thresholds are clinical judgements, and people with the same diagnosis may differ greatly in their functioning and support needs.

Diagnosis can provide understanding and validation, but it has also become the principal gateway to educational, workplace, social care and other forms of support.

The choice is therefore not between diagnosis and needs-led support. A needs-led approach without timely diagnosis risks becoming generic care or implicit rationing, while a diagnosis-led approach without appropriate support results in long waits, fragmented pathways and diagnostic labels without meaningful help. The evidence supports timely diagnosis alongside comprehensive formulation, earlier support, treatment of co-occurring conditions and protected specialist provision for people with lifelong high support needs.

Increasing demand reflects broader recognition, co-occurrence, changing expectations, greater public awareness, increasing complexity and the close link between diagnosis and access to support.

Diagnosis therefore remains important, but it is not sufficient. Its value lies not only in the understanding it provides, but also in the support it can unlock. Rising help- seeking is therefore both a service challenge and an opportunity for a fairer system to respond more effectively. Systems should combine diagnosis with assessment of co-occurring conditions, functioning, participation and support needs in order to provide timely, proportionate and equitable support.

Chapter 9. Co-occurrence and complexity

Neurodevelopmental, mental health and physical health conditions frequently occur together (‘co-occurrence’). Between 50% and 75% of people with autism or ADHD meet criteria for at least one common mental disorder, and between 30% and 40% meet criteria for two or more. International evidence suggests that around 25% of autistic people also have ADHD, and between 10% and 20% of people with ADHD have autism; UK survey data found that 43% of autistic children met criteria for ADHD and 17% of children with ADHD met criteria for autism.

Co-occurrence has become more common, particularly among young people and young adults. Between 2007 and 2024, the proportion of 16–24-year-olds with a common mental disorder screening positive for ADHD traits increased from 36% to 57%, while the proportion screening positive for autistic traits increased from 20% to 61%.

Complexity – the interaction between multiple conditions and adversities – is increasing and is strongly associated with poorer outcomes. Outcomes worsen as conditions accumulate and interact, affecting education, employment, daily functioning, quality of life, healthcare use, self-harm, unmet need and mortality. A relatively small group with multiple overlapping difficulties – around 4–5% of the population – experience the greatest adversity and support needs.

Current systems often struggle to recognise and respond to this complexity. Diagnostic overshadowing, fragmented pathways and sequential assessments can delay recognition of co-occurring needs. The policy implication is that services must look beyond individual diagnoses and understand need in terms of functioning, participation, context and cumulative complexity.

9.1 What is meant by co-occurrence and what is known?

9.1.1 What is co-occurrence?

People rarely present with a single, neatly bounded condition. More commonly, they experience combinations of neurodevelopmental conditions, mental health difficulties

and physical health problems that interact with one another and often change over time.

Co-occurrence refers to the presence of two or more conditions at the same time. The Review uses this term rather than comorbidity because it makes fewer assumptions about why conditions occur together. Multimorbidity is used when several conditions coexist simultaneously.

Co-occurrence may occur between neurodevelopmental conditions, such as autism and ADHD; between neurodevelopmental and mental health conditions, such as autism and anxiety; or across neurodevelopmental, mental health and physical health conditions, such as ADHD and chronic pain. These patterns often overlap but are frequently assessed and treated through separate pathways. Distinguishing between them helps identify where gaps in assessment and care arise.

The high level of co-occurrence may appear surprising. If two conditions each affect only a small proportion of the population, chance alone would predict relatively few people having both. In reality, co-occurrence is far more common than chance would predict. This pattern is seen throughout medicine but is particularly striking in neurodevelopmental and mental health conditions. It suggests that diagnostic categories are not entirely discrete, but overlap in ways that influence how difficulties are experienced, recognised and supported.

9.1.2 Why is co-occurrence so common?

There is no single explanation for why co-occurrence is so common. Rather, several interacting processes appear to contribute.

Many conditions share common risk factors. Genetic influences associated with one condition often increase the likelihood of others. This has been demonstrated across autism, ADHD, intellectual disability and a range of mental health conditions[329]. Environmental influences, including poverty, deprivation, maltreatment, bullying, chronic stress and adverse life events, are likewise associated with increased risk across a wide range of neurodevelopmental and mental health conditions rather than with any single diagnosis[330].

Vulnerability to neurodevelopmental and mental health conditions clusters within families, and people often form relationships with others who share aspects of their cognitive or behavioural characteristics[331]. Children may therefore inherit combinations of strengths and vulnerabilities that increase the likelihood of developing several conditions rather than one alone.

Transdiagnostic models suggest that many of the processes underlying distress and impairment are shared across diagnostic categories. Difficulties such as emotional dysregulation, social avoidance, intolerance of uncertainty and alexithymia occur across autism, ADHD, anxiety, depression and personality disorders[332, 333]. This helps explain why diagnostic boundaries are often less distinct in practice than they appear in classification systems.

Current diagnostic systems also contribute to apparent co-occurrence. Diagnostic categories are based on patterns of symptoms rather than established biological mechanisms. Difficulties with concentration, for example, may arise in ADHD, anxiety, depression, sleep disturbance or trauma-related conditions. Comprehensive assessment can often distinguish between these, but brief assessments and screening tools are less reliable. Some apparent co-occurrence therefore reflects genuine overlap, while some reflects the limitations of current classification systems.

Finally, one condition may increase the likelihood of developing another. Neurodevelopmental conditions influence educational experiences, peer relationships, self-esteem, exposure to adversity and access to opportunities, increasing vulnerability to anxiety, depression and other mental health conditions. Co-occurrence therefore often develops across the life course rather than being present from the outset.

9.1.3 Why co-occurrence matters

Co-occurrence changes how needs present. Outcomes worsen as conditions accumulate, affecting functioning, quality of life and mortality. In APMS 2024, 47% of adults with any mental health or neurodevelopmental condition had two or more conditions, and 23% had three or more. Among 16–24-year-olds with a common mental disorder, 57% also screened positive for ADHD traits, 61% for autistic traits, 36% for probable PTSD, and 52% reported a history of self-harm. It is important to note, however, that screening positive indicates an increased likelihood of a condition, rather than confirming its presence. Studies of autistic children similarly show that between 57% and 80% have additional conditions across multiple domains, including emotional disorders, behavioural disorders and ADHD (Figure 9.1).

Figure 9.1 Overlap of common co-occurring conditions in the Quest and SNAP autism samples of children and young people

Source: SNAP (Special Needs and Autism Project) – A population-based cohort study of 12-year-old autistic children[334]. Quest – A clinical cohort study of children aged 4–8 diagnosed with autism[335].

Personality disorder provides a further illustration. In APMS, it was strongly associated with common mental disorders, ADHD traits, autistic traits, self-harm and substance dependence, forming part of the highly co-occurring group with the greatest levels of complexity and impairment[165].

These findings challenge the current focus within health systems on single diagnoses. Complexity is a defining feature of need, with conditions clustering together and interacting across the life course.

“I can’t access services for mental health because I’m neurodivergent. I can’t access neurodivergent services because I have mental health… I’m just too complex.”

Mind focus group participant, 46–55 age group

9.2 Context, functioning and fit with available support

9.2.1 Co-occurrence as a marker of complexity

Co-occurrence is associated with greater functional impairment than single conditions. Outcomes worsen as additional conditions accumulate across education, employment, relationships, daily living, quality of life and mortality.

APMS data demonstrate this clearly, showing a consistent gradient across successive survey waves: increasing numbers of conditions are associated with progressively poorer functioning (Figure 9.2)[15]. In 2024, adults with three or more condition domains experienced markedly greater impairment than those with none. For example, loneliness and social isolation increased from around 12% to almost 80%, difficulties maintaining close relationships rose from around 7% to over 50%, and difficulties with everyday administrative tasks, such as paperwork, increased from around 5% to 40%[15]. Similar graded patterns were observed across activities of daily living, physical health and cognitive functioning.

Figure 9.2 Functional impairment by number of co-occurring mental health and neurodevelopmental conditions (adults aged 16–64 years)

Source: APMS. 95% confidence intervals (CIs) are shown.

9.2.2 The multiple-difficulties group

One of the most striking findings is the existence of a relatively small group of people characterised by multiple overlapping emotional, behavioural and neurodevelopmental difficulties.

Analysis of the 2017 Mental Health of Children and Young People survey identified a group comprising around 4% of adolescents who experienced difficulties across several domains simultaneously. Although relatively small compared with young people with isolated emotional or attentional difficulties, they experienced substantially poorer outcomes[109]. Young people in this group reported very low levels of wellbeing and self-esteem and very high levels of self-harm and peer difficulties. They were also much more likely to experience physical health problems, adversity, special educational needs, educational disengagement, smoking, substance use and social difficulties. Rather than representing more severe forms of a single condition, they represent a distinct pattern of multiple interacting difficulties.

A similar pattern is evident in adults. Around 5% of adults belong to a highly co- occurring group characterised by multiple overlapping conditions and markedly poorer outcomes across employment, homelessness, self-harm, suicide attempts, wellbeing, mental health treatment and unmet need[336].

9.2.3 Educational outcomes

Co-occurring conditions are associated with poorer educational attainment, increased absence and exclusion, and lower qualification rates than single diagnoses[282, 337, 338]. For example, autistic young people with co-occurring ADHD are substantially less likely to complete upper secondary education than autistic young people without ADHD[339]. Outcomes worsen as conditions accumulate, reflecting cumulative impact rather than diagnosis alone.

Preliminary analyses suggest that, once co-occurring conditions are taken into account, the relationship between several individual disorders and exclusion becomes substantially weaker. This suggests that combinations of difficulties may be more important than any single diagnosis in understanding educational disruption (see the Tables Annex).

Some outcomes are also likely to be more modifiable than others. While intellectual ability influences educational attainment, school attendance, exclusion, engagement and participation may respond more readily to appropriate educational support and environmental adaptation.

9.2.4 Employment, adaptive functioning and quality of life

Adults with co-occurring neurodevelopmental and mental health conditions are less likely to participate in the labour market or sustain employment, and are more likely to experience unemployment, economic inactivity and financial disadvantage than those with single conditions[340-343].

Practical skills needed for everyday life – including communication, relationships, independence and self-care – are consistently poorer among people with co- occurring conditions than would be expected from any single diagnosis alone[344-346]. This pattern is particularly well established for autism co-occurring with ADHD but extends across many neurodevelopmental and mental health conditions.

The same pattern is evident in quality of life. Across the lifespan, people with co- occurring conditions report poorer wellbeing, lower quality of life and greater everyday difficulties than those with single conditions[347-349].

9.2.5 Health, healthcare utilisation and mortality

Co-occurrence is associated with increased healthcare use, self-harm, suicide and premature mortality[350-352]. Mortality risk increases as psychiatric conditions accumulate.

9.3 Has co-occurrence become more common?

9.3.1 Looking beyond individual diagnoses

Understanding changes in co-occurrence helps distinguish between different explanations for rising demand. If conditions increase independently, this suggests a different pattern of need than if they increasingly cluster within the same individuals.

The Review draws on repeated population surveys (APMS and MHCYP), together with primary care data from CPRD. Although these datasets have different strengths and limitations, they tell a consistent story. Across multiple sources, co-occurrence has become more common, with neurodevelopmental and mental health conditions – and their interaction with adverse experiences – increasingly occurring together.

They do not, however, establish why this has occurred. The observed changes may reflect increasing underlying traits or symptoms, greater self-recognition, changing reporting behaviour, increased help-seeking, or genuinely greater clinical complexity.

9.3.2 Co-occurrence has increased among adults

The clearest evidence comes from APMS, which allows comparison across 2007, 2014 and 2024. Analysis of six broad domains – common mental disorders, personality disorder, substance dependence, ADHD traits, autistic traits and probable psychosis – shows a substantial increase in the proportion of adults experiencing difficulties across multiple domains. The proportion of adults aged 16– 64 years with three or more condition domains almost doubled, from 4.4% in 2007 to 8.8% in 2024 (Figure 9.3). Over the same period, the proportion with two condition domains increased from 6.8% to 11.2%, while the proportion with no identified condition fell from 68.0% to 57.3%[15].

Figure 9.3 Number of co-occurring mental health and neurodevelopmental conditions over time among adults aged 16–64 years

Source: APMS. 95% CIs are shown.

The pattern is even more pronounced among people already experiencing common mental disorders. By 2024, 35.9% had three or more condition domains (Figure 9.4). Similar patterns are evident among people screening positive for elevated ADHD or autistic traits.

Figure 9.4 Number of condition domains among adults (16+) with each condition

Source: APMS. 95% CIs are shown.

9.3.3 Young adults show the sharpest increase

The greatest increases in co-occurrence are evident among young adults. Among 16–24-year-olds with a common mental disorder, co-occurring neurodevelopmental difficulties and self-harm became substantially more common between 2007 and 2024. The proportion screening positive for ADHD traits increased from 36% to 57%, the proportion screening positive for autistic traits from 20% to 61%, and the proportion reporting a history of self-harm from 28% to 52%[15].

9.3.4 Similar patterns are emerging in children and young people

Evidence from the MHCYP surveys suggests that similar changes are occurring among children and young people. The correlation between autism and ADHD increased from weak (0.25) in 2004 to moderate (0.66) in 2017. However, some of this increase may reflect changes in diagnostic systems, as DSM-IV did not permit simultaneous diagnosis of autism and ADHD, whereas DSM-5 does[144].

Across both ADHD and autism, the period has seen increasing overlap with emotional difficulties rather than behavioural disorders. This closely mirrors the pattern observed in adults and suggests that emotional distress has become increasingly intertwined with neurodevelopmental conditions.

9.3.5 Evidence from clinical records

Primary care data from CPRD also show that the profile of people receiving autism and ADHD diagnoses has changed over time. Among adults newly diagnosed with autism, prior records of intellectual disability, epilepsy and psychotic disorders became proportionately less common between 2010 and 2023, while previous

records of anxiety, depression, eating disorders, self-harm, sleep disorders and stress-related conditions became increasingly common, consistent with the changing case-mix described in Chapter 8[58]. A similar pattern was observed among adults newly diagnosed with ADHD, with increasing histories of anxiety, depression, eating disorders, headaches and chronic fatigue, alongside declining rates of intellectual disability and psychotic disorders[58].

Patterns of co-occurrence in administrative data require cautious interpretation. Diagnostic overshadowing – where one diagnosis dominates clinical attention and additional conditions go unrecognised[353, 354] – means co-occurring conditions are systematically under-recorded, and as clinical awareness of co-occurrence has grown, changes over time may reflect evolving diagnostic practice and recording as much as changes in underlying need[355, 356].

9.3.6 What might explain why co-occurrence has become more commonly reported?

Several explanations may account for the increasing reporting of co-occurrence, and they are not mutually exclusive.

First, increasing recognition of both neurodevelopmental and mental health conditions may make co-occurrence more visible. Diagnostic systems now permit multiple diagnoses more readily, and clinicians may be more alert to overlapping presentations. At the same time, symptom overlap and evolving diagnostic concepts can make distinctions between conditions less clear, particularly in systems organised around single diagnoses.

Second, changing patterns of help-seeking may contribute, with people increasingly recognising and reporting multiple forms of difficulty.

Third, evidence elsewhere in this Review suggests that emotional distress, self-harm and trauma-related difficulties have increased among young people. If these increasingly occur among people with underlying neurodevelopmental vulnerabilities, co-occurrence would also be expected to increase.

Finally, it is possible that patterns of need have genuinely become more complex. The available evidence cannot fully disentangle these explanations, but it consistently indicates that contemporary demand for assessment increasingly involves combinations of difficulties rather than isolated conditions.

9.4 Co-occurrence between mental health and neurodevelopmental conditions

9.4.1 Co-occurrence within neurodevelopmental conditions

Co-occurrence within and across neurodevelopmental conditions is common and clinically important. Around one in four autistic people also have ADHD[79, 80, 357], while between 10% and 20% of people with ADHD have autism[358]. UK data from the MHCYP survey found that 43% of autistic children met criteria for ADHD, while 17% of children with ADHD met criteria for autism; estimates vary according to the populations studied and the methods used to ascertain each condition. ADHD also co-occurs in around half of children with Tourette syndrome[359, 360], while tic disorders are present in around 10% of autistic people[80]. Around one quarter of autistic children and young people have co-occurring intellectual disability[357, 361], placing them at substantially increased risk of self-injurious behaviour and restrictive interventions.

Specific developmental disorders are common but often overlooked. Developmental language disorder (DLD), developmental coordination disorder (DCD), dyslexia and other specific learning disorders commonly co-occur with autism, ADHD and one another. DLD affects around 7.6% of children aged 4–5 years[362]; its prevalence is higher in more deprived communities and among children in care and those in the youth justice system[363, 364]. Children with ADHD are around three times more likely to have language difficulties than those without[365], while children with mental health conditions are around five times more likely to have language recorded as their primary special educational need[366]. Children with ADHD, dyslexia or dyscalculia are also two to three times more likely to have another specific learning disorder[367].

These conditions also have important mental health consequences. Children with co- occurring ADHD and DCD have higher rates of depression than those with either condition alone[368]. ALSPAC studies link DCD and other neurodevelopmental difficulties to later emotional problems and psychotic experiences[369, 370], while developmental language disorder increases the risk of both internalising and externalising difficulties[371]. Children with low language proficiency are also more likely to experience persistent school absence, even after accounting for SEND status[372].

9.4.2 Co-occurrence of neurodevelopmental and mental health difficulties

A consistent finding is that neurodevelopmental and mental health conditions frequently occur together.

As discussed in Chapters 7 and 8, anxiety, depression and emotional dysregulation occur substantially more often among people with autism and ADHD than in the general population[79, 278, 373]. Difficulties in personality functioning also commonly co- occur with neurodevelopmental conditions[374].

9.4.3 Self-harm, suicidality and trauma

The growing overlap between neurodevelopmental conditions, self-harm and trauma-related difficulties is striking.

Autistic people and people with ADHD report substantially higher rates of self-harm and suicidality than the general population[351, 375], with the strongest associations observed among those with multiple co-occurring conditions. The evidence does not support simple causal explanations for these findings. Neurodevelopmental conditions do not inevitably lead to self-harm, suicidality or trauma-related difficulties. Rather, these associations arise through the interaction of developmental vulnerability, psychological processes and adverse experiences across the life course.

9.4.4 Additional complex presentations

Some important clinical presentations sit outside current diagnostic frameworks. Behaviours that challenge, including self-injury, aggression and interfering repetitive behaviours, occur in around 10–15% of people with intellectual disability and are more common in the presence of autism, lower IQ and limited communication[376]. They remain a major cause of exclusion from education and community activities and an important driver of antipsychotic prescribing, despite NICE guidance recommending psychosocial approaches as first-line intervention[377, 378].

Pathological or extreme demand avoidance remains clinically contested, with ongoing debate about whether it represents a distinct profile or is better understood within existing diagnostic categories[379]. Autistic burnout is also increasingly recognised but remains poorly defined and under-researched[322]. These examples illustrate that clinically important forms of distress and impairment do not always fit neatly within existing diagnostic systems.

9.5 Physical health and multimorbidity

9.5.1 Physical health is part of the picture

Physical health conditions commonly co-occur with neurodevelopmental and mental health conditions and are often the first reason people seek healthcare. Understanding physical health alongside neurodevelopmental and mental health

conditions matters because physical health directly influences functioning, participation and quality of life.

9.5.2 Patterns of physical health co-occurrence

A growing body of evidence demonstrates increased rates of physical health conditions among people with neurodevelopmental and mental health conditions. Among autistic people, higher rates of epilepsy, sleep disorders, gastrointestinal problems, obesity, asthma, immune-related conditions and chronic pain have all been reported[277, 280]. Conditions often grouped under the umbrella of ‘medically unexplained symptoms’, including postural orthostatic tachycardia syndrome (POTS), hypermobile Ehlers-Danlos syndrome, Me/Cfs, fibromyalgia and irritable bowel syndrome, frequently present with overlapping symptoms such as fatigue, pain, dizziness, poor concentration, sleep disturbance and low mood. The diagnostic label often depends on how symptoms are experienced, which are most prominent, and which specialist is consulted.

Population registry studies further illustrate these associations. Compared with the general population, people with Ehlers-Danlos syndrome are five to seven times more likely to be diagnosed with ADHD or autism[380]. Children and young people with neurodevelopmental conditions are around twice as likely to experience chronic disabling fatigue by late adolescence[381], and autism is substantially more common among people with functional neurological disorder[382, 383]. Similar patterns are evident in ADHD, with increased rates of sleep disturbance, obesity, headaches and asthma[384, 385]. People with neurodevelopmental conditions are also more likely to experience multiple physical health conditions simultaneously[10, 15]. Sleep and circadian rhythm disturbances are particularly common, with insomnia, delayed sleep–wake phase and restless legs syndrome potentially contributing to presentations of autism and ADHD. Sleep is discussed further in Chapter 10.

Physical health conditions form part of a broader pattern of multimorbidity and often require approaches to management that recognise neurodevelopmental needs alongside physical illness.

9.5.3 Multimorbidity and cumulative burden

Multimorbidity describes the presence of multiple health conditions within the same individual and helps capture the interaction between neurodevelopmental, mental health and physical health conditions.

Many people with the greatest needs experience difficulties across all three domains simultaneously. The cumulative burden often reflects the interaction between them. Physical health conditions may affect attendance, concentration, social engagement

and independence. Educational difficulties may interact with chronic physical illness; sleep disturbance may worsen emotional regulation and attention; and chronic pain may reduce participation and increase social withdrawal. As conditions accumulate, people experience increasing difficulty managing everyday activities, maintaining relationships and accessing appropriate support. Multimorbidity therefore represents a distinct level of complexity characterised by interacting difficulties across multiple domains of functioning.

“I am an autistic and ADHD individual living with multiple physical health conditions, including hypermobile Ehlers-Danlos syndrome, polycystic ovary syndrome, and chronic migraines. Whilst my journey navigating neurodiversity has been far from straightforward, I have found the greatest challenge to be navigating co-occurring physical health conditions within an often inaccessible healthcare system. I dread medical appointments.

Overcoming logistical, emotional and systemic barriers, including the anxiety of appointments, the fear of being questioned or dismissed, diagnostic overshadowing, and the frustration of repeatedly having to justify my diagnoses.”

Lived Experience Partner

9.5.4 Summary

The evidence reviewed suggests that complexity is best understood as the interaction between neurodevelopmental, mental health and physical health conditions across the life course. As conditions accumulate across these domains, impairment and support needs tend to increase. Autism, ADHD, anxiety, depression, self-harm, trauma-related difficulties and long-term physical health conditions frequently occur together and interact in ways that shape functioning, participation and wellbeing. These patterns appear to be becoming more prominent over time, particularly among children, young people and young adults. They reinforce one of the central conclusions of this chapter: many contemporary presentations are characterised not by isolated disorders but by overlapping and interacting patterns of neurodevelopmental, emotional and physical difficulty. Understanding need therefore requires attention to this complexity as well as to individual diagnoses.

9.6 Diagnostic overshadowing, delayed recognition and incomplete diagnosis

9.6.1 Why co-occurrence creates diagnostic challenges

Symptoms may overlap, evolve over time, or be interpreted through the lens of an existing diagnosis, making it difficult to recognise the full pattern of a person’s needs. Diagnostic systems are designed to identify specific conditions, yet many people present with combinations of difficulties that do not fit neatly within a single category. Different conditions also become apparent at different stages of development. Neurodevelopmental vulnerabilities may be present from early childhood but become clinically significant only as educational, occupational or social demands increase. Anxiety, depression, trauma-related conditions and eating disorders may emerge later, altering how underlying neurodevelopmental characteristics are recognised or expressed.

The challenge for clinicians is therefore often not whether a particular diagnosis is present, but how several interacting conditions contribute to a person’s difficulties.

9.6.2 Diagnostic overshadowing

Diagnostic overshadowing occurs when one condition reduces the likelihood that another will be recognised. Autism may be overlooked because attention is focused on anxiety or depression; ADHD may similarly be interpreted as anxiety or depression; mental health conditions may receive insufficient attention once a neurodevelopmental diagnosis has been made[82-84]; trauma-related symptoms may be attributed to autistic characteristics; and autistic characteristics may be overlooked in people receiving treatment for eating disorders, anxiety disorders or personality difficulties[71, 249].

“Once you’ve got an autism label, no one wants to see any other medical problem. Our son has suffered from years of chronic constipation, eczema, insomnia, aggression, self-harm and anxiety and it is all just explained as part of the autism and not investigated, let alone treated. No one wants to know or do anything about it. They would not treat neurotypical people this way.”

Lived Experience Partner CPRD analyses illustrate these delays. Among children diagnosed with autism before the age of 18, diagnosis was delayed by around three years following a prior anxiety diagnosis, five years following depression, and seven years following intellectual disability. Among children receiving either an ADHD or autism diagnosis before age 18, girls were diagnosed several years later than boys across a range of prior diagnostic categories.

9.6.3 Physical health and diagnostic overshadowing

Diagnostic overshadowing is equally important in physical healthcare. Physical symptoms may be attributed to existing neurodevelopmental or mental health conditions rather than investigated in their own right. Equally, physical health conditions may contribute to emotional or behavioural difficulties without being recognised as the underlying cause. Research involving autistic people and people with intellectual disability has consistently identified serious concerns about delayed recognition of physical health problems[300, 386, 387].

9.6.4 Delayed recognition of co-occurrence and its consequences

Diagnostic overshadowing reflects the reality that conditions often share symptoms and that delayed recognition of co-occurrence is more likely in systems organised around individual diagnoses and service boundaries. Adults receiving an ADHD or autism diagnosis often report years of contact with mental health services before receiving their neurodevelopmental diagnosis[249, 301].

“You get put on a pathway for one thing, you spend a lot of time sitting on that, you get to the end of that, but then you have to go back to the beginning

  • of another one somebody goes in with one thing and they come out with

three.”

Lived Experience Partner Delayed or incomplete recognition can have important consequences. Treating a single condition without recognising co-occurring difficulties is often less effective. For example, in eating disorders, autism-related characteristics are associated with longer illness duration and more complex recovery[388, 389]. In ADHD, untreated illness is associated with increased substance misuse, criminality and broader psychosocial harm[390, 391]. Conversely, earlier ADHD treatment is associated with lower rates of substance misuse and suicidal behaviour[106], while earlier diagnosis has been associated with better educational outcomes[392].

“Our daughter was deteriorating before our eyes, but we kept being told she was a popular, straight-A model student. Repeated referrals to CAMHS from the GP were rejected. …we had done our own research and concluded she may have an underlying neurodevelopmental condition. CAMHS eventually accepted a referral but still insisted on

putting her on a mental health pathway rather than a neurodevelopmental one. Our daughter couldn’t engage with talking therapy or CBT and felt more and more like she was a failure.

Finally, after repeated complaints and escalations, our case ended up with a neurodevelopmental consultant. He referred to our daughter as a ‘missed case’ and confirmed what we had understood for years: she couldn’t engage with the treatment offered because she was struggling to cope with an undiagnosed neurodevelopmental condition. A joint pathway was put in place between CYPMHS and the neurodevelopmental team. I can honestly say this has been life- changing. For the first time in years, I have hope that maybe we can aspire to more than survival, maybe it will be possible for her to thrive.”

Lived Experience Partner

9.6.5 Summary

Co-occurrence poses significant challenges for recognition and diagnosis. Diagnostic overshadowing, delayed recognition and sequential diagnosis are common consequences of conditions that overlap, evolve over time and become apparent in different ways across the life course. For many people, the challenge is not that an earlier diagnosis was incorrect, but that it was incomplete. Different aspects of need may become visible at different times, and multiple conditions may only be recognised gradually as difficulties accumulate or circumstances change. Many of the difficulties associated with co-occurrence arise because systems organised around individual diagnoses are not designed to recognise or respond effectively to complexity.

9.7 Adversity, trauma and developmental vulnerability

9.7.1 Adversity and co-occurrence frequently occur together

Adversity and co-occurrence are closely linked. People experiencing multiple conditions are also more likely to have experienced adverse life events and social disadvantage[109, 393]. They are substantially more likely to report family conflict, abuse, neglect, bullying, household instability and exposure to violence.

Adversity may also be more difficult to manage for people with neurodevelopmental conditions. The COVID-19 pandemic provides one example: although evidence remains limited, available studies suggest that people with neurodevelopmental

conditions were particularly vulnerable to its effects on everyday life. Adversity therefore forms part of the wider context within which multiple difficulties emerge and accumulate[394].

9.7.2 Why adversity may increase complexity

Adversity may contribute directly to emotional distress and trauma-related difficulties. Experiences such as abuse, neglect, victimisation and chronic stress are associated with increased risks of anxiety, depression, PTSD and self-harm across the population[395]. It may also interact with existing neurodevelopmental vulnerabilities. A child with attentional difficulties may struggle academically and experience repeated criticism or failure, while a young person with social communication difficulties may become increasingly isolated or vulnerable to bullying. Over time, these experiences may contribute to additional emotional and behavioural difficulties.

Adversity may also reduce access to the relationships and environments that protect wellbeing. Family instability, poverty, school exclusion and social disadvantage can all limit opportunities for support and increase the likelihood that difficulties persist or accumulate. During the COVID-19 pandemic, online learning, disruption to social relationships, changes in routine and uncertainty about the future may have had a disproportionate impact on people with neurodevelopmental conditions. These explanations are not mutually exclusive and are likely to reinforce one another across development.

9.7.3 Adversity, bullying and victimisation

Certain forms of adversity appear particularly important for people with neurodevelopmental conditions. Bullying, victimisation, social exclusion and exploitation are consistently reported at higher rates among autistic people, people with ADHD and people with intellectual disability[396-398]. These experiences may, in turn, contribute to anxiety, depression, self-harm and trauma-related difficulties. Importantly, vulnerability should not be understood as residing solely within the individual, but also within the social and institutional contexts in which people live, learn and work.

9.7.4 A developmental accumulation model

Complexity often develops gradually. A child may begin with attentional, language or social communication difficulties, which contribute to educational struggles, peer problems or exclusion. These, in turn, increase the likelihood of anxiety, low mood, reduced self-esteem or behavioural difficulties, with additional challenges such as school absence, self-harm and social withdrawal emerging over time.

The result is not a simple progression from one condition to another, but an accumulation of interacting vulnerabilities and environmental experiences. This helps explain why co-occurrence often becomes more apparent during adolescence and early adulthood, as new challenges arise. This process is not inevitable. Supportive families, effective schools and timely intervention can alter developmental trajectories. Nevertheless, adversity plays an important role in shaping how complexity develops across the life course.

9.7.5 Implications for understanding need

Some of the greatest levels of need often arise where neurodevelopmental vulnerabilities intersect with adversity, trauma and social disadvantage. Needs are often better understood through the interaction between developmental vulnerabilities and life experiences than through any single diagnosis alone.

This perspective does not diminish the importance of diagnosis. Rather, it suggests that diagnosis alone provides only a partial understanding of need. Understanding functioning, participation and support requirements also requires attention to developmental context and lived experience.

9.8 Complexity in special settings

9.8.1 What the secure estate reveals

Children and young people in the secure estate (including youth custody, secure training centres and secure children’s homes) experience exceptionally high rates of neurodevelopmental conditions, mental health difficulties and social adversity. ADHD, communication difficulties, intellectual disability, traumatic brain injury, conduct disorder, anxiety, depression and PTSD are all substantially more common[313, 399, 400]. Among adults, screening studies of male prisoners in England and Scotland found that 25% screened above threshold for ADHD, 9% for autism and 9% for intellectual disability[401]. These difficulties commonly cluster within the same individuals, many of whom have experienced significant trauma, attachment difficulties, and co-occurring mental health and neurodevelopmental conditions[253,

402, 403].

Entry into custody or secure care is rarely the result of a single event. More often, it represents the endpoint of a prolonged developmental pathway characterised by repeated missed opportunities for recognition, support and intervention[404, 405].

Children with complex presentations often move repeatedly between education, health, social care and justice services, yet their needs are understood through only one part of the picture at a time. Diagnostic overshadowing, fragmented pathways

and threshold-based models of care all contribute to delays in recognising the full pattern of need[406, 407]. The secure estate should not be viewed as fundamentally different from the populations discussed elsewhere in this chapter; rather, it represents an extreme concentration of processes that are evident more widely.

9.8.2 Other special settings

Similar patterns are evident in other settings where complexity is concentrated. Autistic people and people with ADHD are over-represented in inpatient and secure forensic mental health services, where co-occurring neurodevelopmental and mental health conditions are associated with increased crisis presentations, self-harm, suicide risk and greater service complexity.

Looked-after children also experience elevated rates of neurodevelopmental and mental health conditions, reflecting the interaction of inherited vulnerability with prenatal exposures, maltreatment and neglect[408]. Neurodevelopmental conditions also remain under-recognised in this population[409].

9.9 Treatment modification and co-occurrence

9.9.1 Pharmacological treatments can be differentially effective when there is co-occurrence

Children with co-occurring autism and ADHD respond to stimulant treatment at lower rates than those with ADHD alone (around 50% compared with 70–80%), with smaller effect sizes and more frequent adverse effects[410]. Nevertheless, stimulants remain an effective and recommended treatment for many in this group. Similarly, while SSRIs are clearly beneficial for anxiety in typically developing young people[411], evidence of benefit in autistic children is much weaker. Several well- powered randomised controlled trials have found only small, non-significant effects of SSRIs on anxiety[412-414].

Lived experience accounts echo this pattern. One participant described being prescribed SSRIs for depression before their underlying ADHD had been recognised:

“I went to the doctors feeling entirely overwhelmed and that all happiness and purpose had gone. I was diagnosed as depressed and given SSRIs with no curiosity as to why I felt this way, or any other support. After around 18 months I found the side effects (weight gain, sluggishness and a ‘numb’ feeling) outweighed the benefits. The GP recommended I stop taking them, and not to bother tapering, as they weren’t entirely convinced SSRIs even worked. This directly led to a

very difficult period in my life, and I came close to not being able to live with my family.”

Lived Experience Partner

9.9.2 Adapted psychological treatments may lead to better outcomes in neurodivergent people

Emerging evidence suggests that neurodivergent people benefit more from psychological therapies adapted to their cognitive, communication and behavioural styles[415, 416]. Modified CBT, already recommended by NICE for autistic young people[417], appears particularly promising. Studies report better outcomes than unadapted approaches, although most comparisons have been with usual care or waiting-list controls rather than with standard CBT directly[418-421]. Such study designs are generally given substantially less weight in NICE guideline development because they provide less robust evidence of comparative effectiveness.

In eating disorders, standard approaches are less effective for autistic individuals unless appropriate adaptations are made[422]. For example, family-based approaches may not adequately address the needs of autistic young people, particularly in relation to sensory sensitivities, rigidity and emotional regulation[423]. Autism-informed service adaptations have shown encouraging results, including reduced inpatient stays and associated cost savings[424].

9.10 Complexity, diagnosis and need

9.10.1 Co-occurrence challenges categorical thinking

Although diagnostic categories remain fundamental to clinical practice, research and service organisation, the evidence consistently shows that neurodevelopmental, mental health and physical health conditions overlap in combinations that do not fit neatly within a single diagnosis. The accumulation of difficulties often has a greater influence on outcomes than any individual diagnosis considered in isolation.

Diagnosis alone provides only limited information about how someone is functioning in everyday life. Understanding need also requires attention to functioning, participation, co-occurrence, adversity, physical health and environmental context. Complexity often reflects the interaction between multiple conditions and life experiences. Support should therefore be informed not only by diagnosis but also by the way difficulties affect functioning and participation.

“I wish someone would look at the whole me. Not just each diagnosis in isolation, but how they interact and shape my everyday life. I want to feel listened to, rather than judged, assumed, or leaving yet another appointment feeling rejected and lost.”

Lived Experience Partner

9.10.3 Why single-condition systems struggle

Systems organised around individual diagnoses often struggle to respond to co- occurrence and complexity. People may move repeatedly between services, receive sequential diagnoses and experience fragmented care. The challenge is not simply identifying individual conditions, but understanding how they interact, and providing holistic, evidence-based care.

9.10.4 Summary

Co-occurrence is not a peripheral issue; it is a defining feature of how need commonly presents in practice. Across child and adult populations, conditions frequently overlap, interact and accumulate across the life course. Many of the outcomes that matter most to individuals and families are influenced less by any single diagnosis than by the interaction of multiple difficulties across several domains. The challenge for services is therefore not simply to identify diagnoses accurately, but to recognise and respond to the complexity of need that often extends beyond any single diagnostic category.

Section Three: What is driving these changes?

Chapter 10. Environmental drivers of change in mental health for children and young people

The rise in mental distress among children and young people has unfolded alongside profound changes in the conditions of childhood. No single cause explains it. Family life, education, communities, digital environments and economic security all influence whether children feel safe and connected, have time to recover, can participate in everyday life and can look to the future with confidence.

Among the possible drivers, digital environments and educational pressures show the closest correspondence with the timing and pattern of rising distress. Neither is likely to be the principal cause. The evidence indicates that these environments have become more significant contributors than for previous generations, interacting with family circumstances, economic insecurity and existing vulnerability rather than operating alone. Continuous comparison and visible evaluation, disrupted sleep, reduced opportunities for belonging and recovery, and growing uncertainty may reinforce one another.

The strongest protective factors for mental health are supportive parenting relationships, positive educational experiences, and community participation and belonging. The pandemic, economic insecurity, pressures on families and the decline of community infrastructure do not explain the trend on their own, but they have intensified pressures and weakened sources of protection. These effects are not evenly distributed. Children experiencing poverty, discrimination, SEND, care experience or family stress often bear the greatest burden.

Prevention cannot rest solely on helping children cope with environments that place growing demands on them. It also requires action to improve those environments: more inclusive schools; protected time for rest, play and belonging; stronger community infrastructure; reduced poverty and insecurity; practical support for parents and carers; and enforceable duties on digital platforms to protect children while preserving the connection and support they can provide.

10.1 Introduction

The preceding chapters established that emotional distress, common mental disorders and neurodevelopmental conditions have become increasingly prominent among children and young people. This chapter considers why.

It examines how the environments in which children grow up have changed. Chapter 11 considers how experiences are recognised and understood, while Chapter 12 examines how health, education and care systems have responded. Together, these chapters explore changes in children’s environments, systems of meaning, and institutional responses.

Figure 10.1 shows that the greatest deteriorations among young people concern confidence, worry, self-esteem, appearance and school life, with the largest changes among girls. Smaller changes are evident among boys, and increases occur across the deprivation gradient. Conduct-related difficulties changed less, with some indicators remaining stable or improving. The pattern points towards increasing concerns about evaluation, uncertainty and self-worth, rather than a general deterioration across every domain of mental health.

Figure 10.1 Percentage-point change in 10- to 15-year-olds endorsing negative responses to life satisfaction, mental health and self-esteem questions, by sex, 2010–2023

Source: UKHLS Economic insecurity, educational demands, family stress and disadvantage can accumulate, while supportive relationships, positive educational experiences, community participation and a sense of belonging can protect mental health. Figure 10.2 summarises how these influences and protective factors interact.

Figure 10.2 Social change and shared mechanisms of influence on mental distress for children and young people considered in this chapter

10.2 Family context, adversity and care systems

10.2.1 Family environment: parental wellbeing and changing demands

The strongest evidence linking changes in family life to children’s mental health concerns parental mental health. In England, the proportion of children whose mothers had a recorded mental illness diagnosis increased from 22.2% in 2005–07

to 25.1% in 2015–17[425]. A substantial body of evidence demonstrates associations between parental and offspring mental health, with these relationships generally stronger for mothers than for fathers[426].

Children experience parental distress through its effects on family life: changes in emotional availability, altered family routines, heightened stress, and the ways in which parents themselves respond to adversity. Supportive parenting is associated with better outcomes, and parenting-focused preventive programmes produce lasting improvements in children’s emotional wellbeing[427].

Contemporary parenting must also be understood in the context of economic uncertainty and rising expectations of parenthood. Mothers now spend more time on developmental care – including reading, playing and supporting learning – than previous generations[428], while maternal employment increased from 66.5% in 2002 to 75.6% in 2021[429]. Expectations of close involvement in every aspect of children’s development have grown, while responsibilities within families remain unequal[430]. These pressures may contribute to parental stress and perinatal mental ill health[431].

Children also have less independent mobility than previous generations. They spend less time travelling independently, playing outdoors unsupervised or exploring their environment without adults[432]. This appears to reflect increasing parental perceptions of risk rather than objective changes in safety, contributing to a substantial decline in the proportion of children travelling to school independently, from around 84% in the 1970s to around 25% forty years later[433].

These concerns gained considerable public attention following Jonathan Haidt and Greg Lukianoff’s The Coddling of the American Mind (2018), which argued that ‘safetyism’ – extending from physical safety to ‘emotional safety’ – has interfered with young people’s social, emotional and intellectual development. However, the evidence that over-parenting is a major driver of worsening youth mental health is weaker than public debate often suggests. A meta-analysis of more than 21,000 participants (average age 18-21 years) found only small associations between over- parenting and offspring anxiety and depression, with effect sizes varying by culture, parent gender and developmental stage[434]. Similar conclusions have been reached in systematic reviews of helicopter parenting and children’s independent mobility[435,

436].

10.2.2 Family structure, family stability and children’s mental health

In contrast to the factors discussed above, changes in family structure – such as increases in cohabitation and declines in marriage[437] – are unlikely to explain more than a modest proportion of the rise in children’s mental health problems. From the

mid-1990s to 2024, the proportion of lone-parent families increased only slightly, from around 14-15% to approximately 16%. Parents are also older when children are born, with mean maternal age rising from around 26 to approximately 31 over the same period[437].

Family structure itself is a weaker predictor of children’s mental health than the quality, stability and emotional climate of family relationships. Divorce rates have remained broadly stable, but cohabiting relationships are generally less stable, resulting in more children experiencing parental separation[438]. Parental separation is modestly associated with later mental health difficulties[439, 440], but family instability[441] and repeated family transitions[442] are more strongly associated with behavioural and emotional difficulties. Children generally adapt well when relationships remain supportive, conflict is limited, and caregiving remains predictable following separation. Family instability may therefore increase vulnerability, but gradual changes in family structure over recent decades do not correspond closely with the more recent rise in mental distress.

10.2.3 Mechanisms linking family context to mental health

Emotional security is a principal mechanism through which family life influences mental health. Consistency and emotional availability protect against adversity, while instability and conflict increase vulnerability. The association between parental and offspring mental health is well established[426, 443]. Parental mental health is therefore an important part of the developmental environment in which children grow up.

“I experienced a lot of loneliness when I became a mum. They talk about having a village, but I didn’t have one. That loneliness was a huge part of my mental health getting worse, and there were so many groups that I didn’t even know existed.”

Mind focus group participant, aged 26–35 Evidence from children in care, residential settings and youth justice services reinforces this principle. Across these highly vulnerable populations, stable and enduring relationships with trusted adults are consistently associated with better emotional wellbeing, stronger engagement with services and more positive developmental outcomes[444, 445]. Although these findings arise from children experiencing particularly high levels of adversity, they illustrate a more general developmental principle: children benefit not simply from support, but from support that is reliable and provided through relationships that are meaningful to them. For all these reasons, family context plays a central role in children’s mental health.

10.2.4 Family context, adversity and care systems: likely contribution

Families can both protect and increase risk to children’s mental health. Children exposed to parental distress, repeated instability, chronic conflict, socioeconomic disadvantage, abuse or the care system are particularly vulnerable.

Changes in family structure – and the rise of intensive parenting – do not appear to explain the timing or pattern of increasing distress. Rather, the importance of families lies in the way they mediate children’s everyday experience of wider social change. Supporting parents and carers to provide stable, emotionally available care may therefore represent one of the most important opportunities to strengthen children’s mental health and wellbeing.

10.3 Education, evaluation and exclusion

Schools serve two functions: they are developmental environments in which children negotiate competence, identity and future opportunity, and they are systems through which emotional, behavioural and neurodevelopmental difficulties are first recognised and responded to (see Chapter 13).

The Children’s Society 2025 Household Survey also reported a consistent decline in happiness with both school and schoolwork among 11- to 17-year-olds since 2009, particularly among girls[446]. In a survey of 10,000 children and young people by the Edge Foundation, almost half reported that school was ‘not an enjoyable or meaningful experience’, but rather something they needed to ‘get through’ because of its perceived importance for their future[447]. Many participants, particularly those with creative or practical interests or with special educational needs and disabilities (SEND), described teaching methods as ‘alienating and stressful’.

These concerns related not only to particular teaching methods, but also to broader features of the education system: examination-focused pedagogy, standardised curricula, and limited opportunities for exploration, dialogue and creative engagement. The rigidity of the system and the pressure to perform were perceived to reduce both enjoyment and relevance, increasing stress and disengagement.

Since the early 2010s, educational environments in England have become more focused on high-stakes assessment and school accountability, while opportunities for unstructured time, physical activity, and creative or vocational learning have narrowed. Suspensions, permanent exclusions and persistent absence have risen. Together, these changes may have increased pressure, reduced enjoyment of school and weakened some young people’s sense of belonging.

10.3.1 Evaluation

Assessment is an essential part of education. The question is whether changes in the nature and intensity of assessment have affected young people’s mental health.

There is little doubt that schools have become more continuously evaluative. Since 2010, GCSEs and A levels have largely returned to linear examinations. Scaled attainment measures and accountability frameworks, including Progress 8, Attainment 8, the EBacc and the 2019 Ofsted framework, have increased the prominence of measurable outcomes[448-450]. Entries to arts and design subjects have also declined[451].

The strongest direct evidence for the influence of evaluation on mental distress concerns examinations themselves. Analyses undertaken for this Review using Understanding Society data (2009–2024), excluding the pandemic years when national examinations were cancelled, found that mental distress among 16-year- olds, measured using the GHQ-12, was significantly higher during the GCSE and A- level examination period (May–June) than during the remainder of the year (+0.90 points, 95% CI 0.42 to 1.38; p<0.001). No comparable increase was observed among 17- or 18-year-olds who were not routinely sitting national examinations during the same period. These findings provide clear evidence that high-stakes examinations are associated with a measurable increase in distress.

Figure 10.3 Adjusted difference in mental distress during the May–June examination period compared with the remainder of the year among 16–18- year-olds.

Source: Understanding Society. Adjusted difference in mental distress (measured by the GHQ-12) between the May–June examination window and other months, by age.

Points show the adjusted contrast and 95% confidence intervals. Estimates from linear regression adjusted for sex, ethnicity and calendar year, applied to Understanding Society data on 16- to 18-year-olds in England, 2009–2024 (excluding 2020 and 2021).

HBSC data for England similarly show that reported schoolwork pressure increased between 2014 and 2022 in every age and gender group examined, most markedly among girls. Among 11-year-old girls it increased from 28% to 59%, and among 15- year-old girls from 73% to 82%[452]. Emotional problems followed a similar trajectory, partly explained by schoolwork pressure[453]. Survey evidence is nevertheless mixed. In the HBSC study, perceived school pressure increased more clearly among girls than boys between 2002 and 2014, but did not rise uniformly across all groups[454]. Performance may nonetheless have become more visible, and educational achievement more closely linked to perceptions of future opportunity[447].

Young people may not necessarily describe themselves as feeling more ‘pressured’, while still growing up in environments where performance is more visible, comparison is more pervasive, and educational achievement is more closely tied to perceptions of future opportunity.

“I’d say in general, the GCSE and A-level reforms that happened a few years ago put pressure on more young people. There’s more pressure to go to university because in the news you see people can’t get jobs, so you put pressure on yourself and maybe your parents put pressure on you to go to university. There’s a big emphasis on exams, but no one teaches you what the signs of stress or anxiety are.”

Mind focus group participant, Youth Voice Network, age 14–25

10.3.2 Recovery

Recovery includes opportunities to play, socialise, be physically active, engage creatively and spend time free from formal evaluation. These experiences support emotional regulation and social development. Reduced opportunities for recovery, alongside sustained demands, may worsen the experience of schooling.

Opportunities for recovery within the school day have narrowed. Average break provision has fallen by around 45 minutes per week in primary schools and 65 minutes in secondary schools since the mid-1990s[455]. Secondary school PE fell by approximately 45,000 timetable hours between 2011/12 and 2025[456]. Arts and vocational subjects have declined, while private tutoring increased from around 18% of secondary pupils in 2005 to 30% in 2022[451, 457].

Direct evidence linking these changes to mental health remains limited. The evidence for physical activity is stronger, with longitudinal studies consistently finding that greater activity predicts lower rates of subsequent depression and emotional difficulties[458, 459].

Lived-experience participants linked educational pressure with fewer opportunities for informal play, exploration and time free from adult demands.

“There weren’t youth spaces, and so there just wasn’t anything except social media.”

Mind focus group participant, Youth Voice Network, female/non-binary group, aged 14–25

Reduced opportunities for play, physical activity, creativity and social connection may leave young people less able to absorb other pressures.

10.3.3 ‘Sense of belonging’ to a school

A sense of belonging can weaken as suspensions, exclusions and persistent absence increase, particularly where absence separates children from relationships and support within school.

Suspensions, permanent exclusions and persistent absence have all risen substantially since the early 2010s. Suspensions in England increased from around 287,000 in 2009/10 to 787,000 in 2022/23, while permanent exclusions rose from around 5,000 to 9,400[460]. Persistent absence increased from 10.9% in 2018/19 to 21.2% in 2022/23, before easing slightly to 20.0% in 2023/24[460, 461].

Children with SEND support comprise around 15% of the school population but account for almost half of both permanent and fixed-term exclusions. Exclusion rates are also substantially higher among children living in poverty and among some ethnic minority groups. Exclusions are therefore concentrated among children already experiencing multiple forms of vulnerability. Official figures are also likely to underestimate the scale of the problem, as they do not capture children removed through managed moves or ‘off-rolling’, where a pupil is removed from the school roll without a formal permanent exclusion[462-464] (Chapter 13).

Educational disengagement and mental health influence one another. Secondary- age children with a probable mental health condition were around seven times more likely to miss more than fifteen days of school in 2023[465]. Recognised mental health conditions were associated with nearly six times the odds of exclusion[466], while

persistent absence was associated with approximately twice the likelihood of later mental health difficulties[467].

These findings suggest a self-reinforcing cycle in which vulnerability increases the likelihood of educational disengagement, and disengagement in turn deepens vulnerability. Analysis of the Millennium Cohort Study found that exclusion between ages 11 and 14 was associated with later conduct problems and, among boys, emotional symptoms[468].

Children experiencing emotional dysregulation, learning difficulties, neurodevelopmental conditions, communication needs or trauma may find participation in school increasingly difficult. Where those needs are not recognised or adequately supported, a child’s connection with school may weaken until persistent absence or exclusion follows.

“The education system is not set up for neurodivergent people; mental health support needs to be increased massively in schools, having members of staff that support when needed, safe spaces and accessibility. I go to a lot of youth voice spaces, youth groups, and they’ve literally saved my life.”

Mind focus group participant, Youth Voice Network, female/non-binary group, aged 14–25

Since the early 2010s, more schools in England have adopted highly structured, rule-based approaches to behaviour. These policies can increase order and visible compliance, but they do not necessarily address the underlying causes of behaviour. Where conduct is treated primarily as a disciplinary problem, distress and unmet emotional, developmental or neurodevelopmental needs may be misinterpreted as defiance[469]. Pupils may then experience repeated sanctions, isolation or exclusion while the needs driving their behaviour remain unrecognised. For vulnerable children, this can deepen anxiety, shame, disengagement and loss of belonging. Department for Education guidance requires schools to consider underlying needs before exclusion[470].

In short, schools that foster a strong sense of belonging can provide important protection during childhood and adolescence[471, 472].

10.3.4 Educational transitions

The transition from primary to secondary school coincides with major biological, psychological and social changes. Young people enter larger institutions, friendship networks change, academic expectations increase, and social comparison becomes

more salient. Those with existing emotional, behavioural or neurodevelopmental difficulties, other special educational needs, or socioeconomic disadvantage are at greater risk of persistent distress and poorer adjustment[473, 474]. Post-16 transitions also require decisions about qualifications, employment and future pathways at a stage when identity and aspirations are still developing.

Young people who are not in education, employment or training (NEET) have poorer mental health than those entering education, apprenticeships or employment[475]. Participation provides structure, purpose, social connection and opportunities to develop competence and autonomy.

The abolition of the Education Maintenance Allowance provides a further illustration. Quasi-experimental evidence suggests that removing financial support for low- income young people was associated with poorer mental health among those affected[476].

Transitions are periods during which existing strengths and difficulties become more visible because environmental demands change. They therefore represent particularly important opportunities for timely recognition and support.

Education is unlikely to be the principal driver of worsening mental health, but it appears to have become a more important contributor than for previous generations. Greater emphasis on evaluation, fewer opportunities for recovery, a weaker sense of belonging for some groups – particularly girls – and more challenging transitions have altered children’s developmental environments. Together, these changes are likely to have increased anxiety, uncertainty and reduced confidence about the future.

Although examinations produce measurable distress, they do not explain the long- term trend on their own. The evidence nevertheless suggests that young people experience educational environments as increasingly evaluative, with attainment more closely tied to perceptions of future opportunity. Over time, they report greater unhappiness with school and schoolwork, together with a weaker sense of belonging, particularly girls.

Children with existing emotional difficulties, SEND, socioeconomic disadvantage or other additional needs are more likely to experience these pressures cumulatively. For some children with neurodevelopmental conditions, increasing demands on attention, organisation, sensory regulation, social navigation and independent

learning may reduce participation, even where the underlying condition has not changed[477]. Chapters 7 and 8 consider ADHD, autism and educational demands.

Because schools are the only universal institutions accompanying almost every child throughout development, they are uniquely placed to strengthen belonging, resilience, emotional understanding and participation, and to recognise emerging needs early enough for timely support. The implications for educational policy and needs-led systems of support are considered in Chapters 18–20.

10.4 Changes in community and social infrastructure

10.4.1 Community infrastructure as a developmental environment

Much community infrastructure has eroded over the past fifteen years, particularly since 2010 during the period of austerity. This has reduced opportunities for children to build social capital, develop relationships beyond their families and participate in community life. Community infrastructure – including youth services, libraries, parks and community organisations – forms part of the social foundations that support children’s wellbeing across the population.

10.4.2 What changed?

The contraction of community infrastructure since 2010 has been substantial. Real- terms local authority spending on youth services in England fell by 73% between 2010/11 and 2023/24[478]. Spending on early intervention children’s services fell by 42% over the same period, and by 53% per child in the most deprived local authority areas[479]. This contrasts with the preceding decade, when investment in many children’s services had increased.

The consequences included the closure of youth centres and adventure playgrounds, contraction of Sure Start children’s centres, reductions in family support and parenting services, and erosion of universal community provision such as public libraries[480-482]. By 2018, around one-third of England’s children’s centres had closed, while many of those remaining operated at reduced capacity or served more narrowly targeted populations[482].

These reductions fell hardest on deprived communities, where need was greatest and families had fewer alternatives. More affluent families can often replace free provision with paid-for sport, childcare, tutoring or organised activities. The loss of universal provision therefore widens inequalities in access to activities and

opportunities for play, social connection, trusted adult relationships and family support[483].

“When I was younger, we had a youth club where we spent loads of time. We had somewhere to go, somewhere to be with friends. It’s sad that young people today don’t have that. It almost feels like they’re forced to socialise on screens instead.”

Mind focus group participant, aged 46–55 Opportunities for informal community participation have also declined. More than 400 public playgrounds closed in England during the decade leading to 2022, while local authority spending on parks fell by more than £350 million between 2011 and 2023[484]. Volunteering and civic participation have also declined over the past decade[485].

10.4.3 What does the evidence suggest?

The strongest evidence comes from evaluations of Sure Start children’s centres. Using the staggered introduction of Sure Start across England as a natural experiment, researchers found that greater exposure to Sure Start during early childhood approximately halved the probability of mental health-related hospital admissions between the ages of 12 and 14, with the largest benefits observed in more disadvantaged communities[486].

Evidence relating to other forms of community provision is broadly consistent. A review pooling 24 randomised and quasi-experimental studies involving 23,000 young people found small improvements in emotional wellbeing and self-concept following community-based interventions, although many studies were judged to be at high or unclear risk of bias[487]. Area-level analyses have also linked larger reductions in local authority spending on cultural, environmental and planning services with small deteriorations in population mental health and wellbeing[488]. Although studies of this kind cannot establish causality at the individual level, they are consistent with the broader pattern across the evidence.

Childhood social cohesion also appears to buffer the effects of stressful life events during adolescence[489], while systematic reviews consistently associate loneliness and social isolation with poorer mental health[490].

Participants in the Review’s lived-experience discussions described the loss of community spaces and informal support networks as contributing to loneliness and isolation. They emphasised the importance of places outside home and school where friendships, shared activities and supportive relationships could develop.

Although the causal evidence remains less extensive than for some other environmental changes, communities characterised by stronger social connection, greater opportunities for participation and more accessible local resources are consistently associated with better mental health and greater resilience, particularly among children and young people facing disadvantage.

10.4.4 Mechanisms linking community and social infrastructure to mental health

Belonging appears to be the most important mechanism. Community settings provide opportunities to develop friendships, participate in shared activities, encounter trusted adults and build identities beyond home and school. The loss of community spaces diminishes opportunities for belonging even where families and schools continue to provide support.

Community infrastructure also supports recovery. Parks, sports facilities, youth clubs and recreational activities provide opportunities for physical activity, play, relaxation and time away from educational, family and social demands, allowing children and young people to restore emotional equilibrium and develop confidence without continuous evaluation or performance.

A third mechanism concerns social capital. Communities characterised by trust, reciprocity and participation strengthen informal networks through which practical and emotional support are often provided.

While changes in community infrastructure do not explain the overall rise in mental distress, they correspond closely with increasing loneliness and reduced social connection. The contraction of community provision coincided with broader declines in civic participation and opportunities for informal social connection. Community infrastructure supports children’s mental health by strengthening the social environments in which development, relationships, belonging and recovery take place.

10.5 Social and economic pressures

Young people experiencing poverty, discrimination, disability, neurodevelopmental conditions, care experience or other forms of disadvantage often face greater barriers to education, employment and independent living, while having fewer family or financial resources with which to absorb setbacks.

The route to independent adult life has narrowed. Secure housing, stable employment and financial autonomy are now achieved later and less reliably than by previous generations.

The labour market has also changed, with weaker wage growth, greater employment insecurity and increasing demands for flexibility, self-management and executive functioning. These changes may be particularly challenging for some young people with neurodevelopmental conditions.

The contribution of child poverty and material hardship requires careful interpretation. Poverty rates during parts of the 2010s were lower than in the 1990s, so poverty alone cannot explain the broader rise in distress. However, austerity-era reforms appear to have increased persistent poverty for some families, while the cost-of-living crisis added more recent financial pressures[491]. The latter is likely to be particularly relevant to recent and future trends.

These developments affect more than material circumstances. Economic insecurity constrains housing choices, unstable work delays independence, and poverty restricts educational and social opportunities. Together, they may weaken autonomy and confidence that a secure future remains within reach.

10.5.1 Growing challenges to gaining financial independence

Among 25- to 34-year-olds, home ownership fell from 56% in 2005/06 to 38% by 2015/16, while the proportion renting privately almost doubled from 24% to 46%[492]. Among 20- to 24-year-olds, the proportion living with parents increased from 45% in 2011 to 51% in 2021[493].

For those who leave home, the financial burden can be substantial. In 2021/22, private renters aged 16–24 spent an average of 44% of their income on rent. Among those in the lowest 40% of incomes, 78% spent more than 30% of their income on housing costs[494].

A UK study following households between 2009 and 2019 found that private renters struggling to meet housing payments experienced mental distress at rates around five percentage points higher than those not facing such difficulties, with the strongest association seen in areas most affected by reductions in housing support[495].

“If you’re worried that you’re not going to have anywhere to live, or you’re not going to be able to eat, you won’t have any energy…”

Mind focus group participant, Youth Voice Network, female/non-binary group, aged 14–25

The transition to adulthood has become more delayed, more unequal and more difficult to achieve, with barriers accumulating for young people facing poverty, limited family resources, neurodevelopmental differences or mental health difficulties.

“I’m so done living at home with my parents. They both work from home so I’m literally trapped here. My only escape is seeing my friend who lives an hour away, or going to work. I feel stuck.”

Tellmi user, aged 22 “There’s all this stuff on the news about how young people won’t ever be able to buy a house and it’s all really depressing. How loads of young people, even 30-year-olds, still live at home and it just makes it like a bleak future in my opinion.”

Mind focus group participant, Youth Voice Network, age 14–25

10.5.2 Employment precarity

Work can provide financial security, structure, identity, social participation and a sense of future direction. Through employment, young people hope to translate education, effort and aspiration into an independent adult life. When access to secure and suitable employment becomes uncertain, the consequences may extend beyond current income to how young people understand their abilities, opportunities and futures.

Young people now enter a less secure labour market. Among 16- to 24-year-olds, zero-hours employment rose from around 6% in 2013 to 13% in 2026[496]. Millennials born in the late 1980s earned around 8% less at age 30 than Generation X workers born a decade earlier[497]. Average student debt at graduation in England rose from around £25,000 under the pre-2012 system to approximately £45,000–£50,000 for recent cohorts[498].

“Scrolling through jobs in my area and I actually can’t do any of them. Even the most boring ones I can’t even apply to because they say stuff like ‘Have to have 2 years training’ or ‘have to be skilled in IT’ for a retail job. And a receptionist job said ‘have to be willing to travel’.”

Tellmi user, aged 21 Application procedures, selection criteria and workplace expectations may create barriers for people with communication, executive functioning, sensory or other support needs[499, 500]. Young people’s career aspirations have become increasingly concentrated on a relatively small number of professional occupations, while uncertainty about future careers has increased since 2018[501].

Persistent employment insecurity is associated with poorer self-rated health and more mental health symptoms[502]. Danish data link lower control at work with increased risk of depression[503], while Swedish evidence links insecure employment after graduation with later hospital treatment for mental disorder[504]. Employment insecurity may therefore weaken confidence that sustained effort will lead to an independent adult life.

10.5.3 Child poverty and cost-of-living pressures

The trajectory of child poverty in Britain follows a broadly U-shaped pattern. Among children born in 1991–93, around 25% spent at least half of their first ten years in poverty. This fell to approximately 14% among those born in 1998–99 before rising again to 23% among those born in 2016–17[491]. The recent increase appears to reflect the erosion of the real value of benefits during the austerity period together with rising housing costs[505].

Family income is closely linked to children’s mental health. A UK study found that entering poverty during childhood was associated with subsequent increases in emotional and behavioural difficulties[506]. A review of trials and natural experiments similarly concluded that household income has a causal influence on children’s mental health[507].

The cost-of-living crisis added acute pressures. Inflation reached 11.1% in October 2022, driven particularly by housing, energy and food costs[508]. In 2024, 12% of 18- to 24-year-olds reported repeatedly missing bill or credit repayments, compared with 8% of adults overall, while 16% reported feeling heavily burdened by bills[509]. Childline and NSPCC evidence also describes children worrying about family finances, feeling burdensome and becoming pessimistic about the future[510].

The socioeconomic gradient has remained relatively stable while distress has increased across much of the population. Financial hardship is therefore better understood as an important contributor and amplifier, with its effects greatest when combined with other forms of uncertainty and disadvantage.

10.5.4 Mechanisms linking social and economic pressures to mental health

Autonomy is a central developmental task of adolescence and early adulthood. Housing, employment and financial security support this transition. When access to independent living, secure work or financial stability becomes delayed or uncertain, the transition to adulthood may feel less like a gradual expansion of opportunity and more like a prolonged negotiation with constraint.

Belonging and participation are also affected. Financial strain may restrict access to education, transport, leisure and cultural activities through which relationships are formed and maintained. Within families, financial pressure may alter the emotional climate of everyday life by increasing parental stress. For young adults, insecure employment and unstable housing may also make it harder to establish social networks and develop a stable sense of place within a community.

“My parents are going to try to get PIP so they will make me look like I’m completely disabled which I’m not. I do have a disability but they will defo be making me look worse than I am to get the money and none of it will come to me.”

Tellmi user, aged 16

For younger children, these pressures are experienced principally through family life, influencing parental wellbeing, routines, emotional availability and stability[511, 512]. During later childhood, financial pressures become visible through peer comparison and restricted participation in school and social life[513, 514]. For older adolescents and young adults, the developmental significance increasingly concerns the transition to adult life, with housing, employment and financial independence supporting autonomy, identity and confidence in the future.

Social and economic pressures are therefore best understood as shaping the developmental context of childhood, adolescence and early adulthood in interconnected ways. They contribute to rising distress by reducing security, restricting participation, weakening autonomy and making the future feel less predictable.

10.6 The COVID-19 pandemic

10.6.1 The pandemic as a developmental disruption

The pandemic amplified existing vulnerabilities and interrupted many of the experiences – including education, peer relationships and everyday activities – that support belonging, competence, autonomy and confidence in the future.

“It changed everything for me. Suddenly having that massive change that broke all my routines…”

Mind focus group participant, Youth Voice Network, female/non-binary group, aged 14–25

Its cumulative impact arose from disrupting the ordinary business of growing up. For many young people, recovery was rapid. For others, particularly those already facing vulnerabilities or approaching key developmental transitions, the effects proved more enduring.

10.6.2 Disruptions to education, belonging and developmental transitions

Between March 2020 and March 2021, most pupils in England missed around half a normal school year[515, 516]. Around 24% of pupils in the most deprived secondary schools lacked a suitable device for home learning, compared with 6% in the least deprived schools[517].

The disruption extended well beyond formal education. The proportion of young people reporting loneliness ‘often’ or ‘always’ increased from 7% before the pandemic to 17% during it, with larger increases among girls and among young people from Black, African, Caribbean and Black British backgrounds[518]. Around 60% of children reported spending less time outdoors during the first pandemic year[519], while more than 100,000 fewer children and young people met recommended physical activity levels[520].

The COSMO study provides a clear illustration of disrupted transitions. Among more than 13,000 young people in Year 11 during 2020–21, 64% reported changes to their educational plans and 60% to their career plans as a direct consequence of the pandemic[521]. Two years later, around one-third continued to report negative effects on education and mental wellbeing[522]. Young people whose late adolescence coincided with the pandemic also described reduced confidence in their futures and greater uncertainty about educational and occupational opportunities[523]. Comparisons with earlier cohorts suggest delays in some conventional markers of

adulthood, including leaving the parental home, forming resident partnerships and becoming parents[524]. The risk of becoming NEET may also have increased for some whose education was disrupted, although this is better understood as reflecting broader disruption to developmental transitions than educational interruption alone.

“I went from SATs in Year 6 to my A levels. I didn’t do GCSEs because of Covid. I had no idea how to do exams. Life just totally paused for everyone. But Covid’s kind of been forgotten about a bit. Education- wise, we still feel like we’re two years behind.”

Mind focus group participant, Youth Voice Network, aged 14–25 The pandemic disrupted far more than education. It restricted relationships, recreation and many of the everyday experiences through which children develop belonging, confidence and independence. Most young people showed considerable resilience, but the scale and duration of disruption left a measurable legacy.

10.6.3 Mental health during and after the pandemic

The disruptions described above were accompanied by measurable changes in children’s and young people’s mental health, operating both directly and indirectly through parental wellbeing and family strain[525]. Pregnancy, childbirth and the early postnatal period were also profoundly affected, with reduced access to partners, family members and community support during a period of heightened vulnerability for perinatal mental health[526, 527]. However, it is too early to determine whether children born during the pandemic constitute a distinct cohort with enduring mental health difficulties[528].

Reviews of longitudinal studies following the same young people before and during the pandemic generally report increases in anxiety, depression and emotional distress, particularly during periods of stringent restrictions and school closures, with effects generally larger among girls and young women[529-533]. In England, the prevalence of probable mental disorder among 5–16-year-olds increased from 11% in 2017 to 16% in 2020 and remained elevated in subsequent waves[534]. Among 17– 19-year-olds, prevalence increased from 10% in 2017 to 18% in 2020 and remained substantially above pre-pandemic levels in 2023[535].

Figure 10.4. Prevalence of probable mental disorder among children and young people, 2017–2023

Source: MHCYP. Note: differences in age ranges and methods between waves mean exact comparisons should be made cautiously. The overall pattern – rise during the pandemic and only partial recovery – is consistent across indicators

Evidence following the end of restrictions shows partial recovery[536-538]. Symptoms declined from their peak but often remained above pre-pandemic levels. Continuing disruption to education, developmental transitions, family life and social connection appears to have prolonged the effects for some young people.

“That period, the pandemic, shone a spotlight on the fact that we’re all quite lonely and living quite lonely existences.”

Mind focus group participant, age 36–45

10.6.4 Differential effects across groups

Girls and young women experienced larger increases in anxiety, depressive symptoms and emotional distress than boys during and after the pandemic[534, 535,

539].

Young people from disadvantaged households experienced greater educational disruption, higher levels of distress and slower recovery. Unequal access to digital devices, study space and educational support widened inequalities during periods of remote learning[517, 522].

Many autistic young people and young people with ADHD experienced worsening anxiety, disrupted routines, sleep difficulties and reduced wellbeing[540-542]. At the same time, a minority reported benefits from reduced sensory demands, greater flexibility and relief from challenging school environments[543-545].

For children with autism with severe or profound intellectual disability, severe or multiple disabilities, and those reliant on special schools, the disruption was different in scale. School had provided therapies, communication support, routines, social contact and family respite as well as education. Its loss affected behaviour, communication, sleep and family wellbeing[546, 547].

10.6.5 Mechanisms linking the pandemic to mental health

Belonging was immediately affected as schools, colleges, workplaces, sports clubs and community activities closed. The increase in loneliness during the pandemic is consistent with this loss of friendship, shared identity and everyday participation[518].

Recovery was constrained by reduced access to sport, recreation, outdoor activity and informal social contact. For children reliant on specialist schools or group provision, opportunities for recovery could disappear altogether[548].

Certainty also declined as education, examinations, employment, relationships and future plans were disrupted simultaneously. COSMO findings demonstrate lower confidence in future opportunities and a reduced sense of control over life trajectories[523].

Developmental progression was interrupted. Educational transitions, entry into employment, increasing independence and other milestones of adolescence and early adulthood were delayed or experienced under unusual circumstances[549]. For younger children and those with substantial developmental needs, these interruptions may also have affected the acquisition and maintenance of communication, learning, social and daily living skills, particularly where specialist support was withdrawn[528, 548].

These effects accumulated. The pandemic simultaneously disrupted belonging, recovery, certainty, developmental progression and systems of care, while access to protection and support remained unequal.

10.6.6 Likely contribution of the pandemic

The pandemic made a substantial contribution to rising mental distress. Anxiety, depression, loneliness and emotional distress increased between 2020 and 2022 and recovered only partly after restrictions ended[534, 535]. The pattern suggests both a population-level deterioration and an acceleration of trends that were already under way before the pandemic. Administrative CPRD data also show a marked increase in recorded autism and ADHD following the pandemic (see Chapters 7 and 8).

The effects of Covid were uneven. Girls and young women, children from disadvantaged households, children with neurodevelopmental conditions, and those dependent on specialist provision experienced the greatest and most persistent difficulties. For children with severe disabilities, the pandemic often meant the simultaneous loss of education, specialist support, social contact and opportunities to participate beyond the home, while families assumed greatly increased and, in some cases, unsustainable caring responsibilities[550, 551].

The pandemic cannot explain the increase in mental distress that began years before it. Its impact was nevertheless profound. Children already at greater risk experienced heightened fear, disruption and isolation while access to education, care and support narrowed. Many others who had previously been coping lost the routines, relationships and services that had helped keep distress manageable. School, friendships, community activities and contact with trusted adults were abruptly reduced or removed. As these protective structures fell away, manageable problems became more severe, prolonged and complex.

The Review does not revisit the wider public health decisions taken during the pandemic. It does conclude that the prolonged loss of education, social participation, developmental opportunities and specialist care had serious consequences for children’s mental health and development. For children who relied on special schools, therapies, respite and other specialist support, these services were not optional additions to ordinary life. They were fundamental to communication, learning, safety and family stability.

Future emergency planning should treat this support as essential provision, with specific safeguards to maintain access and prevent children with SEND and their families from again bearing a disproportionate share of the harm[552].

10.7 Digital environments and social media

10.7.1 Digital environments as a developmental setting

By 2025, almost all children and young people were online, most owned mobile phones, and digital engagement occupied a substantial proportion of waking hours[553, 554]. Exposure now begins in early childhood, often well before children enter school.

Digital technologies enable communication across distance, provide access to information and educational resources, facilitate creativity and self-expression, and create opportunities for friendship and community. These opportunities may be particularly valuable for young people who are geographically isolated, disabled or

members of minority groups. For some autistic children and young people with ADHD, digital technologies may provide tools for emotional and sensory self- regulation, predictable forms of interaction and opportunities to communicate at a more manageable pace[555-557]. Online communities may reduce isolation and enable young people to connect with others who share similar experiences. Restricting access to social media may therefore have unintended consequences for those who rely on it for connection, regulation or identity-related support.

Digital environments also support parents and carers. Online peer networks can help families of children with SEND find practical advice, navigate assessment and education systems, and learn from other parents with similar experiences[558, 559].

The combination of continuous availability, portability, interactivity, personalisation, social feedback and algorithmic selection is unprecedented. Young people now participate in systems that monitor behaviour, shape what they encounter next and connect individual activity to wider social networks. The key question is how these systems alter children’s developmental experiences and whether particular features help explain rising distress.

Researchers continue to disagree about the magnitude of digital effects and their independent contribution to worsening mental health. The following sections therefore consider the platform features and experiences associated with the greatest risk, the groups most affected, and the developmental mechanisms through which these effects may operate.

10.7.2 What can and cannot be known

Estimating the effects of digital environments on mental health remains difficult, but the limitations do not arise from research methods alone. Platforms hold detailed information about what children encounter, how recommendation systems shape their exposure, how patterns of use change over time, and what their own internal research has found. Much of this evidence remains unavailable for independent scrutiny.

The Online Safety Act 2023 did not create an enforceable right for independent researchers to access these data. The Data (Use and Access) Act 2025 now permits the Government to require platforms to provide them, but access remains limited and largely controlled by the companies under scrutiny. This represents an important regulatory gap. Public policy is being asked to evaluate systems whose operation and effects are most visible to the companies that design them.

Secure and enforceable access to platform data and internal research would allow much more robust assessment of long-term effects, harmful design features and

risks for particular groups of children. Some uncertainty would nevertheless remain because digital technologies are now so embedded in everyday life that no meaningful comparison group exists of children growing up entirely outside digitally mediated environments.

These limitations require proportionate interpretation. Modest individual effects may still have important population consequences when exposure is widespread, repeated or concentrated among vulnerable groups. Equally, statistical associations do not establish causation.

A further challenge is that digital environments do not represent a single exposure. They encompass devices, platforms, design features, content, social interactions and patterns of use, each of which may operate through different mechanisms. Aggregate measures such as screen time may therefore obscure more than they reveal. Two young people spending the same number of hours online may have profoundly different experiences: one may be learning, creating or sustaining supportive friendships, while another is exposed to harassment, harmful comparison or compulsive engagement.

The UK DIORA (Dynamic Interplay of Online Risk and Resilience in Adolescence) study examined how different aspects of young people’s digital experiences related to later anxiety and depression, moving beyond simple measures of screen time[560]. Appearance- and ability-related comparison, risky online activities and exposure to harmful content were more strongly associated with poor outcomes than overall engagement[561]. Social comparison emerged as one of the strongest predictors of later emotional difficulties.

The study relied on adolescent self-report using the newly developed DAFI (Digital Activities and Feelings Inventory), rather than objective platform data, creating the possibility of recall and social desirability biases. Future research combining self- report with passive telemetry and platform-level data should allow more precise understanding of what young people encounter, how they respond, and which experiences precede changes in mental health.

10.7.3 Social comparison, evaluation and identity

Social media platforms expose young people to a far wider range of comparison targets than was previously possible, potentially involving hundreds or thousands of peers, influencers and celebrities. These comparisons are often selective, presenting unusually attractive or successful versions of other people’s lives.

Systematic reviews report associations between appearance-focused social media use, body dissatisfaction, lower self-esteem and depressive symptoms, particularly

among girls and young women[562-566]. The strongest associations generally relate to active comparison with others rather than total time spent online. Experimental studies likewise suggest that exposure to idealised appearance-related content can increase body dissatisfaction and negative mood, while longitudinal research indicates that upward social comparison may partly explain the relationship between social media engagement and lower self-worth[567, 568].

Likes, follower counts and comments make social life visible and measurable. Young people can see how others look, how their own posts are received, who appears to be included and who has been left out. Young people with lower self-esteem, loneliness or pre-existing emotional difficulties may engage more frequently in upward social comparison and may also be more affected by it. At the same time, repeated exposure to idealised comparison targets may further reduce self-worth. The relationship is therefore likely to be reciprocal rather than one-directional.

“…you kind of weren’t going home and then comparing yourself to everyone online.”

Mind focus group participant, Youth Voice Network, female/non-binary group, aged 14–25

Digital environments did not create social comparison. They increased its frequency, reach, visibility and persistence. This provides a plausible pathway to lower self- esteem, body dissatisfaction, feelings of inadequacy and reduced confidence, although it does not explain these trends on its own.

10.7.4 Attention, engagement and cognitive control

Concerns about attention have grown as parents, teachers and young people report greater difficulty concentrating, sustaining attention and disengaging from digital devices.

The clearest experimental evidence concerns interruption and distraction. Notifications and digitally mediated interruptions can impair performance on attention-demanding tasks, reducing concentration and increasing errors[569, 570]. Some studies have also reported that the mere presence of a smartphone can reduce cognitive performance by drawing attention towards the possibility of engagement[571]. However, this finding has not been consistently replicated[572, 573].

Experimental studies therefore show that digital interruptions can affect attention in the short term, although effects are generally modest and vary between studies. They do not establish enduring attentional impairment.

Reviews report associations between problematic smartphone or social media use and ADHD symptoms, depression, anxiety and sleep disturbance[574, 575]. Longitudinal studies likewise identify small associations between greater digital engagement and later attentional difficulties[576-578]. However, young people with attentional or emotional difficulties may themselves be more likely to engage with highly stimulating digital environments, making reverse causality and residual confounding difficult to exclude.

Particular concern has focused on short-form video platforms such as TikTok, Instagram Reels and YouTube Shorts. These platforms present rapid sequences of highly engaging, algorithmically selected content and may encourage frequent shifts of attention. A review of 71 studies involving almost 100,000 participants reported associations between short-form video use and attention problems, stress and anxiety[579], although most studies were cross-sectional and could not determine the direction of causality.

Repeated exposure to highly stimulating content may encourage rapid scanning rather than sustained attention. Continuous novelty and reward may reduce tolerance of boredom and increase sensitivity to distraction. Equally, young people who already find sustained attention difficult may be particularly drawn to fast-paced and highly stimulating content. Current evidence does not allow these possibilities to be distinguished confidently. Neuroimaging studies reporting associations between digital media use and brain regions involved in attention and executive control are similarly preliminary and cannot establish causality[580, 581].

“I found that when I was unemployed… scrolling can be really detrimental to your ability to focus, to concentrate, to maintain your ability to follow a thought process and articulate yourself.”

Mind focus group participant, Youth Voice Network, aged 14–25 Digital environments clearly create short-term demands on attention through interruption, novelty, distraction and the continual possibility of engagement. Whether these effects are sufficient to explain population-level increases in attentional difficulties remains uncertain.

Digital technologies have nevertheless altered the attentional environment in which children develop. These changes may have greater consequences for children with existing difficulties involving attention, executive functioning or emotional regulation (see Chapter 7). They also interact with sleep, educational pressure and reduced opportunities for recovery, rather than operating independently.

10.7.5 Sleep, recovery and constant connectivity

Recovery includes sleep, rest, reflection, recreation and psychological disengagement from everyday demands. Digital environments may influence mental health partly by making this disengagement more difficult.

The strongest evidence concerns sleep. A systematic review of 127 studies involving more than 867,000 adolescents found that digital device use was associated with shorter sleep duration, later bedtimes and poorer sleep quality, with the strongest associations observed for smartphones, social media and problematic internet use[582].

Emotionally arousing content, social interaction and anticipation of messages may make it harder to settle and fall asleep. Notifications and habitual checking can also interrupt sleep after it has begun. Evening exposure to the short-wavelength light emitted by screens may delay the body’s circadian preparation for sleep by suppressing melatonin, the hormone that signals biological night. The relative importance of this light-related mechanism remains uncertain, however, and behavioural factors such as delayed bedtime, emotional arousal and repeated checking may be at least as important[583].

Although causality cannot be established conclusively, sleep remains one of the most consistently supported pathways linking digital environments with poorer emotional wellbeing.

“I barely slept and now I can’t think straight. I’ve got to delete my social media apps. I was just doomscrolling for hours and I know that if I could put the same amount of energy into study I’d be acing my exams.”

Tellmi user, aged 17 The relationship is unlikely to operate in one direction alone. Young people experiencing insomnia, anxiety or low mood may themselves use digital devices late at night to regulate emotions or cope with difficulty falling asleep[584].

Recovery extends beyond sleep. Digital environments are characterised by a degree of continuity that is historically unusual. Social interaction, entertainment, news and peer communication remain available throughout the day and often into the night. Notifications, messaging and continuously updated content can create expectations of responsiveness and blur the boundaries between activity and rest. Digital technologies can provide enjoyment, relaxation, emotional regulation and social connection, but they may also make recovery more difficult when engagement becomes continuous or compulsive.

Digital engagement may also displace activities that support recovery, including physical activity, outdoor recreation, creative activity and unstructured leisure. The evidence for broad displacement is less consistent than the evidence concerning sleep. In particular, there is little evidence that social media has generally replaced close friendships. Adolescents commonly use digital platforms to maintain existing relationships, especially where face-to-face interaction is difficult[585, 586].

The broader significance lies in the interaction between digital environments and the other developmental changes described in this Review, including educational pressures, economic insecurity and reduced community opportunities for recreation. Digital environments may compound these pressures by carrying stimulation, social comparison and expectations of responsiveness into periods that would previously have allowed greater disengagement.

Recovery is therefore one of the more plausible pathways linking digital environments with anxiety, emotional dysregulation, attentional difficulties and reduced wellbeing. The key question is not total time spent online, but whether patterns of engagement – and the urge to check – leave sufficient time for sleep, psychological distance, recreation and recovery.

10.7.6 Belonging, loneliness, connection and bullying

Belonging is a central developmental task of adolescence. Friendships, peer groups and wider social networks increasingly shape identity, self-esteem and emotional wellbeing. Messaging platforms, gaming communities, online interest groups and social media allow young people to maintain friendships, organise their social lives and remain connected, including across distance and periods of transition.

These opportunities may be especially valuable for young people who are geographically isolated, disabled, autistic or members of minority communities. Online environments can provide understanding, support and forms of community that may be difficult to find locally or through face-to-face contact.

“Due to my problems with expressing myself… I am mostly non-verbal, but I can talk to autistic people on Tellmi and I feel that I am part of the community.”

Tellmi user, aged 16 “I had cancer as a teenager… the only way I found support was online. Without those communities my mental health would have been so much worse.”

Mind focus group participant, aged 14–25 These benefits rule out any account of digital environments based solely on harm. Online relationships generally complement rather than replace offline friendships[585, 586].

Belonging involves feeling recognised, valued and securely connected to others. Digital environments can support these experiences while also making exclusion unusually visible. Young people can see how others look, who is together and who appears to have been left out.

Bullying was strongly associated with later anxiety, depression and other mental health difficulties long before social media emerged. Considerable overlap exists between young people bullied online and those bullied offline, with the same child often experiencing both forms of victimisation[587]. What digital environments may change is the reach, persistence and visibility of these experiences. Bullying can continue beyond school hours, be witnessed by much wider audiences, and follow a young person into spaces that would previously have offered respite.

Reviews consistently associate cyberbullying victimisation with anxiety, depression, self-harm and suicidal thoughts[588, 589]. Digital platforms may therefore extend and intensify difficulties already embedded in a young person’s social environment[587].

“Young people now can’t get away from it. If they avoid social media, they’re out of the loop with their friends. It’s a double-edged sword.”

Mind focus group participant, aged 26–35 Digital environments may therefore both strengthen and undermine belonging, sometimes for the same young person. A platform that provides connection, identity and support may also expose that person to comparison, exclusion or harassment. Their effects depend less on the amount of communication than on the quality of relationships, the young person’s social position and the wider environments within which online interaction occurs.

10.7.7 Harmful content and algorithmic amplification

Concern about digital environments centres not only on the time young people spend online, but also on the content they encounter.

Material relating to self-harm, suicide and eating disorders raises particular concern. Repeated or sensationalised exposure to suicidal behaviour can increase risk among

vulnerable people[590]. Digital systems increase the accessibility, scale, repetition and personalisation of such material, and may actively recommend harmful content.

Molly Russell’s death brought these concerns into sharp public focus. The coroner concluded that depression, self-harm and suicide-related content encountered through Instagram and Pinterest contributed to her death[591]. The case illustrated how recommendation systems can repeatedly expose a vulnerable young person to increasingly distressing material.

Audit studies and independent investigations are broadly consistent with this concern. Studies using simulated adolescent accounts have found that initial engagement with self-harm, eating disorder or distress-related material can lead recommendation systems to provide substantially more of the same content[592-594]. These studies cannot establish effects on individual young people, but they do demonstrate the capacity of recommendation systems to create digital environments that users may find difficult to recognise or control.

Eating disorder-related content presents a similar concern. Idealised appearance content is associated with body dissatisfaction, appearance comparison and poorer wellbeing[562, 566]. Some online communities promote restrictive eating, starvation, extreme exercise or rapid weight loss. Risk is not shaped by social media alone: advertising, celebrity culture, television, fashion and wider body ideals also contribute. Renewed thinness ideals and the increasing visibility of weight-loss medication may also have altered the cultural meaning of weight and appetite[595]. Such messages may be particularly harmful during recovery from an eating disorder[596].

A related but distinct concern is the growing volume of mental health and diagnostic content online[597]. Accurate information about ADHD, autism, trauma and other psychological conditions can reduce isolation, give young people language for previously unexplained experiences and encourage appropriate help-seeking. However, much of this content is simplified, decontextualised or inaccurate. Symptom lists, personal testimonies and repeated recommendations can present common experiences as diagnostic evidence, blur important distinctions between conditions and encourage inaccurate self-diagnosis. Repeated exposure may also intensify symptom monitoring and turn a possible explanation into premature certainty. Misinformation may then influence self-care in unhelpful ways, including avoidance, self-medication, use of unproven remedies, changes to prescribed treatment or delays in seeking professional support. The risk arises when authoritative-sounding content is repeatedly promoted without appropriate safeguards for accuracy, context or clinical nuance.

This material is not necessarily harmful in the same way as content promoting suicide, self-harm or starvation. Rather, it influences how young people interpret and make sense of their experiences (see Chapter 11 for discussion of diagnostic narratives, identity formation, cultural understandings of mental health and online meaning-making).

Recommendation systems prioritise material according to predicted engagement. Content that generates strong emotion or repeated interaction can therefore become disproportionately visible.

Most young people exposed to harmful content will not experience serious adverse outcomes. Digital environments are not uniquely responsible for self-harm, suicide or eating disorders. These difficulties are complex and multifactorial. However, digital environments may influence how existing vulnerabilities are expressed, reinforced and sustained. For a relatively small but important group of highly vulnerable young people, repeated and personalised exposure may have important adverse consequences.

10.7.8 Mechanisms linking digital environments to mental health

There is no single pathway linking digital environments and mental health. Instead, a small number of developmental mechanisms recur and may affect different young people in different ways, depending on their vulnerabilities.

Evaluation appears to be among the most important. Digital environments make social comparison unusually visible, continuous and quantifiable. Likes, follower counts, comments and other forms of social feedback provide immediate signals of social approval[598]. Combined with exposure to a much larger range of comparison targets, these environments may intensify evaluative pressures during a developmental period in which self-esteem is especially sensitive to social feedback.

Belonging represents a second important mechanism. Digital technologies create valuable opportunities for connection, particularly for young people who might otherwise experience isolation. At the same time, they may make exclusion, comparison and bullying more visible, altering how belonging is experienced rather than changing its developmental importance.

Recovery is a third recurring theme. The evidence linking digital engagement with sleep disruption is strong. More broadly, constant connectivity may reduce opportunities for psychological disengagement by extending stimulation, social interaction and information exposure into periods that would previously have been reserved for rest.

Recommendation systems, personalised feeds and changing online norms create highly engaging but largely opaque environments. Algorithmic amplification may reinforce concerns, increase repeated exposure and narrow the range of material encountered. Frequent interruptions and continual opportunities for engagement may impede recovery, tolerance of uncertainty and resistance to evaluative pressure.

These demands may be greater for some young people with ADHD, autism or difficulties involving executive functioning, sensory regulation or emotional control. At the same time, the same technologies may support regulation, communication and participation. Their effects therefore depend on both the activity and the individual.

Digital environments are likely to have contributed to worsening mental health among children and young people, although neither causality nor the scale of their contribution can yet be established with confidence. Their rapid expansion coincided with rising adolescent distress. Several associated experiences – including social comparison, body dissatisfaction, sleep disruption, cyberbullying, harmful content and visible exclusion – also correspond to the difficulties that increased most sharply elsewhere in the Review. Correlation does not establish causation, but the convergence of timing, plausible mechanisms and observed outcomes suggests that digital environments have made a meaningful contribution.

The evidence is strongest for particular experiences rather than for digital engagement as a whole. Studies of sleep disruption, appearance-related comparison, cyberbullying and harmful content point in a consistent direction. Total screen time and overall social media use are much weaker explanatory measures, although for some individuals severe disruption to recovery or social relationships may have adverse effects. These aggregate measures flatten very different experiences into a single number. Risk depends on what children encounter, how platforms respond to their behaviour, which activities and relationships are displaced, and the circumstances of the individual child.

This creates a difficult policy challenge because digital technologies have both benefits and risks. The same technologies that expose children to comparison, harassment or harmful content can also provide friendship, information, identity, self- regulation and community. Girls and young women, children with lower self-esteem or existing emotional difficulties, and those exposed to bullying or exclusion may be especially vulnerable. For some children with ADHD or autism, continuous stimulation, interruption and rapidly changing social demands may increase difficulties with attention, regulation and recovery. For others, digital environments

provide communication, predictability and connection that may be harder to find elsewhere. Blunt restrictions risk removing these benefits while leaving the most harmful features of platform design unchanged.

Children born since 2010 are among the first cohorts to spend their entire childhood in a world where smartphones, social media and algorithmically selected content are ordinary features of everyday life. They are often described as digital natives, but familiarity should not be mistaken for mastery. Children are not born knowing how to recognise persuasive design, manage continuous stimulation, evaluate health information, protect sleep or seek help following harmful online experiences. These skills need to be taught and reinforced as technologies continue to evolve.

The response must therefore combine platform accountability with stronger support for children to manage their digital lives. Platforms should be held accountable for harmful design choices, recommendation systems, data access and the repeated promotion of damaging content. Children and young people need practical education about attention, self-regulation, relationships, privacy, misinformation, conflict and recovery, supported by parents, schools and services. Digital environments are neither the sole cause of rising distress nor a neutral backdrop to childhood. They are now part of the conditions in which children develop and should be governed accordingly.

10.8 Changing developmental environments and mental health

10.8.1 Stakeholder views of rising mental health distress and neurodevelopmental conditions

To understand how public stakeholders view the drivers of rising mental health conditions and neurodevelopmental diagnoses, the Review undertook a rapid qualitative analysis of interviews with health and mental health professionals, education and school-based practitioners, researchers and academic experts, youth sector and community organisations, and policy and advocacy organisations. Together, these perspectives provide a broad understanding of how ADHD, autism and mental health conditions are experienced, recognised and responded to across different settings.

Stakeholder interviews highlighted five broad drivers of rising ADHD, autism and mental health-related help-seeking:

  • Rising population-level distress
  • Loss of protective environments
  • Digital environments
  • Changing awareness and identity
  • Structural inequalities

Stakeholders did not view rising ADHD, autism and mental health conditions as being driven by any single factor. Rather, they emphasised the combined effects of social and economic pressures, developmental vulnerability, digital environments and changing awareness.

They reported that more young people are experiencing distress, more individuals are recognising traits in themselves, and more are actively seeking support. At the same time, access to help remains uneven, early support is often limited, and diagnosis continues to function as the principal route into care.

As a result, increasing numbers of people are entering diagnostic pathways, meaning that recorded prevalence reflects both underlying need and the way systems recognise and respond to that need.

10.8.2 Conclusions

The deterioration in children’s and young people’s mental health cannot be traced to a single cause. It has emerged as pressures accumulated and sources of protection weakened. Economic insecurity, housing pressures, educational demands, changing family life, declining community infrastructure, the COVID-19 pandemic and digital technologies have each plausibly contributed. None, on its own, however, explains the timing, breadth and pattern of change described elsewhere in the Review. The rise in recorded neurodevelopmental conditions followed the earlier increase in common mental health conditions, with a marked acceleration in ADHD and autism diagnoses after the pandemic. Chapters 7 and 8 discuss the additional factors contributing to these trends.

Across this complexity, a common pattern emerges. Children are growing up with greater uncertainty about the future, more continuous evaluation and social comparison, fewer opportunities for belonging, and less protected time for rest and recovery. These pressures reinforce one another. Strain at home can reduce a child’s capacity to manage pressure at school. Exclusion can remove both education and a sense of belonging. Digital comparison can extend evaluation beyond the school day, while disrupted sleep reduces opportunities for recovery. Different forms

of social change can therefore interact, even where no single factor is sufficient to explain rising distress.

The burden is not evenly shared. Poverty, discrimination, disability, care experience, family stress and reduced access to educational or community support increase exposure while weakening the resources available to respond. Changes that increase distress across the population can also deepen existing inequalities in functioning, participation and opportunity.

This interaction may be particularly important for neurodevelopmental conditions. ADHD and autism are not created by social change. Their underlying traits are continuously distributed across the population (Chapters 7 and 8). However, the degree of functional impairment depends partly on the relationship between an individual’s needs and the demands of their environment. Educational, social and digital environments that place greater demands on attention, organisation, sensory regulation, flexibility and social navigation may make neurodevelopmental differences both more visible and more impairing[599]. This may influence when difficulties are recognised, how they affect everyday life, and whether available support is sufficient.

Digital environments and educational pressures correspond most closely with the observed rise in mental distress. The widespread adoption of smartphones and social media coincided with increases across socioeconomic groups. Social comparison, body dissatisfaction, sleep disruption and visible social evaluation also correspond to several of the difficulties that increased most sharply. Educational pressures similarly align with the timing of worsening mental health and its greater concentration among girls. Young people consistently describe educational environments as increasingly evaluative and report greater unhappiness with school and schoolwork, together with a weaker sense of belonging.

Other contributors fit the national trend less closely but remain important. Material hardship alone cannot explain increases across the income distribution, although economic and housing insecurity can affect expectations, stability and confidence in the future. The pandemic did not create the deterioration, which was already under way, but it accelerated it and removed education, relationships, routines and support from many children who depended on them. Diagnoses of common mental disorders, and particularly ADHD and autism, also increased sharply following the pandemic. Family relationships and community infrastructure changed more gradually or unevenly. Their importance lies in determining which children are protected, whose difficulties deepen, and where support can be found.

Research cannot yet assign a precise proportion of the increase to each factor. Stronger longitudinal and quasi-experimental studies, better population data, and

secure access to platform evidence are all needed. The inability to produce exact estimates should not become a reason for inaction.

Chapter 11. Changing understanding of mental health and neurodevelopmental conditions

People, particularly children and young people, are experiencing higher levels of distress. Greater awareness of and the wider use of psychological language shape how that distress is recognised, interpreted and responded to, including whether or not people seek help.

For many, this has reduced stigma, improved recognition and self-understanding, and brought previously unrecognised needs into view, including ADHD and autism. However, greater awareness can also shape how experiences are interpreted and reported.

Awareness programmes are active interventions whose effects depend on the message, the individual and the context. Recognition without assessment or support may narrow the search for other explanations, attach certainty to an untested label and leave the person more distressed.

Trusted information, responsible communication and support based on need are essential.

11.1 Recognition, interpretation and help- seeking

Chapter 10 concluded that no single environmental change explains rising mental distress. Economic insecurity, educational pressures, changing family and community life, the COVID-19 pandemic and digital environments have together altered the conditions in which children and young people grow up and develop. Although this Review focuses primarily on children and young people, some of the evidence considered in this chapter draws on studies across the life course.

Over the past two decades, mental health literacy has increased, stigma has reduced in some settings, and psychological language has become part of everyday conversation. This has helped people recognise difficulties and seek support. It has

also influenced what people notice about themselves, how they interpret their experiences and the help they expect.

11.2 Public awareness, mental health and neurodevelopmental literacy

Discussions once largely confined to clinical settings now occur throughout public life. Campaigns, education and anti-stigma programmes have challenged historical under-recognition, when many people concealed distress, managed alone or did not recognise that their difficulties might reflect ADHD or autism.

Awareness has also expanded through workplaces, charities, media, digital applications and, increasingly, AI-based systems. Many young people now encounter mental health and neurodevelopmental information before they reach health services. This can help individuals and families recognise difficulties, often to great benefit, but it also shifts interpretation into settings where evidence, context and alternative explanations may be harder to evaluate.

A systematic review of 15 campaigns across eight countries found improvements in mental health knowledge, attitudes and intentions to seek help[600]. In England, Time to Change was associated with improved knowledge and public attitudes, while Every Mind Matters increased knowledge of anxiety, depression and stress, and confidence in seeking support[601, 602]. Evidence relating specifically to autism and ADHD remains more limited, and some broader gains in public attitudes may now be reversing[27, 603, 604].

Campaign exposure has also been associated with greater disclosure and help- seeking[602, 605]. A Canadian social media campaign was followed by increased use of mental health services, while Australia’s R U OK? campaign increased confidence in supporting others[606, 607].

Most evaluations measure knowledge, attitudes and intentions rather than longer- term mental health outcomes. Increased service use may indicate that barriers have been reduced, which is an intended benefit. Where awareness grows faster than assessment and support, however, greater willingness to seek help may translate into unmet demand.

Greater awareness is an important public health achievement, not simply an explanation for rising demand. The more difficult question is whether awareness only reveals existing difficulties or also shapes how people understand and report them.

11.3 Psychological language, self- understanding and explanatory frameworks

Psychological and neurodevelopmental concepts give people ways to recognise, interpret and communicate experience. They can replace confusion and shame with explanation. They can also influence which experiences are noticed and the meaning attached to them.

Box 11.1. How societies understand distress changes over time Psychological distress and disability are not new, but the language used to understand and explain them changes over time. Neurasthenia grouped fatigue, anxiety and poor concentration as a disorder of industrial modernity before giving way to concepts such as anxiety and depression. Shell shock disappeared as a diagnosis but transformed understanding of psychological trauma. Hysteria once encompassed paralysis, seizures, mutism and loss of sensation without an identifiable physical cause; its features are now distributed across several diagnostic categories.

These histories do not show that contemporary conditions are invented or invalid. Rather, they show that diagnostic categories develop alongside advances in science, institutions and culture, particularly where no biological test exists to confirm them. Hacking described this as the looping effect: classifications describe people, but people also come to understand and express their experiences through those classifications[608, 609]. The experience is real; the available framework influences what is noticed, named and reported.

What distinguishes the present period is the speed and reach with which psychological concepts circulate. Some scholars describe the broadening of concepts to encompass an increasing range of experiences as concept creep[91]. Linguistic studies similarly demonstrate the growing use of psychological and diagnostic language in everyday communication and on social media[245, 610, 611]. Whatever the reasons, young people today have access to a much wider psychological vocabulary than previous generations.

This does not mean that psychological language has displaced real distress. It suggests that explanatory concepts have become more widely available and their boundaries more openly debated. The same term may provide one person with a precise explanation while, for another, collapsing important distinctions.

This vocabulary can bring relief, validation and reduced self-blame. Online communities can reduce isolation and connect people with others who share similar

experiences[612, 613]. Diagnosis in adulthood can also bring substantial relief and greater self-understanding, particularly for autistic people[614].

For people whose difficulties were previously overlooked or misinterpreted, a new explanation can reorder years of experience and open access to adjustments. These benefits are greatest when the explanatory framework accurately reflects the person’s difficulties and is followed by appropriate support.

11.3.1 How awareness shapes self-understanding

Explanatory frameworks are not neutral. They suggest causes, identities, expectations and routes to support. Research over several decades shows that information about psychological conditions can influence how people interpret their own experiences[615-617]. Participants informed about previously overlooked ADHD symptoms were more likely to believe they had ADHD despite reporting no more symptoms than comparison groups[618]. Social media content normalising anxiety similarly increased self-diagnosis without increasing reported symptoms[619].

These findings do not suggest that awareness creates symptoms or that self- identification is usually inaccurate – for depression or insomnia, for example, it is generally broadly accurate. Rather, they show that explanatory frameworks can influence classification, particularly where the experiences described are common or shared across several conditions. These effects are likely to vary according to prior distress, personal circumstances and access to assessment and support.

Experimental evidence also suggests that exposure to mental health content can broaden what young people regard as symptoms, even where reported symptom levels do not change[617-619]. Awareness does not manufacture mental ill health, but it can influence what people notice, how they interpret it and what they report.

Evidence from universal awareness programmes in schools is similarly mixed. The Aware trial within the Department for Education’s Strategies for Safety and Wellbeing programme found greater emotional difficulties at long-term follow-up among pupils receiving an awareness intervention than among controls[620]. Reviews consistently find improvements in knowledge and attitudes, but mixed evidence for mental health outcomes, including no benefit and occasional adverse effects[617, 621,

622]. Awareness may give young people language for distress without providing a solution. Where recognition is not followed by timely assessment and support, difficulties may be more clearly identified but no easier to manage.

The reasons remain uncertain. Higher reported difficulties may reflect better recognition of existing problems, unintended effects of the intervention, or both. These findings indicate that awareness programmes are active interventions rather

than a neutral transfer of information. Their effects depend on the message, the individual and the context.

This is not an argument against awareness. Earlier recognition can reduce shame, encourage help-seeking and prevent difficulties from worsening when effective support follows. The challenge is to retain these benefits while minimising unintended harms[623]. Recognition without assessment or support may narrow the search for alternative explanations, attach certainty to an untested label and leave people more distressed.

11.3.2 Self-identification before assessment

Clinicians report increasing numbers of people arriving for assessment already convinced that they have a particular condition[624]. Most psychiatrists surveyed for this Review also reported substantial growth in self-identification[625]. In a pilot study of 57 young people entering treatment, most believed they had a condition that was not subsequently diagnosed by a clinician, and most reported that social media had contributed to this belief[626]. Interviews with 16 clinicians similarly suggest that wider access to diagnostic language, particularly online, influences how adolescents understand and present their difficulties[627].

This Review uses the term self-identification rather than self-diagnosis. It refers to someone recognising their experiences in a condition before assessment, without assuming either that their conclusion is incorrect or that a diagnosis has been established. Awareness is only part of the explanation. Long waits and limited support leave many people trying to understand their difficulties on their own.

Surveys consistently find substantially more adolescents and young adults identifying with ADHD or autism than reporting a formal diagnosis[14]. Self- identification may lead to an accurate diagnosis, appropriate adjustments or a better understanding of need. It may also prove not to be the best explanation. Evidence remains limited regarding these different outcomes.

Online environments are now central to how many young people understand their experiences. Personal accounts, symptom descriptions and diagnostic explanations can help people recognise difficulties long before assessment. They can also reduce isolation, challenge stereotypes and connect people with others who share similar experiences[612, 613].

Recommendation systems may then repeatedly present similar material, reinforcing an initial interpretation[67]. Repetition can make one explanation appear more certain than the evidence allows (see also the clinician roundtable discussion of self- identification in autism in Chapter 8). Personal testimony provides valuable insight

into lived experience, but it cannot establish whether someone meets diagnostic criteria or whether another explanation may better account for their difficulties.

The risk lies neither in people sharing their experiences nor in young people finding meaning in them. It arises when one person’s account becomes a diagnostic template and is repeatedly reinforced without context or qualification. Some young people may therefore reach assessment with a preferred explanation already established. This is understandable, particularly when professional help is difficult to access, but it may narrow an open exploration of what is causing their difficulties and what support is likely to be most helpful.

11.3.3 Neurodiversity, identity and diagnosis

The concept of neurodiversity, defined in Chapter 4, has reshaped public understanding of autism and ADHD. Emerging from autistic self-advocacy communities in the 1990s, it begins with the proposition that human minds vary naturally[46]. It has since developed into a movement arguing that autism and ADHD should be accommodated as forms of human difference rather than viewed solely as disorders requiring treatment[628, 629]. Its adoption across charities, schools, workplaces and research illustrates how quickly an idea can enter wider culture and reshape self-understanding[47].

The neurodiversity framework allows autism and ADHD to be understood as forms of human variation while recognising that they may also involve substantial disability. For many people, it has reframed lifelong difficulties as difference rather than personal deficit, reducing shame and fostering community[630]. Alongside wider research and advocacy, it has also challenged narrow childhood models that left many girls, women and adults unrecognised[631]. To interpret this shift only as a driver of rising demand would overlook these substantial gains.

“The algorithms on social media can be incredibly accurate. In the year before I sought an autism assessment, I was experiencing worsening mental health, and a crisis felt imminent. I began encountering neurodivergent advocates whose lives appeared to mirror my own. Their content challenged the stereotypes that had held me back from

  • seeking diagnosis sooner However, public awareness has outpaced

public understanding of what disorder and impairment actually look like,

  • and how they translate into everyday life more people may begin to

attribute ordinary human traits or differences to neurodivergence, without understanding the severity, persistence and impairment that distinguish a diagnosable condition.”

Lived Experience Partner Here lies the paradox. Neurodiversity challenges the idea that autism and ADHD should be understood only as disorders, yet formal diagnosis remains one of the principal routes to support. The Equality Act 2010 does not require a formal diagnosis before the duty to make reasonable adjustments can arise, although practice often differs. The neurodiversity movement asks society to accommodate difference; the health system, by contrast, often requires diagnosis before providing treatment, for example for ADHD.

Self-identification often precedes assessment. Where diagnosis is expected to confirm an identity that has already formed, but does not, a different clinical conclusion may be experienced as a rejection of the self[632]. This is Hacking’s looping effect[608, 609] in a contemporary context.

Neurodiversity describes human variation, a social movement and a theoretical framework. In this context, the concept of a spectrum has proved useful because autistic traits are continuously distributed, co-occurrence is common and support needs vary greatly, but it does not remove the diagnostic threshold for disorder[633-

635]. That threshold does not imply that people below it have no genuine difficulties – they may still require support – or that those above it have uniform needs. Rather, it distinguishes continuously distributed traits from a diagnosable pattern of clinically significant impairment. Support should respond to functional need without making diagnosis either meaningless or the only route to help. Diagnostic clarity about disorder and support based on need are distinct and can coexist.

This distinction underlies debate about whether inclusion requires replacing the concept of disorder with one of difference, particularly across a spectrum that includes intellectual disability and profound support needs[636-639]. Views within autistic communities are considerably more diverse than this debate sometimes suggests[640].

University data illustrate a broader shift in the language of mental health. Around one quarter of students report a professional diagnosis, while more than twice as many describe themselves as experiencing ‘mental health problems’[641]. This broader category includes loneliness, homesickness, examination stress and uncertainty about the future. Such experiences can be severe and deserve support, but they do not necessarily indicate a mental disorder. Recognising distress is not the same as diagnosing disorder.

These findings do not suggest that young people are exaggerating or misunderstanding their experiences. Rather, they show that mental health language is now used more widely than formal diagnosis. Distress can be real and require support without meeting the threshold for disorder. Comparable data on self- identified neurodivergence among university students are not yet available.

In summary, greater awareness can identify previously overlooked conditions, but it can also extend diagnostic language to experiences that do not meet clinical thresholds. Long waits and limited support mean that self-identification may become the first, and sometimes the only, route to understanding and help.

11.4 The awareness environment: how psychological ideas spread through society

Analysis of Charity Commission data undertaken for this Review reflects the expansion of mental health-related activity. Between 2003 and 2024, the number of charities referring explicitly to mental health in their objectives increased by approximately 44%, compared with 15% across registered charities overall. Autism- focused organisations more than doubled in number and their combined income almost tripled, while ADHD-focused organisations also grew substantially. These organisations provide vital support. Through education, training, advocacy and peer support, they also help shape public understanding.

This is a process of social diffusion. People interpret information through their own history, circumstances, distress, identity and unmet need. The same information may bring accurate recognition to one person and premature certainty to another. Formal models suggest that as awareness reaches more people, self-identification and referrals that do not lead to diagnosis will also increase, even when the information itself is accurate. Demand may therefore exceed assessment capacity. Awareness should be judged by its overall benefits and risks, not by accuracy or good intentions alone[70].

The recognition of inattentive ADHD illustrates both the benefits and the risks. Research corrected its historic under-recognition among girls and women and was incorporated into professional guidance, education and advocacy[642-644]. This helped many people understand previously overlooked difficulties. However, inattention may also arise from anxiety, learning difficulties and other causes. When the distinction between a characteristic and a diagnosis is lost, useful information can narrow rather than deepen understanding.

Digital platforms accelerate this process. In an analysis of 100 popular ADHD TikTok videos, 55% of the characteristics presented did not align with DSM-5-TR criteria. Nevertheless, 59.6% of the most-liked commenters identified with the experiences shown, often accepting them as evidence of ADHD[645]. Analysis of obsessive- compulsive disorder content similarly identified stereotype amplification and misinformation[646]. At the same time, these platforms can improve recognition of less familiar presentations. Their influence is therefore not neutral.

“ADHD awareness is rising, but what is presented for the most part are

  • the minor irritations We have to ensure that the more difficult and

serious symptoms that aren’t even a little bit entertaining aren’t lost in this more aware society. ADHD can ruin lives, it’s not fashionable and it’s not funny.”

Lived Experience Partner Awareness, advocacy and lived-experience storytelling can reduce shame, challenge neglect and encourage help-seeking. However, information that changes how people understand themselves also carries responsibility. Common characteristics should not be presented as proof of a condition. Alternative explanations and uncertainty should be acknowledged, and serious conditions should not be reduced to relatable quirks. A personal story can illuminate a condition, but it cannot define it. Responsibility is greatest for organisations and platforms that commission, package or amplify this material at scale. Public mental health information should therefore meet standards of evidence, context and accountability proportionate to its influence.

11.5 Environmental change and changing meaning

Chapters 10 and 11 provide complementary explanations for why this Review has found increasing levels of distress among children and young people. Changing environments shape what children and young people experience. Changing systems of understanding shape how those experiences are recognised, interpreted and communicated. Greater distress creates greater need for recognition; greater awareness makes existing distress more visible. These processes reinforce one another.

Language does not create underlying experience. It provides people with ways to understand what they are feeling, communicate it and decide whether to seek help. Recognition is the point at which lived experience becomes visible to health, education and care systems.

Recognition shapes how disability is understood and supported; it does not create profound disability. An autistic child with profound intellectual disability, limited communication and lifelong support needs remains disabled whatever language is used. Social context may determine whether those needs are recognised and met, but not whether they exist.

Distress, recognition, help-seeking and diagnosis are related but distinct. A person may experience distress without understanding it through mental health language. Recognition may lead to help-seeking, but diagnosis is only one possible outcome of professional assessment.

Chapter 12 begins where recognition meets the system. It examines how assessment, diagnostic thresholds, waiting times and access to support determine whether recognition leads to timely help, prolonged uncertainty or increasing pressure for diagnosis.

Chapter 12. How systems shape recognition, diagnosis and support

Rising distress and need are real, but referrals, waiting lists and recorded diagnoses also reflect how systems recognise, assess and record need. Diagnosis – particularly for ADHD, autism and access to educational support – has increasingly become a gateway to help, concentrating demand on assessment services.

Long waiting times, fragmented pathways and unequal local provision can worsen outcomes, increase costs and widen inequalities. Administrative data therefore measure where need meets the system rather than population need directly, and geographical differences may reflect commissioning, capacity and recording practices as well as underlying prevalence.

The response cannot be simply to increase assessment capacity. Faster assessment remains important, but it needs to be part of a system that responds earlier to need through better-connected pathways, and that is judged by how well it meets need rather than by the volume of activity it records.

Four headline figures illustrate the pressures facing the system. In 2023, around one in five children and young people in England screened positive for a probable mental disorder, up from around one in eight in 2017[535]. At the end of 2025, around 254,000 people were waiting for an autism assessment, with nine in ten waiting longer than the recommended maximum[12]. By 2024, mental health services in England received around 17,000 referrals per month for suspected ADHD, an increase from 5,000 in 2019, excluding those referred to community paediatric services for ADHD assessment, although ADHD referrals nevertheless represent a stable proportion of mental health service referrals overall (see Chapter 7, Section 7.3)[7]. NHS spending on independent-sector ADHD services increased by around 252% in just three years, with a projected overspend of £164 million[18].

These figures are often interpreted as a single story of rising need overwhelming services. The rise in distress and need is real, but it is not the whole explanation for increasing service demand. Demand also reflects the interaction between people and systems. Recognition, help-seeking, referral thresholds, waiting times,

diagnostic practice and the availability of support all influence which needs become visible and what is ultimately recorded.

Administrative data do not simply count illness; they record the point at which human experience encounters a system. Read alongside epidemiological evidence, they show not only the needs people bring but also where they enter services, where they wait, and how the organisation of support shapes what becomes visible. The interaction between people and systems is therefore one of the most informative sources available for understanding why demand has changed and where pressures arise.

This chapter follows that journey from recognition and help-seeking (Sections 12.1– 12.3), through the gateway effect of diagnosis (Section 12.4), waiting and access to support (Sections 12.5 and 12.6), to what the resulting data measure (Sections 12.7 and 12.8).

12.1 From need to recognition

Systems can respond only to needs that become visible to them. Children and young people may experience significant distress or impairment for years before anyone identifies it as requiring help. Recognition marks the point at which private experience first becomes visible to families, professionals or services.

Recognition is not diagnosis. Recognising oneself in a description of autism, ADHD, anxiety or depression may prompt further enquiry, but it does not establish the presence of a disorder. Recognition by systems is also unevenly distributed. Women and girls, adults whose difficulties were missed in childhood, and people living in disadvantaged circumstances remain systematically less likely to have their needs recognised (as discussed in Chapters 6, 7, 8 and 17)[156, 647].

Recognition should therefore be understood neither as a measure of prevalence nor as a proxy for diagnosis. As Figure 12.1 illustrates, it is one stage in the process through which population need becomes recorded demand.

Figure 12.1. From need to system demand: how interactions with the system shape our understanding of demand

12.2 From recognition to diagnosis

Many people recognise difficulties without seeking professional advice; others seek assessment but do not receive a diagnosis. Between recognition and diagnosis lies assessment, which is more demanding than public debate often assumes.

Diagnostic criteria provide an essential framework, but they do not substitute for professional judgement. Assessment requires consideration of developmental history, persistence over time, expression across settings, functional impairment, co- occurring conditions and social context[260]. The task is to understand what symptoms mean for the individual in everyday life.

Assessment for ADHD and autism illustrates this complexity. The characteristics of both are continuously distributed across the population, there is no biological test, and the functional impairment that distinguishes disorder from normal variation is context-dependent and can be difficult to judge consistently[212]. Neurodevelopmental and mental health conditions frequently co-occur[79]. Good assessment requires formulation and clinical understanding before classification. Diagnosis should follow understanding, not replace it (see Chapters 5, 7 and 8).

12.3 The many routes in

There is no single pathway into mental health support or neurodevelopmental assessment. People seek help through primary care, schools and colleges, emergency departments, universities, employers, voluntary organisations and, increasingly, through digital information, social media and online communities. Many are seeking an explanation, reassurance or help to remain engaged in education, work or everyday life, rather than a diagnosis.

Multiple entry points create opportunities to recognise difficulties earlier and in settings that people know and trust. However, these routes have evolved rather than been designed as a coherent system. Provision varies between areas according to local commissioning decisions, workforce capacity and the availability of early support[203]. Different entry points understand needs differently and have different options available.

Where a person first seeks support may influence what help they receive, how quickly they receive it and whether they reach specialist services. The challenge is not simply that there are many routes into support, but that they are often poorly connected.

12.4 Why pathways converge on diagnosis

Pathways to support increasingly converge on diagnosis. Accurate and timely diagnosis remains one of the most valuable contributions that clinical services can make. It can guide treatment, make longstanding difficulties more understandable, support appropriate adjustments and, for many people, provide both relief and explanation. Difficulties arise when diagnosis also becomes the principal mechanism for determining access to wider support. A process designed to improve clinical understanding then carries educational, health and social care responsibilities for which it was never intended.

This gateway effect is particularly evident in education and neurodevelopmental support. Where schools cannot meet increasingly complex needs from core budgets, assessment for ADHD or autism may become the most reliable route to additional support[648]. Schools may identify additional needs long before a diagnosis has been made, yet access to support often requires evidence beyond professional observation, increasing pressure to obtain a diagnosis. The number of children with an Education, Health and Care Plan (EHCP) rose by around 11% in a single year[649]. NHS services also remain largely organised around diagnostic pathways, while needs-led alternatives are often less well developed or inconsistently

available[203]. Under these circumstances, seeking a diagnosis is not evidence of inappropriate demand; it is a rational response to the way support is organised.

This gateway effect should not be generalised across all mental health care. It is strongest for ADHD, autism and EHCP-related provision, where diagnosis frequently determines eligibility for support. Across Children and Young People’s Mental Health Services (CYPMHS) more broadly, only a minority of referrals result in a formal diagnosis, and much care remains organised around formulation, distress and risk rather than a named disorder[60]. The argument that follows applies most directly to neurodevelopmental pathways but, given the high level of co-occurrence, is also relevant to many mental health difficulties.

12.5 Living within the system: waiting and fragmentation

For most people, waiting has become a defining feature of seeking support. At the end of 2025, around 254,000 people were waiting for an autism assessment in England, with around nine in ten waiting longer than the 13-week maximum recommended by NICE[12]. For ADHD, the Independent Taskforce reported waits of around two years in many areas, and substantially longer in others[203]. The Review’s own analysis found median waits ranging from under two months to more than 1,000 days depending on where a person lived[650].

During prolonged waits for ADHD assessment, symptoms and impairment may increase, motivation to engage may diminish, and opportunities for early intervention may be lost[72, 651]. National survey evidence suggests a similar relationship between longer waits and worsening mental health. In the Care Quality Commission’s 2024 survey of community mental health service users, the proportion of people reporting that their mental health had deteriorated increased steadily with waiting time, rising from around one quarter of those waiting less than two weeks to more than seven in ten of those waiting longer than six months. These findings should be interpreted with appropriate caution given the survey response rate of 19%[19]. These effects are not evenly distributed. People with greater initial impairment, fewer family or financial resources, or less ability to navigate services are likely to be affected most severely. Long waits may therefore widen inequalities as well as delay support[652].

People understandably adapt to these delays. Some seek private assessment, growing numbers use Right to Choose, and many return repeatedly to their GP while waiting. NHS spending on independent-sector ADHD services rose by around 252% over three years, with a projected overspend of £164 million across the ICBs analysed[18]. One ICB reported a 30-fold increase in spending on autism and ADHD assessment and treatment services over three years, largely driven by private

providers operating through Right to Choose, and now commissions services from 42 private providers[650]. A diagnosis obtained outside the NHS does not necessarily provide continuity of care. Many people return to primary care for prescribing and monitoring without shared-care arrangements in place, so one pathway reconnects with another rather than replacing it, often to the detriment of the patient.

General practice absorbs much of this complexity. GPs help people navigate thresholds, complete documentation, reconnect with services and maintain continuity while they wait. This role has grown largely by default rather than by design and is seldom commissioned or resourced accordingly. Fragmentation is less about the number of services than the weak connections between them. Organisations each see part of the pathway; only children, young people and their families experience the whole journey.

12.6 Why systems behave as they do

Health and care systems do not simply respond to demand. Through the way access is organised, they also shape the demand they subsequently experience. They influence how need becomes visible, how it is interpreted, how quickly it is met and, in some circumstances, whether the experience of seeking help itself becomes an additional source of distress.

The Review has approached this through the perspective of complex adaptive systems. Organisations and professionals act within their own responsibilities, yet together they may generate outcomes that nobody intended. Three mechanisms recur throughout the evidence:

Feedback. As NHS waiting times lengthen, families seek alternatives such as private assessment or Right to Choose provision. These responses alter patterns of spending, commissioning and demand, creating feedback loops in which system behaviour influences subsequent pressures as much as underlying need[18].

Gatekeeping. Where diagnosis determines access to finite resources, demand converges on diagnostic pathways even when diagnosis is not the person’s primary objective. Waiting lists then shape behaviour by influencing where people seek help and what they seek from services.

Path dependency. Differences early in a person’s journey – such as which professional they encounter, which service receives the referral or where they live – can lead to very different experiences later, even among people with similar needs. These differences help explain the substantial geographical variation in waiting times and recorded prevalence identified elsewhere in the Review.

These system effects do not imply failure by any single organisation. Fragmented commissioning can create gaps in oversight. The Review identified providers delivering NHS-funded ADHD assessments without clear commissioning arrangements and others operating without Care Quality Commission registration[18], although further evidence is needed to understand the scale of these arrangements[653].

The broader lesson is that systems often evolve in ways that were never intended. In complex systems, pressures are rarely removed by strengthening one part of the pathway alone; more often, they are displaced elsewhere.

12.7 What administrative data are really measuring

System interactions shape the administrative data used to understand demand. They explain why findings that appear contradictory can nevertheless all be true. Recorded ADHD diagnoses have risen sharply since 2020 yet remain below community prevalence estimates in most age groups[7]. Prescribing has also increased substantially over the same period[654]. Probable mental disorder among children and young people has likewise increased markedly since 2017[535].

As described in Chapter 3, administrative prevalence reflects the point at which need encounters a system. A person must first recognise a need, seek help and reach a service. Administrative data may therefore underestimate need when people remain unrecognised, are still waiting or receive support outside statutory services. Equally, administrative prevalence may increase because recording practices, pathways or service availability change even when underlying prevalence remains stable.

These system effects help explain why neighbouring areas with similar populations may report different diagnosis rates or waiting times. Such differences may reflect variation in local commissioning, provider availability and recording practices as well as differences in population health and help-seeking.

Routine data nevertheless remain indispensable. They are among the best sources available for understanding how systems function: where people enter services, where they wait or leave, how resources flow and where pressures accumulate. Administrative data should not, however, be interpreted as direct measures of population need. Epidemiological surveys and administrative data answer different questions.

12.8 Synthesis: from need to demand

The evidence shows that the pressures reflected in referrals, waiting lists and recorded diagnoses arise from both increasing need and the way systems are organised. Where support is tied to diagnosis, demand accumulates at diagnostic gateways. Where pathways are fragmented, children, young people and families experience delay, duplication and uneven access, which may itself worsen outcomes.

The response cannot be simply to increase assessment capacity. Faster assessment remains important, but it needs to be part of a system that responds earlier to need, reduces unwarranted variation and maintains continuity across services, while protecting timely diagnosis, evidence-based treatment and specialist care for those who need them most (see Chapter 20, Recommendations 2 and 3).

Section Four: How well is the current system set up to respond?

Chapter 13. Education, SEND and participation

Nurseries, schools, colleges and universities play a central role in identifying mental health and neurodevelopmental needs. Chapter 10 describes how educational environments influence mental health and wellbeing, and identifies education as one of the strongest influences on how children’s lives subsequently unfold.

Education cannot meet rising and increasingly complex needs alone, nor can it continue to rely on clinical assessment and diagnosis as the principal gateway to support. The Equality Act duty to make reasonable adjustments is triggered by substantial functional impact, and an EHCP describes barriers to participation and the provision needed to overcome them. Neither framework requires diagnosis. In practice, however, both have unintentionally increased demand for clinical assessment, widened inequalities between families who can and cannot navigate systems, and delayed support until difficulties are formally documented.

The policy implication is that education should move towards needs-led assessment and provision, supported by integrated neighbourhood health systems that provide timely access to coordinated multidisciplinary advice with expertise in neurodevelopmental and mental health conditions. Children and young people with complex and enduring needs should retain a clear individual entitlement to the specialist provision they need, on which families can rely. That entitlement is distinct from the Equality Act duty to make reasonable adjustments, which applies to every disabled child irrespective of diagnosis or plan.

13.1 Why education is central

For most children and young people, education is the first public system in which difficulties become visible and concerns are raised. Struggles may be reflected in poor attendance, challenging behaviour, low attainment, difficulties with friendships or emotional distress. Nurseries, schools, colleges and universities are much more than places of learning: they are settings in which participation is established and resilience develops through the ordinary experiences of growing up. Educational participation is one of the strongest influences on later participation in adult life and is closely associated with subsequent employment, physical and mental health,

economic independence and wellbeing across the life course. Chapter 10 considered the evidence that educational environments may contribute to common mental health difficulties. This chapter instead examines education as the system through which need becomes visible, and through which timely, flexible support can make a difference.

The Good Childhood Report 2025[446] household surveys identify school as the aspect of life with which children and young people aged 10–17 years are most commonly unhappy. Similarly, the latest Understanding Society data[655] show a marked decline in happiness with school among children aged 10–15.

Figure 13.1 Understanding Society survey of children’s (aged 10 to 15) happiness with school and schoolwork

School School work

Mean score (1 = not at all happy, 7 = completely happy) Source: Understanding Society Current longitudinal evidence suggests that the relationship between wellbeing and educational participation is bidirectional. Chapter 10 reviewed evidence that educational environments can contribute to emotional distress, while longitudinal studies also show that emotional and behavioural difficulties predict poorer attendance, lower attainment and higher rates of exclusion, even after adjustment for socioeconomic disadvantage and special educational needs. Rising internalising symptoms predict later reductions in school connectedness, whereas changes in attendance alone do not consistently predict subsequent emotional difficulties[656-659].

Attendance is one of the clearest indicators of how a child is faring, although prolonged absence may itself deepen the difficulties that first prompted it. Illness- related absence has risen sharply since the pandemic, from 2.5% of possible sessions in 2018/19 to 3.46% in 2023/24, easing to 3.22% in 2024/25[660]. Recording codes do not distinguish mental from physical illness, but survey evidence indicates that anxiety and mental health now account for a substantial and growing share of the reasons parents and pupils give (30% at key stage 4 against 48% for physical illness)[661]. The relationship between exclusion and poor mental health is also bidirectional: children who are struggling are more likely to be excluded, while exclusion is associated with subsequent deterioration in mental health[468]. A recent Department for Education survey found that, among pupils who had missed school, around one quarter cited anxiety or mental health, and around one in ten a neurodevelopmental condition. Most who cited mental health difficulties referred to worries about school itself – including other pupils’ behaviour, lessons, examinations and relationships with teachers – rather than to a diagnosed condition[662].

Another form of non-attendance is elective home education (EHE), which has continued to increase. Between autumn 2024 and autumn 2025, the number of children in EHE rose from 111,700 to 126,000, representing 1.5% of children aged 5–16 years[663]. Many families describe it as a way of escaping bullying or other sources of poor mental health, while offering greater flexibility and an opportunity to respond to their child’s particular strengths and needs. However, it also inevitably reduces opportunities to establish and maintain peer relationships. Most parents will not have formal teaching experience, and home education can place substantial financial pressures on families.

The Welsh data-linkage study linked absence and exclusion specifically with neurodevelopmental conditions, mental disorders and self-harm[664]. The drivers of absence are transactional and may lie partly within the educational environment itself. While the priority is a supported return to education rather than prolonged absence, participation has therapeutic value only when attendance is positive – when it promotes belonging and psychological safety, provides access to trusted adults, and reduces in-school sources of distress such as bullying and punitive disciplinary practices[665]. Returning a child to an unchanged environment that contributed to the original distress is unlikely to help and may cause further harm. Supported participation is therefore more important than attendance alone.

Attitudes towards attendance have also changed. Around one third of parents report being more relaxed about attendance since the pandemic, but this should not be interpreted simply as permissive parenting[666, 667]. The same evidence shows that around one third of parents felt their child was unsupported when anxious about attending school, while one in five reported that their child refused to attend despite

wanting them to[666, 667]. The picture is therefore fundamentally relational. Practices such as requiring medical evidence from the first day of absence may erode trust and encourage families to seek clinical documentation for distress that might otherwise have been managed within education, contributing to increasing demand for assessment. Recent statutory guidance appropriately emphasises a support-first approach[668]. The aim should be shared responsibility supported by the wider system.

“My son has not been to school since last year cos he can’t leave the house. His ADHD is so bad that he has a panic attack if he even looks at the front door. I feel so lonely, and trapped cos I worry about leaving him home alone. I really want him to get help but I’m on benefits and school can’t help if he’s not there. I want to give up.”

Tellmi user, aged 38 A strong sense of belonging at school – feeling accepted, included and valued – is consistently associated with lower levels of anxiety and depression and buffers the effects of childhood adversity into adulthood[669, 670]. Conversely, bullying, chronic evaluation and punitive disciplinary environments are associated with poorer mental health, while perceived academic pressure during adolescence independently predicts later depressive symptoms[671]. The conditions that support learning and participation are therefore also the conditions that support good mental health.

For many young people, a trusted member of school staff is the first person they approach outside family and friends, and school staff are consulted more often than formal mental health services[535, 672]. These adults see children continuously rather than episodically and are well placed to recognise emerging need before it reaches clinical thresholds.

For some children and families, however, education is also where support proves hardest to obtain. Difficulties that are masked in the classroom, or that emerge mainly at home, may go unrecognised, and families who raise concerns are not always believed when a child appears to be coping in school. Parent-led organisations such as Not Fine in School and Define Fine have documented how often support is withheld because a child is judged to be ‘fine in school’ and how damaging it can be when families approach trusted professionals but do not feel heard.

Schools fulfil these responsibilities under considerable pressure. They are expected to respond to needs that originate beyond the school gates while working within demanding accountability frameworks, constrained resources and rising levels of increasingly complex need. Emerging evidence suggests increasing stress, burnout

and poorer mental health among teachers and school leaders, particularly in secondary education[673, 674]. Better teacher wellbeing is associated with better pupil wellbeing, while teacher distress is associated with higher levels of pupil distress, partly through its effects on teacher–pupil relationships[675]. Efforts to improve children’s mental health therefore cannot be separated from efforts to support the adults who teach and care for them.

Finally, the student population has become increasingly complex. Special educational needs, mental health conditions and neurodevelopmental differences overlap extensively (Chapter 9), while communication difficulties – among the most common forms of special educational needs and disabilities (SEND) – remain one of the least recognised influences on educational outcomes[372]. When a child’s participation begins to decline, it is therefore important to understand how multiple interacting factors, operating within a particular educational and social context, affect that child’s ability to learn, participate and flourish.

13.2 Diagnosis, identified need and access to support

This Review finds that diagnosis has become one of the principal routes through which children and young people obtain educational support. This is not, however, what the legal framework intended. Under the Equality Act 2010[50], reasonable adjustments should be based on functional need rather than diagnostic status. Similarly, the SEND Code of Practice is organised around identifying and responding to educational need rather than diagnostic category. Education, Health and Care Plans (EHCPs) describe the barriers to educational participation and the provision required to overcome them; there is no legal requirement for a diagnosis in order to access an EHCP. The graduated approach described in the SEND Code of Practice – assess, plan, do, review – was explicitly designed to identify emerging needs and provide support before specialist assessment becomes necessary. In practice, however, diagnosis is often regarded as the only way to obtain a fair, objective and robust assessment of need.

This has understandable consequences for everyone involved. Families seek diagnostic assessment first and foremost to understand what their child is experiencing, but increasingly also to secure appropriate support. Schools encourage referral in order to obtain the provision a child requires. Clinicians increasingly find themselves providing evidence that extends beyond clinical decision-making into educational planning and wider administrative processes.

“After years of pushing, Sophie finally received an EHCP, which at last allowed appropriate support to be put in place at school. Unfortunately, this

only arrived in Year 11, as she was approaching her GCSEs… Nevertheless, the EHCP represented a kind of passport – one that finally opened the door to the additional support she needed at school and beyond.”

Parent case study, Tellmi These arrangements may also unintentionally widen inequalities. Families with greater knowledge of education and healthcare systems, stronger advocacy skills, or the financial means to obtain independent assessment are better placed to secure a diagnosis and navigate administrative processes. Children with comparable needs, whose families have fewer resources, may therefore experience longer delays before receiving appropriate support. This pattern is consistent with geographical and institutional variation in SEND provision[676, 677].

Nevertheless, diagnosis remains a fundamental component of good clinical care. It guides treatment, supports specialist intervention and facilitates communication between professionals. For many children and families, it also provides an explanation that makes sense of long-standing difficulties and reduces self-blame.

The challenge is therefore not to weaken the role of diagnosis, but to restore it to its proper place within education. At the same time, educational support should be available without waiting for diagnostic certainty. How this might be achieved – while ensuring that those with complex and enduring needs retain a clear individual entitlement to specialist provision, and making fuller use of the reasonable- adjustments duty that already applies to every disabled child without a plan or diagnosis – forms the central argument of the remainder of this chapter.

13.3 What SEND currently gets right and wrong

The SEND system occupies a distinctive place within education, and two different kinds of legal protection operate within it. The Equality Act 2010 places a duty on schools to make reasonable adjustments for any disabled child, based on the substantial impact of an impairment rather than on a diagnosis or a plan. Individual statutory protection does something narrower: it specifies the specialist support a particular child with complex or enduring needs should receive. At present this is done through an EHCP, under the reforms proposed in the Schools White Paper, the individual entitlement would attach to the provision set out in a Specialist Provision Package. The two are often spoken of together as SEND rights, but they differ in whom they cover and what they guarantee and treating them as one has reinforced the assumption that a plan is the only route to support. For many children with complex or enduring needs, individual statutory provision has made the difference

between exclusion and meaningful participation in education. For many more, the reasonable-adjustments duty, applied well, would achieve the same without it.

Demand for SEND provision has increased dramatically. More than 1.8 million pupils (14.8%) in England are now identified as having special educational needs, while 718,800 children (6%) hold an EHCP – the largest annual increase since plans were introduced in 2014 and around 140% higher than in 2015[678, 679]. Growth has substantially outpaced the capacity of the system to respond. Of the 105,300 statutory assessments completed during 2024, 93.6% resulted in an EHCP, yet fewer than half of newly issued plans were completed within the statutory 20-week timescale[678].

The Review recognises that the pressures on teachers are substantial and our own survey of teachers confirmed this. Some expert teachers in that survey saw diagnoses as sometimes unwarranted.

“I have seen students with unnecessary diagnoses whose educational outcomes have been harmed by the belief that they have a disability.”

Specialist SEND Teacher Autism has become the most common primary need recorded within EHCPs, while speech, language and communication needs remain the most common category among children receiving SEND Support, followed by social, emotional and mental health needs, moderate learning difficulty and specific learning difficulties[678].

Figure 13.2 Percentage of total pupils on school rolls with SEND, 2010–2025 Source: DfE public data These educational categories only partially overlap with the diagnostic classifications used in healthcare. The same child may therefore be described differently depending on whether they are viewed through educational or healthcare systems, illustrating one of the central themes of this Review: different systems often describe the same underlying difficulties in different ways.

The rapid growth in demand has no single explanation. Improved recognition, increasing complexity of presentation and the organisation of resources all contribute. Schools often require statutory plans before they can access specialist resources beyond their core budgets. Families and schools therefore respond rationally by seeking the diagnoses most likely to secure provision. The data also reveal inequalities. Boys continue to predominate among those with an EHCP, although the gap between boys and girls has narrowed[678]. More strikingly, almost 44% of children with an EHCP and 39% of those receiving SEND Support are eligible for free school meals, compared with just over 22% of pupils without identified SEND[678].

The Public Accounts Committee concluded that ‘parents increasingly seek EHCPs because SEND Support cannot always be relied upon’[679]. The same report also highlighted examples of highly inclusive schools where strong early provision reduced reliance on statutory plans. The substantial variation between local authorities and schools suggests that institutional practices influence recorded SEND need[676, 677].

The consequences of inadequate support are profound. Young people with SEND are substantially more likely to become NEET (not in education, employment or training) – 22% compared with 7% of their peers without SEND[308] (see Chapter 14 for evidence on inequalities in access to support and the role of Education, Health and Care Plans). Conversely, almost two thirds of young people who subsequently became NEET had ever previously been identified as having SEND. They were also more likely to experience poverty and contact with social care, suggesting that systems continue to struggle to translate early recognition into sustained participation.

Yet despite increases in high-needs funding – reaching £10.7 billion in 2024–25, a 58% real-terms increase over the previous decade – there has been no consistent improvement in educational or employment outcomes since 2019[680]. Children receiving SEND Support finish Reception around one year behind their peers, with the gap widening to around two years by the end of secondary school[681]. Attainment gaps remain wide and vary substantially between local authorities, representing another postcode lottery[682]. The Public Accounts Committee concluded that the system does not currently represent value for money[679].

However, these findings do not necessarily indicate that SEND provision is ineffective. Children receiving SEND support have, by definition, greater underlying needs[308]. Robust evidence regarding the effectiveness of SEND provision remains limited, making evaluation an essential component of future reform.

Special educational needs rarely occur in isolation. Internationally, around one quarter of autistic children also have ADHD, while between 50% and 75% of people with autism or ADHD have at least one common mental disorder (see Chapter 9). Language and communication needs, meanwhile, cut across almost every SEND category[372, 683]. The concentration of these difficulties among the most vulnerable children is striking. Language and communication needs are present in around 60% of looked-after children and 71% of children in the youth justice system[364, 684]. The challenge for schools is that diagnostic labels rarely map directly on to need or provision. Children with the same diagnosis may require very different forms of support.

Current SEND reform seeks to address these challenges. Earlier, graduated and needs-led support should be strengthened so that more children can participate successfully within mainstream education, reducing both educational disadvantage and pressure on diagnostic pathways[685].

13.4 The role of schools in early identification and support

Schools are well placed to identify emerging needs and initiate support. They see children daily over many years, across learning, friendships, behaviour and emotional wellbeing. However, schools are most effective when they operate as part of a wider system, with timely access to specialist advice and support rather than working in isolation. Equally, not every whole-school intervention improves outcomes. Some are effective, some make little difference, and a small number may inadvertently cause harm. Understanding what schools can realistically achieve, and where specialist services remain essential, is central to designing proportionate systems of support.

“A lot of people in my family have autism, including my dad and my brother, and my nephew so mine should have picked up when I was in primary. In my school they don’t offer any help to the younger years. It’s only when you’re closer to year nine or ten and they think you’ll fail your GCSEs, that they start to give people the help that they need, but they will only do it if you have had a formal diagnosis.”

Tellmi user, aged 17

13.4.1 Early identification: building on pre-school surveillance

For many children, important indicators of future need are already available before they enter the classroom. For others, difficulties emerge only as developmental demands increase.

Preterm birth provides a good example. Around one in five children born very preterm (≤33 weeks) or with very low birthweight show the characteristic ‘preterm phenotype’. Meta-analyses demonstrate increased risks of educational difficulties, attentional and executive-functioning problems, and internalising symptoms that become more pronounced with increasing prematurity[686]. Longitudinal evidence from the Born in Bradford study suggests that entering school according to chronological rather than corrected gestational age may compound these educational disadvantages. Greater flexibility in school-entry decisions where developmental readiness is uncertain may therefore mitigate some of these risks[687].

Screening for early learning needs is challenging. Universal screening tools identify many children who will subsequently require support, but they neither identify all children with later difficulties nor avoid over-identifying some who do not have enduring needs. For example, the ASQ-3 administered at two years predicts later developmental outcomes, but identifies only around one quarter of the approximately three in ten children in England who do not achieve a Good Level of Development on the Early Years Foundation Stage Profile, performing best for children with more severe developmental difficulties[688, 689]. Screening for common mental health difficulties, particularly anxiety, also shows promising evidence when integrated with effective support, as demonstrated in the iCATS-i2i trial[690].

Developmental surveillance – including information from health visiting, neonatal services and the early years – should therefore be understood as the beginning of a continuous process extending throughout a child’s educational journey, rather than as a definitive assessment. As developmental expectations increase, new difficulties emerge while others become more apparent.

13.4.2 A layered model: whole-school culture and targeted intervention for mental health and wellbeing

The foundation of effective support is a school culture characterised by positive relationships, inclusion, psychological safety and a strong sense of belonging. These features are consistently associated with better mental health, improved behaviour and greater educational engagement[691-694]. Longitudinal studies indicate that pupils’ sense of safety at school is associated with subsequent improvements in wellbeing and attendance[695]. Schools with more inclusive cultures, reflected in lower exclusion rates, also appear less likely to lose young people from education, reducing later risk of becoming NEET.

Teachers occupy a distinctive role, but they are not mental health professionals, nor should they be expected to become them. Nevertheless, the OxWell Student Survey demonstrates that school staff are among the most frequently accessed sources of support outside family and friends, and are rated as helpful by around 82% of pupils who seek their support[14, 696].

“In my school I received support from the receptionists and whenever I felt overwhelmed or upset, I could sit in the medical nurse’s room to calm down in a quiet environment. In college I do not get much support, specifically for me.”

Tellmi user, aged 16 Whole-school (universal) approaches have an important population role because they reach children experiencing mild or moderate distress who are unlikely ever to meet thresholds for specialist services. Consequently, even relatively modest individual benefits may translate into meaningful improvements across the school population[697].

The evidence for both universal and targeted programmes is encouraging but nuanced. The strongest UK studies are summarised in Table 13.1. Overall, interventions based on cognitive behavioural principles generally outperform mindfulness-based approaches; programmes tend to be more effective for internalising difficulties (depression, anxiety, social withdrawal and low self-esteem) than for externalising difficulties (aggression and rule-breaking); practitioner- delivered programmes show a small advantage over teacher-delivered ones; targeted approaches produce larger effects and are more likely to meet the needs of children experiencing current difficulties; and implementation quality matters. Leadership, staff training and fidelity to programme design are at least as important as the intervention itself[622, 698].

Table 13.1. Summary of evidence for whole-school and targeted interventions

Study/interventionLevelDesignKey finding
Learning Together (INCLUSIVE)[691]Universal/ whole-schoolCluster RCT, 40 secondary schoolsReduced bullying victimisation; improved mental wellbeing and health behaviours
Strategies for Safety and Wellbeing[699]Universal, teacherdeliveredEvaluation, primary & secondaryImproved help-seeking; success depended on leadership and implementation fidelity
DISCOVER workshops (BESST)[700]Targeted, sixth formRCTImproved emotional outcomes; cost-effective at scale
MYCATS[701]Targeted, primaryCluster RCT, 95 schoolsLower anxiety symptoms and impact (disorder-rate difference not significant)
ICATS-i2i[690]Targeted, primaryCluster RCT, 84 schoolsLower anxiety symptoms and significant impact at two year follow-up
MYRIAD (mindfulness)[702]Universal, secondaryCluster RCT, 85 schoolsNo overall benefit; harm signal among higher-risk pupils

Importantly, some whole school-based interventions do not improve outcomes and some may unintentionally cause harm. Myriad, a cluster-randomised trial involving 85 secondary schools, found that universal mindfulness training produced no overall benefit compared with usual provision and, among pupils at greater risk of mental health problems, was associated with slightly poorer wellbeing and a small increase in depression risk one year later[702]. This was attributed, in part, to the possibility that

asking adolescents already experiencing difficulties to focus closely on their emotions in a classroom setting, without individual support, may heighten rather than relieve distress[703]. More broadly, systematic review evidence suggests that around 9% of universal programmes report at least one adverse outcome, rising to almost one third among the highest-quality evaluations[621]. The implication is not that schools should do more, but that they should do less of what does not work — commissioning universal programmes only where benefit is evidenced, evaluating them for harm as well as for gain, and directing the resource released towards selective and indicated approaches.

13.4.3 Mental Health Support Teams

Mental Health Support Teams (MHSTs) represent one of the most important recent developments in education (see Chapter 15). Introduced following the 2017 Green Paper, they have created a new workforce of Education Mental Health Practitioners (EMHPs) working alongside schools and colleges. Coverage now extends to 60% of pupils in England, with national expansion intended to reach all pupils by 2029/30.

MHSTs provide brief, evidence-based interventions, predominantly informed by cognitive behavioural principles, for children and young people experiencing mild-to- moderate anxiety, low mood and behavioural difficulties. For younger children this support is frequently delivered through parents and carers. MHSTs also contribute to whole-school approaches to wellbeing, strengthen links with wider services and help schools recognise when specialist support is required.

As with many national programmes, implementation has varied, but the emerging evidence is encouraging. Early evaluations found that MHSTs were well received, and around 70% of staff reported that children were seen within four weeks of referral. MHSTs appeared to have greatest impact in supporting individual children with mental health needs rather than through whole-school activity[704]. A large real- world evaluation involving observational data from more than 26,000 children and young people across England found significant improvements in anxiety, depression and the personalised goals identified by children and families. Importantly, interventions were cost-efficient, typically involving only two to eight sessions delivered largely by trainees and newly qualified practitioners[705]. Reliable improvement occurred in 47% of children on symptom measures and 56% on goal- based outcomes, while more than half of those above clinical thresholds at the start of treatment no longer met those thresholds by the end.

Recent examples of innovative practice also demonstrate how MHSTs can increase capacity and accessibility through digitally supported treatment delivery[706], including delivery through school pastoral support workers who are trained and supervised by

MHST staff. This approach builds on existing trusted relationships within schools while strengthening the skills of school staff.

“I want to acknowledge the hard work of our school Emotional Literacy Support Assistant (ELSA), who made me feel safe and confident throughout each module. I felt able to be honest without judgement – something that is rare as a SEND parent. I genuinely believe that this type of intervention… could change the trajectory for so many children who are currently living with selective mutism – many in crisis, with nowhere to turn. Early intervention is often all they need.”

West Sussex WOWSI parent MHSTs are also well placed to overcome barriers to accessing effective support for common mental health difficulties, for example, through integrated screening linked directly to intervention[690]. A recent randomised controlled trial involving 84 primary schools in England evaluated universal screening for anxiety combined with a targeted parent-delivered CBT programme delivered online with remote therapist support. This integrated approach produced significant reductions in child anxiety together with improvements across a range of wider outcomes, including teacher- rated impact of anxiety within the classroom, which were maintained for at least two years[690].

“My daughter was 8 years old when we were offered help and I can honestly say it completely changed our lives… Although her fears and worries were

  • really impacting both of our lives I don’t think I would have sought help I

thought it would be unlikely that we would ever get seen given the pressure on services… but after completing a questionnaire, my responses showed she was struggling and I was offered some help. I was so excited when I got

  • that letter I feel like I have a different child now. She went from not wanting

to be apart from me at all to being able to sleep in her own bed, go on sleepovers, and go on school trips and dance residentials. That wouldn’t have happened without the help we received, and I am so glad that we got it when we did, before the teenage years. She is now a 13 year-old who is unstoppable. She is confident and wants to try everything, and I feel really positive about her mental health for the future.”

iCATS-i2i study participant These findings suggest that MHSTs can deliver meaningful, cost-effective improvements at scale, although further robust trial evidence remains important. Particularly encouraging are early indications that children with more complex presentations, including autism, may also benefit.

The evidence also clarifies the limits of this model. MHSTs were designed primarily for children with mild-to-moderate difficulties, and evaluations consistently identify an important gap for children whose needs exceed this level but who do not meet thresholds for specialist services. Concerns have also been raised about limited specialist SEND training, workforce shortages and difficulties retaining trained staff[707]. MHSTs are therefore one important component of a wider system of support.

13.4.4 What schools cannot do alone

Schools can identify emerging difficulties earlier than almost any other public service and can respond effectively to many of them. They cannot, however, be expected to meet all children’s needs independently. The most successful school-based systems are those that work in partnership with families and are embedded within neighbourhood networks providing timely access to specialist advice, enabling children to remain successfully supported within their local schools wherever possible.

Language and communication difficulties provide a good example. They are highly prevalent among children with SEND and frequently underlie both educational and emotional difficulties, with language skills often around three years behind those of peers[708]. Universal and targeted programmes such as the Nuffield Early Language Intervention (NELI) produce modest improvements in language at scale, but appear to confer smaller benefits for children with the greatest needs, potentially widening inequalities[709]. Language interventions are more effective when delivered by trained specialist assistants working under clinical supervision than by school staff alone because implementation is more consistent[710, 711].

Mental health interventions are also commonly verbally mediated and therefore need to recognise the particular needs of neurodivergent children. Masking among autistic children, which educational environments may inadvertently encourage, is associated with anxiety, depression and autistic burnout[712]. Co-produced research consistently highlights the importance of personalised, neurodiversity-informed approaches, while randomised evidence demonstrates that CBT adapted specifically for autistic children produces greater reductions in anxiety than standard CBT[421]. Specialist expertise therefore adds value by helping the adults around a child recognise communication needs, sensory differences and the need for predictability, and by adapting support accordingly.

Effective early help also depends on shared understanding between families and schools. Parents and carers usually know their child best, and what they observe at home constitutes important evidence rather than claims to be verified against classroom behaviour. The long-term aim should therefore be to strengthen schools

by embedding them within neighbourhood systems where specialist knowledge is readily available.

13.5 Behaviour, neurodevelopmental conditions and adversity

Behaviour that disrupts learning, relationships or classroom routines is often the first visible sign that a child is struggling and frequently prompts a teacher or parent to take action. Similar patterns of behaviour may arise from neurodevelopmental differences, emotional distress, language or learning difficulties, adversity at home, physical ill health, or, more commonly, some combination of these.

Managing behaviour is not straightforward. Schools do it while meeting attainment expectations, preparing for inspection, and carrying workload and staffing pressures, and a classroom needs order if any child is to learn. The harder task is recognising when behaviour signals an unmet need – something schools do best when specialist advice is readily available to them. The accounts of young people show why this matters.

“For much of my time at school, my behaviour was taken at face value, rather than anyone exploring what was happening beneath the surface. Looking back, much of it was a cry for help. Lashing out was the only way I knew to communicate the distress I was feeling. If people had looked beyond the behaviour itself, they might have recognised the unmet emotional needs driving it.”

Lived Experience Partner Behaviour emerges from the interaction between a child and their environment – family, peers, classroom, teaching and the wider community – rather than from the child alone[713]. Changes to the environment, including consistent routines, supportive relationships, clearer communication and thoughtful classroom adaptations, can therefore improve behaviour substantially without changing the child’s underlying characteristics.

Behavioural difficulties frequently precede some of the most damaging outcomes in a child’s education. Suspensions and permanent exclusions have risen sharply, from 438,265 suspension episodes and 7,894 permanent exclusions in 2018–19 to 913,000 suspensions and 9,906 permanent exclusions in 2024-25[714]. The permanent exclusion rate for pupils with SEND is 0.37, six times higher than the rate for pupils with no identified SEND, with nearly 1,200 excluded pupils having an EHCP in place.

Excluded children collectively lose more than 32 million days of learning each year, and those affected are disproportionately likely to be disadvantaged, have SEND or already be struggling with their mental health. Longitudinal analysis of the MHCYP found that exclusion predicts later psychiatric disorder even after adjustment for earlier difficulties[692], while linked population data from Wales associate exclusion with subsequent depression, self-harm, substance misuse and poorer educational outcomes[338]. The balance between the two directions varies by age and sex. In the ALSPAC birth cohort, boys excluded during primary school already showed elevated difficulties at age three, with the gap widening as they moved through school, whereas for girls exclusion by the age of 16 was followed by deteriorating mental health[715]. For some children exclusion is the visible end of a long-established difficulty; for others it marks the beginning of one. The most robust attempt to isolate the effect of exclusion itself points the same way. Applying inverse probability of treatment weighting to the Next Steps cohort, and adjusting for prior health and a wide range of earlier characteristics, childhood exclusion was associated with worse psychological distress, life satisfaction, sleep and general health at age 25[716]. Exclusion is not only a marker of difficulty; it appears to add to it. Attendance and poor mental health are self-reinforcing: poor mental health increases the risk of absence, while prolonged absence may in turn worsen mental health[467].

“I missed all of year ten because of my anxiety. I think schools should have a way of teaching students what would happen if something like panic attacks start happening or like what happened to me, or just a place where you could go rather than hiding away from it at home and not telling anyone because you feel like nobody’s going to listen.”

Tellmi user, aged 16 ADHD illustrates this pattern. Among 5–15-year-olds in the 2017 survey, 15% of children with ADHD had been excluded at least once, compared with fewer than 1% of children with no disorder, and permanent exclusion was around 16 times more likely[143]. Exclusion and low participation in school clubs were also more common among children with ADHD than among children whose difficulties were primarily emotional, suggesting that schools may be less well equipped to recognise externalising behaviour as an expression of need than quieter, internalising difficulties.

Disciplinary behaviour-management strategies can increase order and visible compliance, but they do not necessarily address the underlying causes of behaviour. Pupils and many staff favour approaches that seek to understand the causes of behaviour while retaining a proportionate framework of sanctions[717].

Engaging with the complex needs underlying the most challenging behaviour requires specialist training and support that MHSTs do not currently provide. Under the new SEND arrangements, ideally supported through neighbourhood hubs applying the ‘Expert at Hand’ principle set out in this Review (see Chapter 19), schools and families will need access to additional mental health and neurodevelopmental expertise. It is critical that SEND reform sets out a clear strategy and professional support for schools so that they can understand why behavioural difficulties have emerged and respond proportionately, reducing the risk of exclusion.

It is particularly encouraging that this approach converges with the work of the Neurodivergence Task and Finish Working Group commissioned by the Department for Education, which became available as this Review went to press[718]. Its conclusions closely reflect a central theme of this Review, in its words, ‘moving beyond both diagnostic and deficit-based models to offer a coherent conceptual foundation for a genuinely needs-led approach to educational support’.

13.6 How education shapes demand

Systems influence demand (see Chapters 11 and 12), and schools play an active role in shaping how need is recognised, interpreted, documented and ultimately translated into demand for both educational support and specialist health services.

Around one fifth of accepted referrals to child and adolescent mental health services originate within education[696]. However, the same evidence also suggests that substantial need remains unrecognised by schools, consistent with families’ accounts of having to press for recognition themselves.

The process of identifying need may itself influence the pathway that follows. Longitudinal linkage undertaken by the Education Policy Institute found that children recognised by their school as having social, emotional and mental health needs were more than twice as likely to receive specialist mental health support as children with comparable baseline difficulties who were not formally identified[719].

The Nuffield Foundation’s Sense study similarly found that statutory documentation altered everyday classroom practice, increasing the likelihood that pupils would be taught in smaller groups or receive additional support from teaching assistants[720].

Once documentation becomes the principal route to support, demand for diagnosis inevitably increases. As discussed in Section 13.2, families, schools and clinicians each respond rationally to these arrangements.

13.7 Transitions

13.7.1 Rising distress and demand in higher education

As young people move on from school, they encounter new systems – higher education, occupational health, primary care, community services, adult mental health services and social care – each operating according to different criteria and holding different records. For some young people, these transitions remain abrupt and consequential. The move from children’s to adult social care, and from child and adolescent to adult mental health services, is frequently experienced as a cliff edge: eligibility is reassessed against different and often narrower criteria, established support may fall away, and what remains varies considerably between local authorities. Transitions should instead become opportunities for coordinated planning.

By 2024, around half of young adults had entered higher education by the age of 25[721]. The student population is both larger and substantially more diverse than in previous generations, and includes many more students with disabilities, neurodevelopmental conditions and mental health difficulties.

The evidence suggests that psychological distress has increased alongside this expansion. Within Understanding Society, the proportion of students meeting the threshold for psychological distress increased from approximately 34% to 43% between 2015 and 2019, while APMS shows rising rates of common mental disorder, particularly among young women[188, 722].

Importantly, these trends broadly mirror those seen among young adults who do not attend university. Factors specific to higher education, including tuition fees and financial pressures, may contribute for some students, but are unlikely to provide a sufficient explanation. Universities are responding to wider changes in the mental health and neurodevelopmental profiles of young adults.

In response, university counselling and wellbeing services now commonly see 10– 12% of the student population each year, while national benchmarking suggests that almost one in five students accessed counselling or wellbeing services during the preceding year[723]. Universities consistently report both increasing numbers of students seeking support and greater complexity, including higher rates of severe distress, self-harm, suicidality and co-occurring mental health and neurodevelopmental conditions.

“I messaged my university wellbeing team as I’ve been struggling with anxiety and other mental health issues, but they just sent me a list of things that I

already know about. I wanted them to help me, not signpost me to the Samaritans.”

Tellmi user, aged 19 The Association of Colleges reports that mental health need rose sharply following the COVID-19 pandemic and has remained high, yet fewer than half of colleges currently receive support from a Mental Health Support Team[724]. Colleges therefore need to be considered explicitly within any coherent national approach to young people’s mental health.

“Whenever I tried to access healthcare support at university, I found myself spending more time explaining and justifying my diagnoses, and having their validity questioned, than discussing the reason I had sought help in the first place. It often felt less like a clinical consultation and more like an interrogation. As a result, I became increasingly reluctant to seek support when I needed it. Instead, I delayed accessing care until I reached crisis point, often resulting in an entirely preventable A&E admission.”

Lived Experience Partner Analyses of NHS Talking Therapies suggest that young adults – and students in particular – experience lower recovery rates than older adults or similarly aged young people in employment[725]. This may indicate that services designed primarily around the wider adult population do not always align with the developmental circumstances of emerging adulthood. Further and higher education should therefore be understood as central components of the mental health system.

13.7.2 Reasonable adjustments

Requests for formally assessed adjustments within higher education have increased, with mental health and neurodevelopmental conditions accounting for a growing proportion. Figure 13.3 illustrates this trend using data from the University of Cambridge.

Figure 13.3 Assessment adjustments at the University of Cambridge by reason, Easter Term examinations, 2016–2024

Source: University of Cambridge, Annual Reports. Included here as a cross-setting illustration.

Reasonable adjustments are valuable and necessary, but their implementation is constrained by two factors. First, providing highly individualised support to around one fifth of the student population is increasingly difficult for universities and colleges to manage in practice. Second, there is little robust evidence demonstrating which adjustments improve longer-term outcomes. Occasionally, commonly used adjustments may have unintended adverse effects; for example, experimental studies suggest that trigger warnings can increase rather than reduce anxiety[726-728].

One response is inclusive or universal design, in which teaching, assessment and learning environments are developed from the outset to accommodate the widest possible range of students, reducing reliance on highly individualised adjustments[729]. Robust evaluation of the effects of both reasonable adjustments and universal design on student mental health is urgently needed.

Young people describe clear challenges in accessing support: “Whenever I tried to get help, the question was ‘What do you need?’ rather than ‘Here is what we can offer – let’s work it out together.’ I had been diagnosed only a year earlier and didn’t yet have the language to name my needs; I was expected to advocate for myself before I had been given the tools to do so. The problem wasn’t unwillingness to engage – it was a system that assumed I already knew how to navigate it.”

Lived Experience Partner This illustrates why responsibility for continuity of support at transition points should rest with the system rather than the individual – for example, through portable support plans or student, employer and care passports – reducing reliance on a young person’s ability to advocate for themselves.

These arrangements must not come at the expense of privacy. Mental health information is particularly sensitive, and young people may confide in one service precisely because they trust that information will not automatically be passed elsewhere. Portable records should therefore remain under the control of the individual – and their family where appropriate – rather than institutions. People should know what information is held about them and decide when and where it is shared (see Annex A).

13.8 Review of support and reasonable adjustments

Education is inherently developmental. Children grow, environments change and expectations evolve. Support that is appropriate at one stage may become insufficient, unnecessary or even restrictive at another. Regular review is therefore an essential part of supporting development.

This principle applies at every level. Within schools, statutory annual review of EHCPs exists precisely because children’s needs are expected to change over time. In practice, however, joint Ofsted and Care Quality Commission inspections have identified weaknesses in the timeliness, coordination and quality of annual reviews[730]. The Public Accounts Committee has also highlighted backlogs and the continuing need for many families to pursue complaints or appeals in order to obtain support that should follow from routine review[679].

In further and higher education, there is very little evidence regarding how reasonable adjustments are reviewed once established. Sector commentary has described an ‘accountability vacuum’ around ongoing review[731], while disability advisers report caseloads that leave little opportunity for proactive monitoring[732]. In practice, support plans often remain unchanged unless students themselves request reconsideration or difficulties escalate, precisely when review becomes least preventive.

Complaints reviewed within higher education repeatedly identify problems arising because support plans failed to adapt to changing course requirements, placements, laboratory work, independent study or fluctuating health[733].

Two broader principles follow. First, review should be understood as a dynamic process in which support can be increased where needed or appropriately reduced, reshaped or withdrawn as children’s confidence, skills and independence develop. Without regular review, assessments made at one point in development may become fixed descriptions of a child’s capabilities, shaping expectations long after they cease to reflect current functioning. Second, regular review strengthens both fairness and public confidence by demonstrating that provision continues to reflect current need.

“I have an EHCP but I don’t even need it anymore. It was to help me when I used to get in trouble at school with my ADHD but now I want to get into college and I don’t need it at all. I’ve changed so much and it’s slowing down me getting into college. I wish I didn’t even mention it on my application”

Tellmi user, aged 17 The Review concludes that regular review of support is essential to good educational practice. Provision should evolve alongside children’s functioning, participation and aspirations rather than remain anchored to an earlier assessment. This principle lies at the heart of the direction of current SEND reform, to which the chapter now turns.

13.9 Implications of current SEND reform

The direction of current SEND reform, set out in the Government’s White Paper Every Child Achieving and Thriving[734], is consistent with many of the conclusions reached in this chapter. At its heart is a shift towards inclusive, high-quality universal provision that is expected to meet the needs of most children, reducing unnecessary escalation to individual statutory support. Properly implemented, this approach offers a genuine opportunity to improve earlier access to support. Whether it succeeds will depend on how planned reforms are translated into everyday practice.

Closer integration between education and health is essential. Schools should build on information already available before children enter education, including developmental surveillance and school-readiness measures. Early concerns can then be followed by timely support, continued observation and proportionate specialist input.

The clearest organisational expression of this approach is the neighbourhood multidisciplinary model described in Chapter 19. Teams based around family hubs and primary care could bring together educational psychology, speech and language therapy, occupational therapy, school nurses, MHSTs, paediatric

neurodevelopmental expertise and CAMHS into a coordinated local resource that supports schools rather than operating separately from them. In such a system, specialist expertise is brought to the child rather than the child being referred repeatedly between different services. Support for autism and ADHD could increasingly resemble existing models used for hearing or visual impairment, where specialist expertise is embedded within education while remaining closely connected to health services. Diagnosis would continue to guide clinical treatment wherever appropriate.

Achieving this vision also requires a shared understanding of children’s needs. Developing a common language of functioning, participation and support need would allow professionals across sectors to work together more effectively (see Chapter 18).

Digital infrastructure is equally important. Integrated care depends on shared identifiers, interoperable records, and routine monitoring of participation, equity of access and outcomes. Born in Bradford illustrates the potential of integrated, data- informed approaches to supporting children’s development across multiple public systems. Interoperability does not require clinical records to be pooled or freely shared. Rather, shared identifiers enable services to coordinate care and link de- identified data, allowing outcomes to be monitored across systems and at population level.

If inclusive provision is expected to meet greater need without corresponding investment, schools will continue to rely on diagnosis to justify access to additional support. If health and education do not develop a genuinely shared language of need, integration will remain difficult to achieve in practice. If access to specialist advice continues to vary geographically, inequalities will persist even within a reformed system. Most importantly, if families lose confidence that support will be available on the basis of need, they will understandably continue to seek formal diagnosis as the most reliable route to securing help for their children.

The evidence reviewed here leads to a clear conclusion. SEND reform should strengthen early identification, improve timely access to integrated specialist expertise – for example, speech and language therapy working alongside mental health teams – and ensure that children with the greatest needs retain a clear individual entitlement to the specialist provision they require, while making far fuller use of the reasonable-adjustments duty that already applies to every disabled child without a plan. Such reforms are likely to reduce reliance on diagnosis as the gateway to support while improving participation. For many children, accurate diagnosis will remain essential, but support should no longer have to wait for it.

Chapter 14. Health services, access and pathways

People enter mental health and neurodevelopmental support through multiple routes, but thresholds, responsibilities and commissioning arrangements often do not align. The result can be delay, repeated assessment and fragmented pathways.

Co-occurring needs are common, yet services are often organised around separate pathways, sectors and age boundaries.

Waiting is not neutral. Delays can worsen distress, disrupt education, employment and family life, and increase the likelihood of people reaching crisis. Fragmentation can also widen inequalities, particularly for those least able to navigate the system.

Increasing capacity within individual services will not, by itself, solve problems of access if people continue to be redirected between disconnected pathways.

‘No wrong front door’ should become a reality. Wherever people seek help, they should receive timely assessment, proportionate support and clear onward navigation, with specialist expertise available when needed.

Pathways across health, education, social care and community services should be connected through shared information, linked data, coordinated commissioning, clear accountability and named responsibility for continuity of care.

Integrated mental health and neurodevelopmental pathways should be jointly commissioned to reduce duplication, repeated assessment, separate waiting lists and fragmented care for people with co-occurring needs.

This chapter examines what happens after a person recognises a need and seeks help: whether they can enter an appropriate pathway, move between services without losing continuity, and reach assessment and support in a timely way. People may enter through general practice, schools, colleges, NHS Talking Therapies, emergency departments, neurodevelopmental services, social care, workplaces, voluntary organisations and digital routes. Chapter 15 examines what happens once

people have entered the system, including access to evidence-based treatment and the delivery of care.

Many people receive timely help, but others encounter multiple misaligned service thresholds, repeated assessments and unclear responsibility, particularly where needs overlap or people move between services. Improving access requires sufficient capacity alongside clearer entry points, shared information, coordinated responsibility and continuity across organisational and age boundaries.

14.1 Entering the system: multiple pathways to support

Mental health care has multiple front doors. People seek help through general practice, NHS Talking Therapies, schools and colleges, workplaces, voluntary organisations, digital platforms, social care, the criminal justice system or emergency departments, depending on how need is recognised, where help is sought, local provision and urgency. These routes can widen access, but they operate with different thresholds, waiting times and patterns of support and do not always connect into a coherent network of care.

Research in children’s services illustrates this challenge. A national evaluation commissioned for the Review, covering ten sites across England, found that even where a named single point of access existed, other entry routes, including local authority, safeguarding, crisis and online services, operated alongside it. A family in a typical area in 2018/19 could therefore encounter three to five separate so-called ‘single’ points of access, each with different eligibility criteria and no mechanism for coordinating access between them.

General practice occupies a distinctive position within this landscape. General practitioners (GPs) recognise need, provide initial support[735], connect people with non-medical help through social prescribing[736], make referrals to other health services, and help people navigate the wider system. The extent of this role varies with age, setting, local service arrangements and the nature of a person’s difficulties. Prescribing activity illustrates its scale: antidepressant prescriptions dispensed in England increased from 47.3 million in 2011 to 85.6 million in 2022–23, with an estimated 7.2 million identified patients receiving 24 million antidepressant items in the final quarter[737].

As independently commissioned services have expanded, GPs have increasingly interpreted local referral pathways, explained thresholds and waiting times, and reconnected people with services when pathways fail to join up. This coordinating role has largely evolved by default and varies according to local capacity.

Direct access routes have also expanded. Adults, and in some areas young people aged 16 years and over, can self-refer to NHS Talking Therapies, while children and young people may access support through schools and colleges, including Mental Health Support Teams (MHSTs), and other community settings. In two English children’s mental health services, introducing a single point of access increased self- referrals from around 5% to 15–25% of all referrals. Self-referrals were slightly more likely to be accepted than GP referrals but were concentrated in the least deprived areas. Rejected self-referrals were more likely to involve boys, younger children, those from more deprived families and those not of White British ethnicity[738]. Self- referral can widen access, but referral pathways must also be designed to be equitable and accessible for all population groups.

Emergency departments provide another front door, particularly when earlier routes are unavailable or needs escalate. They provide an essential safety net during acute crises but were not designed to deliver ongoing mental health care[739].

England’s urgent and emergency mental health pathway and liaison psychiatry standard requires face-to-face assessment within one hour of presentation to an emergency department[740]. This standard exists to minimise delays in assessment for people experiencing mental health crises[741]. Further waits may nevertheless follow where admission, specialist beds or community placements are required.

Access to care is therefore better understood as a network than as a single gateway. Different entry points serve different populations, operate under different rules and connect to different onward pathways. A ‘no wrong front door’ approach is therefore needed to ensure that, wherever people first present, they receive timely assessment, navigation and access to support appropriate to their needs.

Analysis undertaken for the Review suggests that the effectiveness of ‘no wrong front door’ arrangements depends less on the entry point itself than on the pathways behind it. Single points of access and open-access hubs can improve visibility and simplify entry into support, but they do not by themselves create integrated care. The Review heard repeatedly that approaches based on ‘no wrong front door’ principles, strong professional relationships and effective connections between local services were more likely to provide coherent and responsive pathways. The challenge is therefore not simply to create a single gateway, but to ensure that multiple entry points operate as part of a genuinely connected system of care.

14.2 Cross-sector pathways and transitions

Mental health and neurodevelopmental needs may first be recognised at home or through education, employment, welfare, housing, justice, voluntary-sector or

healthcare settings, many of which also provide support. These systems serve different purposes, use different professional languages and evidence, apply different eligibility criteria and maintain separate records. Education focuses on participation and reasonable adjustments; employment on occupational functioning and workplace support; welfare on eligibility for financial assistance; and health services on diagnosis, risk and treatment. As a result, people may be asked to repeat the same information and undergo multiple assessments as they move between them.

This discontinuity is particularly evident during the transition from children’s and young people’s mental health services to adult services. Although 0–25 models of care have softened the age boundary in some areas, evidence gathered by the Review suggests that many young people with ongoing mental health needs experience a reduction in support during transition, creating a risk of unmet need and poorer outcomes.

“Our autistic and ADHD daughter’s transition from CAMHS to adult mental health teams was delivered very poorly. An adult care coordinator was not allocated until two weeks before her 18th birthday. Support then stopped for three months; weekly support and therapy under CAMHS became a care coordinator appointment once every six weeks with no therapy. Her mental health dropped dramatically and she stopped going to college.”

Lived Experience Partner Discontinuity also occurs when young people move from school into further or higher education or employment, where educational, health and occupational support often operate independently (see Chapter 13). Transitions from children’s inpatient services to adult mental health services present a particular patient safety concern when young people cross age, service and organisational boundaries without adequate planning, preparation and continuity of support[742].

For young people with ADHD or autism, the move from paediatric or children’s and young people’s mental health services into adult provision can expose gaps in commissioned pathways and specialist expertise[743]. Responsibility for prescribing, medication monitoring and review may become unclear, while support available at school is not always matched in higher education or employment. For some lifelong conditions, services remain organised primarily around childhood[260].

Improving transitions requires clear responsibility for coordination, shared information, and support that follows people across age, organisational and sector boundaries.

14.3 Mental health and neurodevelopmental pathways

Mental health and neurodevelopmental services have largely evolved as separate specialist pathways. Their organisation reflects historical diagnostic frameworks, commissioning decisions and patterns of service development rather than the way people experience their needs. Children, young people and adults rarely present with a single isolated condition: co-occurrence is the rule rather than the exception (see Chapter 9). Mental health difficulties frequently occur alongside neurodevelopmental and physical health conditions, while communication difficulties, adverse childhood experiences and social disadvantage further shape presentation. Separate assessments and sequential waiting lists are therefore inefficient where complexity and co-occurrence are commonplace[271].

A child with attentional difficulties and anxiety may be referred separately to neurodevelopmental and mental health services, each conducting its own assessment without developing a shared understanding of the child’s needs. In some areas, children’s and young people’s mental health services are not commissioned to provide support where a neurodevelopmental condition is considered the primary concern unless a mental health difficulty also meets the service threshold. Neurodevelopmental services may, in turn, defer intervention while mental health difficulties are addressed. Families can therefore find themselves moving repeatedly between services without a shared understanding of need or a coherent plan for ongoing care.

This duplication has prompted growing interest in more integrated models. Within neurodevelopmental services, emerging evidence suggests that combined pathways for ADHD and autism may shorten decision-making and waiting times compared with separate condition-specific pathways.

Box 14.1 Regional analysis – combined neurodevelopmental pathways

Review analyses found shorter waits in combined neurodevelopmental pathways, although service models differed. In one city, by 2024 the median wait for assessment was around 21–24 months in an autism-only pathway compared with seven months in a combined pathway, while median referral-to-diagnosis time was around 50 months versus 25 months. Over the same period, median patient- level time to diagnosis was 16 months in the combined pathway compared with 25 months in the ADHD-only pathway.

A second combined pathway showed a different pattern. Waiting times to assessment increased, while referral-to-treatment remained at around two months. Median patient-level time to diagnosis was 4.9 months overall in the combined pathway, increasing to around 14 months among adolescents in 2023.

These observational comparisons do not establish causality.

Separate pathways also affect diagnostic accuracy and treatment. Adults presenting to mental health or substance misuse services may have previously unrecognised ADHD or autism. Unrecognised ADHD appears to be common among people presenting with first-episode psychosis, bipolar disorder or substance misuse, yet neurodevelopmental expertise is often less available within these services[743]. Conversely, anxiety, low mood and other treatable mental health conditions may be attributed to an existing neurodevelopmental condition. Diagnostic overshadowing can therefore operate in both directions and across the boundaries between mental health and physical health[277, 280, 744]. The implications for treatment are considered in Chapter 15.

Linked records from a medium-sized, diverse city illustrate how delayed recognition of additional conditions may reflect diagnostic overshadowing (see Chapter 5). Among children with both autism and ADHD, the second condition was recorded, on average, four to five years after the first. Children with a neurodevelopmental condition were more than twice as likely subsequently to have anxiety or depression recorded (19.5% versus 8.7%), with the proportion consistently higher when the condition was recognised later rather than earlier. Among children with both autism and ADHD whose autism was recognised first, around three in five subsequently had anxiety or depression recorded.

Service design itself can create barriers. Review analyses found that autistic adults and people with learning disabilities experienced greater difficulty accessing primary care than their peers, including making appointments, communicating with

professionals and managing clinical environments. GPs also reported lower confidence in recognising co-occurring mental health conditions in these groups. Autistic people similarly describe mainstream mental health services as a poor fit when communication, sensory and environmental adjustments are lacking[278, 315].

Children’s services demonstrate a similar mismatch at the front door, depending on how difficulties present. In three services examined as part of a national evaluation, referrals involving self-harm or suicidal thoughts were seen most rapidly, typically within five days. However, they were also most likely to be assessed and discharged without treatment, with around half subsequently re-referred. Children presenting with autism or ADHD waited longest for assessment. Autism presentations had the highest re-referral rate, at just over half. Mixed or difficult-to-categorise presentations were most likely to be screened out without the child being seen, regardless of underlying need[745].

Services therefore often require people to identify the correct diagnostic pathway before assessment, whereas assessment is frequently needed precisely because that distinction is unclear. Assessment should help determine which conditions are present and therefore which pathway is most appropriate.

Integrated pathways respond differently by assessing strengths, impairment, differential diagnosis, co-occurring conditions and immediate priorities together. A named professional can coordinate communication, reduce duplication and help individuals and families navigate services. Specialist expertise remains essential, particularly for complex or high-risk presentations, but it can be brought into a shared pathway rather than requiring separate queues for each suspected condition. The Review therefore recommends commissioning integrated neurodevelopmental and mental health pathways (Chapter 20, Recommendation 3).

A more proportionate model separates self-report screening, which identifies symptom burden but cannot establish diagnosis; structured, criterion-based triage, which determines whether a person should proceed to full assessment, follow an alternative pathway, receive interim support or be escalated; and comprehensive diagnostic assessment. Triage is not diagnosis by another name. It directs specialist capacity towards those most likely to require it while identifying people who may need a different or more immediate response.

Structured screening and triage already operate within NHS services. Examples include prioritisation categories for ADHD and autism referrals in South West Yorkshire, six possible triage outcomes in Bath and North East Somerset, Swindon and Wiltshire, and a central triage function within Greater Manchester’s transformed pathway. Preliminary evidence from a prospective validation study in an adult NHS

service suggests that criterion-based triage can be implemented safely while improving efficiency.

Box 14.2 Service-level analysis – triage, clinical appropriateness and an end- to-end adult pathway (South West Yorkshire)

South West Yorkshire Partnership NHS Foundation Trust operates one of the few end-to-end adult neurodevelopmental pathways, combining structured screening and triage, diagnostic assessment and structured post-diagnostic support. Since January 2025, trained Physician Assistants have administered a DSM-5-based triage assessment of around one hour for adult ADHD referrals through a criterion-based Single Point of Access.

In a prospective validation study, 48 adults referred for ADHD underwent both structured triage and a full NICE-compliant assessment. Triage identified all confirmed ADHD cases in this small sample (observed sensitivity 100.0%, 95% CI 61.0–100.0%; negative predictive value 100.0%, 95% CI 83.2–100.0%), while classifying 39.6% of referrals as unsuitable for a full ADHD assessment. Estimated cost was £731 per referral compared with £1,011 for universal full assessment, representing a saving of around £280 per referral, or approximately £600,000 annually at current referral volumes.

These single-site findings should be regarded as proof of concept rather than definitive evidence of safety or effectiveness.

14.4 Thresholds, gaps and the ‘missing middle’

The ‘missing middle’ describes people whose needs exceed what can be managed in primary care but who do not meet thresholds for specialist mental health services, or who face prolonged waits before those services become available[746] (see Chapter 15). The term is increasingly used more broadly to describe people with mental health needs who receive no NHS mental health support, estimated at around 8 million people with mental health needs who are not in contact with NHS mental health services – estimated by NHS England in 2021 based on indicative modelling in the absence of current prevalence data[747]. In practice, people may be ‘too well’ for one service, ‘too complex’ for another, or repeatedly redirected between services.

Neurodevelopmental pathways illustrate the distinction between gaining an assessment and receiving support. Many services are commissioned primarily around diagnosis[748], yet the practical benefits of diagnosis do not always consistently follow[260, 743]. A national clinician survey conducted for the Review found that only around half of autism services offered any support while people were waiting for assessment.

Increasingly, fixed eligibility criteria and diagnostic thresholds fail to capture the cumulative impact of several co-occurring difficulties. A person may be considered too complex for NHS Talking Therapies because they require multidisciplinary support, while simultaneously being regarded as insufficiently unwell for secondary mental health services. Others experience combinations of mental health, neurodevelopmental, trauma-related, physical and social difficulties that cross several service boundaries. Repeated referral or ‘signposting’ may then transfer responsibility between services rather than coordinate care.

Risk thresholds can create a similar distortion. Risk assessment is essential, but people with substantial distress, functional impairment or developmental disruption may remain below urgent-care thresholds, making help easier to access after needs have escalated into crisis than at an earlier stage.

Children’s mental health data illustrate what can be lost at the front door. In 2018– 19, around 133,000 referrals (25%) to specialist children’s mental health services were not accepted[749]. Across ten children’s mental health services, almost one in three referrals was rejected, redirected or closed without the child being seen, and only around half resulted in recorded treatment[745] (see Figure 14.1).

“Despite seeking help for my daughter for years, it took until she ended up in A&E in a full shut down for anyone to listen, as she masked well at school and achieved well, and the education setting is the gatekeeper of getting a child neurodevelopmental (ND) support. Even after the A&E and overnight admission it took multiple letters and a formal complaint to get the services to do what they said they would… it took a further 9 hours of phone calls to get CAMHS and the ND team to work together as they agreed they would. If I were less able to advocate for her, she would still not be receiving care and would still be coming to harm.”

Lived Experience Partner Rejection did not end need. Around three in ten children turned away without being seen were referred again within a year. In one area, only 17% of those whose first referral was rejected received an assessment within the following year, while one third of those referred again were turned away a second time. Repeat referral roughly doubled the likelihood of receiving treatment, making access partly dependent on whether someone was able to persist on the child’s behalf. Minority ethnic families were less likely to refer again after rejection, and Asian or Asian British children were 1.3 times more likely never to receive an assessment[745].

Figure 14.1 Outcomes of referrals to children’s mental health services, from receipt of referral to recorded treatment, ten study services, financial years ending 2016–2019

Source: Referrals for under-18s supplied to a national evaluation by ten services in England, financial years ending 2016–2019. Referrals whose episode remained open when data were extracted are excluded. Percentages represent the proportion of all referrals pooled across the ten services.

Needs that are both clinical and social are particularly vulnerable to falling between pathways. Review data found that referrals describing family or social difficulties were the most likely to be screened out without the child being seen, and 27% of referrals from social care did not lead to assessment[745]. Professionals described a risk that needs are allowed to escalate before referrals clearly meet eligibility thresholds[750].

Commissioning arrangements can either bridge or reinforce these gaps. Successive national initiatives, including the Community Mental Health Framework and neighbourhood mental health centres, have sought to bridge primary and specialist care but have been implemented inconsistently[751-753]. Pathway mapping across ten areas illustrates this. Between 2015 and 2018/19, the number of children’s mental health services increased by around one quarter, from 454 to 573, with growth concentrated in access and early intervention services, while provision for some

groups with the greatest needs changed little[745]. Adding services without shared referral arrangements, shared records or shared accountability may increase complexity, while repeated organisational change can itself weaken coordination[750].

The consequences are both organisational and personal. People described prolonged periods without support, repeated assessments, uncertainty about responsibility and abrupt endings to care without clear routes back[754, 755]. Collaborative discharge, planned follow-up, gradual reductions in contact and clearly defined routes back into care can improve continuity and reduce feelings of abandonment[756].

Additional capacity may be necessary, but capacity alone will not resolve fragmentation. Effective pathways require clear entry points, a shared understanding of need, active coordination and explicit responsibility when care moves between services. Services need to adapt to the complexity and changing nature of people’s needs rather than requiring individuals repeatedly to adapt to service structures.

Regional data further illustrate this point. In one region, replacing paper triage with an assessment centre reduced the monthly proportion of referrals closed without assessment from 21–32% to 3–5%. Referral numbers and the case mix remained broadly unchanged. Median waiting time for assessment fell from 63 to 29 days, monthly throughput increased from 178 to 227 referrals, and median waiting time for treatment fell from 173 to 84 days. Across fourteen areas, services that closed the highest proportion of referrals without assessment subsequently treated the smallest proportion of referrals. Turning people away without assessment did not increase treatment capacity[745].

14.5 Waiting, deterioration and fragile commissioning

Waiting is not neutral. Among children and young people alone, more than one million had an active referral to mental health services within a single year. Around one third were still waiting at the end of that year, and approximately 60,000 had waited for more than two years[757].

Long waits can change people’s needs, help-seeking and the demand subsequently experienced by services. In the Care Quality Commission’s 2024 national survey of more than 14,000 community mental health service users, 42% reported that their mental health had deteriorated while waiting for care. The proportion reporting deterioration increased steadily with waiting time, rising to 71% among those waiting longer than six months[758]. As noted in Chapter 12, these findings should be interpreted with appropriate caution given the survey’s response rate was 19%. A

separate survey reported similar findings, with two in five adults with a diagnosed mental illness reporting that waiting between referral and treatment had worsened their mental health[759].

“Whilst waiting 2 years for an autism assessment our 14-year-old daughter’s mental health became much worse and into crisis with several trips to emergency departments and then admission to inpatients. I often think of what would have happened if support in school and mental health support would have been given without having to wait so long.”

Lived Experience Partner Delay can disrupt education, employment and family life, erode trust, and alter the needs that services eventually encounter. At its extreme, referral can cease to provide a meaningful route into care, leaving some children formally on a waiting list but effectively outside the system[757].

Waiting also varies geographically and should be considered alongside referral acceptance. In 2018–19, providers in London declined around 17% of referrals to children’s mental health services, compared with approximately 28% across the South, Midlands and East, but recorded the longest median waits among accepted referrals[749]. Services that accept more referrals may develop longer waiting times if capacity does not increase accordingly. Although several mental health waiting-time standards have now been introduced, coverage remains incomplete and expectations continue to differ between pathways.

Variation between services also suggests that long waits are not inevitable. The Review identified local systems combining comparatively short waits with low rates of referral rejection through open-access, simplified pathways and strong partnerships between NHS, voluntary-sector and community providers. Some also offered support while people waited for specialist intervention. These examples suggest that pathway design and integration may influence waiting times alongside service capacity.

Box 14.3 Regional analysis – patient-level waiting times were substantially longer than referral-level waiting times

National waiting-time statistics generally treat each referral as a separate episode, restarting the clock after discharge, redirection or re-referral. Linking referrals for the same child produces a different picture. In one city, median time to diagnosis was 38.9 months at patient level compared with 18.2 months at referral level. Within the same city, a combined autism and ADHD pathway reduced these figures to 16.0 and 9.3 months, respectively. Referral-level measures may therefore substantially underestimate the cumulative waiting time experienced by children who repeatedly re-enter pathways.

Waiting also redistributes demand. Families may return repeatedly to general practice, schools and employers for advice, while emergency departments or crisis services become the point of contact when difficulties escalate. Others seek privately funded assessment, meaning that delay may widen inequalities as well as shifting pressure elsewhere.

“After getting a GP referral for ADHD assessment, I entered the ‘waiting list’ phase with no idea when I might be assessed – I was told it could be up to eight years. My distress worsened. It took written complaints and escalation to get an assessment after 18 months, made harder because no local GPs accepted shared care. The subsequent diagnosis and medication has been genuinely life changing, but I was made to feel unreasonable for wanting an assessment quickly – it is not after all my fault the current system is overwhelmed and not fit for purpose.”

Lived Experience Partner Long ADHD and autism waiting times have coincided with substantial growth in independent provision, privately funded assessment and Right to Choose referrals. As waiting times in the independent sector became shorter than those in local NHS services, NHS-funded activity increasingly shifted to independent providers through Right to Choose. Provider estimates suggest that independent providers delivered more than half of NHS-funded ADHD assessments and more than one third of autism assessments in 2025/26, with around 90% of this activity occurring through non-contracted Right to Choose arrangements. Within the same providers, NHS- funded patients waited a mean of 151 days for ADHD assessment and 170 days for autism assessment, compared with 21.5 days for privately funded patients. Timely assessment increasingly depends not only on where people live, but also on how care is funded.

The effects of delay accumulate. Needs may increase, family resources and trust may erode, and opportunities in education, employment and everyday life may be lost. Waiting therefore changes the needs that services eventually encounter rather than simply postponing care.

14.6 Crisis, escalation and step-down

Mental health crises may arise rapidly, but many follow longer periods during which difficulties have accumulated, functioning has deteriorated and opportunities for earlier intervention have been missed. Crisis is both a clinical event and an indicator of how well earlier parts of the pathway have functioned. In a survey of adults on so- called hidden waiting lists, more than three quarters had used a crisis line or emergency service while waiting for treatment, including around one in eight who attended an emergency department and one in fourteen who called 999[759]. For many people, crisis is the point at which longstanding difficulties are first recognised and responded to with urgency.

Emergency departments provide an essential safety net for acute psychological distress, self-harm, suicidal crisis and severe behavioural disturbance. They are available without referral thresholds or diagnostic categories, but they were not designed to provide continuing mental health care. Heavy reliance on emergency departments may therefore reflect gaps in accessible community alternatives. Attendance has increased, placing additional pressure on acute hospitals, where specialist psychiatric liaison is not always available[760] (see Chapter 17).

Following a crisis, people may move rapidly between services. Abrupt discharge, unclear responsibility and difficulty re-entering care can disrupt continuity. Housing insecurity, disrupted education, employment, family stress and financial pressures do not disappear once the immediate crisis has resolved. Effective pathways therefore require clear follow-up, good communication and reliable routes back into care. The aim is not only to respond to crises but to reduce their recurrence and support recovery.

The period immediately after discharge is a particularly high-risk phase and requires careful handover. Well-designed step-down pathways can improve continuity and reduce pressure on acute services. For children and young people, intensive community care combining home treatment, assertive outreach and supported discharge can reduce hospital admission while maintaining or improving outcomes and is likely to be cost-effective[761]. Among homeless adults in England, step-down

intermediate care following hospital discharge was associated with 18% fewer emergency department attendances, although this benefit diminished once funding ended[762]. These findings reinforce the importance of flexible pathways that allow support to step up and step down as needs change.

14.7 The earliest pathways: infancy, early childhood and families

Many mental health and neurodevelopmental trajectories begin before people reach specialist services. Emotional regulation, communication, attachment, social development and early learning develop rapidly during infancy and the preschool years. Infant mental health is inherently relational, so support must often work with parents or caregivers and infants together[763, 764]. Early recognition of neurodevelopmental differences can help families understand a child’s needs and obtain appropriate support or adaptations. Emotional difficulties in infancy are often under-recognised and may persist if left unaddressed[765]. At least 17% of children under five in England are estimated to have a mental health need, with higher rates in more deprived areas[685].

Specialist perinatal mental health services have expanded substantially, with England exceeding the ambition to support 66,000 women annually in 2025 [766]. However, demand continues to exceed specialist capacity in some areas[767], and many women still struggle to access coordinated support. This is particularly true for those whose needs fall below specialist thresholds[768], or whose difficulties span mental health, physical health, neurodevelopmental, social care and safeguarding services. Perinatal mental health difficulties affect around one quarter of women, and maternal suicide remains the leading cause of death between six weeks and one year after pregnancy[185, 769].

Provision for infants and young children remains uneven. The ‘baby blind spot’ describes the gap between stated priority and actual provision[770]. The national model specification for children’s mental health services did not mention babies or children under five[763], while evidence submitted to the Review by the Parent-Infant Foundation highlighted continuing gaps in commissioning, national guidance and service provision for infant mental health, alongside the importance of relational and attachment-based approaches that consider prenatal and wider family contexts. Only a minority of children under five who need support currently receive it[771].

Funding and provision also remain limited and uneven. No national funding is ring- fenced specifically for under-five mental health, which competes for attention and resources within wider services for children and young people aged up to 25[764]. Around 50 specialist parent–infant teams operate across the UK, but most accept

referrals only up to the age of two, leaving limited provision for three- to five-year- olds[763]. Family Hubs and Start for Life funding initially covered 75 local authorities, while specialist parent–infant mental health teams operate in only two of Wales’s seven regions[763]. Consequently, access to support remains highly variable and often depends on where families live.

Early childhood pathways are inherently cross-sectoral. Midwives, health visitors, general practice, early-years settings, paediatrics, social care and voluntary organisations may each hold part of the picture. Where responsibility for coordination is unclear, emerging needs can remain unrecognised until school age or until difficulties become sufficiently severe to trigger statutory intervention. Reductions in preventive provision, including children’s centres and health visiting, have been followed by rising demand elsewhere, including increasing emergency department presentations among children under one year of age[772], while reducing opportunities to identify and respond to emerging needs early.

Early intervention therefore requires maternal and child health, education, family support and community services to work together around children’s developmental needs. The Royal Foundation for Early Childhood estimates that more than £16 billion is spent annually in England on remedial action associated with inadequate support during infancy and early childhood[773], while specialist parent–infant services have been estimated to return around £13 in public savings, in addition to wider social value, for every £1 invested[764]. The priority should be to provide coherent support early, before difficulties escalate and pathways become more complex.

Support for families must also continue beyond the early years as children encounter new developmental challenges and concerns about mental health or neurodevelopment emerge. As highlighted throughout this Review, where mental health support is available it is often episodic, reactive and short term, and restricted to those meeting specific eligibility criteria. By contrast, mental health conditions, autism and ADHD frequently follow enduring developmental trajectories with consequences across the life course[774, 775].

Young people consistently describe parents and carers as the people they turn to first, and value most highly, when they experience mental health difficulties[696]. Similarly, in the context of neurodivergence, qualitative evidence suggests that young people often prefer support from a parent or carer with whom they already have an established relationship and who is present at the times they are struggling[776].

Young people often also rely on parents and carers to navigate complex systems in order to access support[696]. Yet parents and carers report substantial barriers to

obtaining help for their children, including difficulties recognising emerging problems, concerns about the potential consequences of seeking help, uncertainty about how to access support, and structural barriers once they try to do so[777, 778].

“You know I think actually the parents should have help too, to help the

  • children, I think in the first place you know it’s part of his life, part of our
  • life… we need to support him here first and take it as a family basically.”

Parent participant[778] Parents and carers also told this Review that, once they reached services, they often felt sidelined, despite remaining the people their children relied on most in everyday life. Consistent with this, qualitative studies describe parents and carers wanting professional support that enables them to support their children more effectively rather than replacing their role[701, 779].

Failing to support parents and carers when children are experiencing mental health difficulties or unmet needs associated with ADHD or autism not only misses an opportunity for early intervention, but also places a substantial personal burden on families and wider costs on society. Caring for a child with mental health difficulties, autism or ADHD is associated with high levels of parental stress[780, 781]. Research by Deloitte estimated that the impact of children’s mental health on working parents costs UK employers around £8 billion annually, with half of concerned parents reporting an impact on their work performance, 10% taking up to five days’ leave each year to support their child, and around 1% leaving employment altogether because of their child’s mental health difficulties[782.].

Fragmented child mental health systems, characterised by poor access and limited coordination, can compound these difficulties by disempowering parents and increasing stress[783]. Conversely, support that enables parents to help their children effectively benefits both child and parent wellbeing[701, 779], with emerging evidence of longer-term cost-effectiveness[784, 785].

14.8 Inequalities in access, treatment and outcomes

Inequalities arise not only because some groups experience greater underlying need, but because recognition, referral, waiting times and continuity of care differ once people seek help. Small disadvantages may accumulate at each stage of a pathway, meaning that those with the greatest need do not always have the easiest access to support.

Geographical variation is considerable. Spending per referred child varies by around fourfold between areas, from £569 to £2,410[757]. Patterns of recognition also differ. Girls and women have historically been identified later than boys and men for neurodevelopmental conditions, while more adult women than men are now newly diagnosed with autism[58].

Socioeconomic disadvantage influences both need and movement through services. People living in more deprived communities are more likely to experience adversity, poorer physical health, insecure housing and financial insecurity, all of which increase vulnerability to mental health difficulties. Children living in more deprived areas also have higher referral rates to mental health services[757]. Recorded autism is around 45–50% more common in the most deprived than the least deprived areas, with the steepest gradient among children under five[58]. ADHD shows a different pattern: although recorded prevalence remains higher in more deprived populations, recent data suggest that rates of new diagnosis have become more similar across deprivation groups (see also Chapter 7). More deprived families may also face greater difficulty attending appointments, accessing local provision or paying for alternatives.

Many minoritised ethnic communities also experience inequalities in recognition and access to care. One quarter of Black children and around one in six Asian children referred to mental health services were referred in crisis, compared with fewer than one in twelve White children, despite Black and Asian children being under- represented among referrals relative to their share of the child population[757].

Box 14.4 Regional analysis – inequity may be concentrated at the point of access

In one city, White British children accounted for 67% of referrals to assessment pathways but around half of the child population, whereas South Asian children accounted for 21% of referrals but around one third of the population. Among children who entered pathways in this city or a second city, analyses found no consistent differences in waiting times by ethnicity or deprivation. Children’s Global Assessment Scale (CGAS) scores at first clinical contact also showed no consistent gradients by ethnicity or deprivation.

Small subgroups and incomplete ethnicity recording limit conclusions, but these findings suggest that inequalities may arise before people enter the pathway.

Recording of need also differs between groups. Autism-related special educational needs were recorded less often for some communities: 2.11% of Black pupils and 0.85% of Roma and Irish Traveller pupils had autism-related SEND recorded, with

lower rates also observed among children whose main language was not English[306]. Autistic people and people with learning disabilities may face additional barriers relating to communication, appointment systems, clinical environments and fragmented care. Inequalities may also continue after access: autistic adults were less likely than comparable non-autistic patients to benefit from standard NHS Talking Therapies and more likely to deteriorate, although average benefit remained positive[786]. Inequalities therefore reflect not only differences in health but also how accessible and appropriately adapted services are.

As recognition broadens, systems must continue to protect timely and appropriate support for people with severe or profound learning disabilities and the highest continuing needs. Transitions may further widen inequalities because family support, financial resources and confidence navigating services influence people’s ability to maintain continuity of care. Those with stable family support, stronger social networks and greater confidence navigating systems are better protected when pathways fragment.

Among young people with special educational needs, 22% became not in education, employment or training (NEET), compared with 7% of those without SEND[308] (see also Chapter 13 on participation and life outcomes). Similarly, carers, people in insecure employment and those outside education or work may face additional barriers. Reducing inequality requires pathways that are accessible, culturally responsive and less dependent on individual advocacy or private resources. The Patient and Carer Race Equality Framework (PCREF)[787] and the Culture of Care Standards[788] provide important foundations, but inequalities should also be measured across recognition, referral, waiting, continuity and outcomes.

14.9 Chapter synthesis

This chapter has followed people from the point at which they seek help. The evidence shows that access is shaped by more than the size of waiting lists. Multiple front doors, inconsistent thresholds, separate diagnostic pathways and fragmented commissioning all influence who reaches assessment and support, how long it takes, and how much responsibility falls on individuals and families to navigate the system.

The first implication is that multiple entry points should operate as a connected network. ‘No wrong front door’ means that general practice, schools and colleges, emergency departments, community services, digital routes and other front doors can recognise need, provide or arrange proportionate support, and make effective onward connections.

The second is that continuity should be designed into pathways. Shared information, clear responsibility and named coordination are particularly important during

transitions between children’s and adult services, between education and employment, and between crisis and community care. Linked data should make the whole journey visible to commissioners rather than recording each organisational episode in isolation, enabling pathways to be understood from the perspective of the person rather than of individual services.

The third is that co-occurring mental health and neurodevelopmental needs require integrated assessment rather than repeated condition-specific queues. Screening and triage can help direct specialist expertise to where it adds greatest value, but should not become new barriers to care or substitutes for comprehensive clinical assessment. Commissioning arrangements should increasingly support integrated mental health and neurodevelopmental pathways that reflect the reality of co- occurring needs.

The fourth is to reduce the role of thresholds and waiting in determining who receives help. Support should respond proportionately to need before people deteriorate, reach crisis or have to seek repeated referrals. What happens once someone is through the front door – whether evidence-based treatment and continuing support are available and appropriately matched to need – is the focus of Chapter 15.

The fifth is that access begins early. Pregnancy, infancy and the preschool years depend on connected pathways across maternity, health visiting, primary care, early- years services, family support and community services that promote mental health, support healthy development and identify emerging needs. Early support is most effective when it values and empowers families and is available consistently over time rather than depending on geography or short-term programmes. Every local area should commission integrated early-years mental health and neurodevelopmental pathways, ensuring coordinated support from pregnancy through the preschool years for children and their families.

Finally, access must be judged by equity. Differences in recognition, referral, waiting and continuity can compound the effects of poverty, ethnicity, disability, neurodevelopmental differences and other disadvantage. A pathway may be open in principle while remaining much easier to navigate for people with greater time, confidence, family support or financial resources.

Adding services is necessary in some places, but it is not sufficient. Sure Start, Future in Mind, Thrive, Family Hubs, MHSTs and Community Mental Health Transformation have all sought to improve coordination and enable earlier intervention. The recurring challenge has been creating the commissioning, workforce, data and organisational conditions that allow effective models to be sustained and spread.

People move through systems, and systems need to move with them. Better- connected pathways should make access earlier, more continuous and less dependent on navigating organisational boundaries, ensuring that support is shaped by need rather than diagnosis, thresholds or waiting lists.

Chapter 15. Service models, mental health treatment delivery and implementation

We identify a persistent implementation gap. Services can often measure access, activity and diagnoses, but cannot consistently demonstrate whether people receive timely, evidence-based care or report on the outcomes achieved. Among adults with severe and complex needs receiving community mental health care, only 2.7% receive a NICE-recommended psychological therapy.

The chapter also examines the ‘missing middle’: children and young people whose needs exceed what universal services can provide but who are not considered severe enough for specialist support. Early intervention can be strengthened through community, digital and other non-clinical forms of support, but these need to be integrated within coherent pathways.

This chapter also reviews the evidence for prevention in mental health. Chapters 7 and 8 consider prevention and early intervention in ADHD and autism. The strongest evidence supports structural interventions, particularly those that improve family income and the early-years environment. Evidence for universal population-level interventions is more mixed, although targeted preventive approaches have been shown to be effective.

The principal barrier is not a lack of evidence about what works, but the system’s ability to deliver it. Workforce shortages, generic roles that dilute specialist expertise, activity-based commissioning, fragmented pathways, weak post-diagnostic support, uneven access, and the diversion of resources towards acute response rather than prevention mean that care is too often shaped by capacity rather than need.

The preceding chapter examined how people enter services and move through pathways. This chapter focuses on what happens once they arrive: what services deliver, how consistently care is provided, and how delivery is shaped by workforce, commissioning, service design and implementation.

Diagnosis can provide a shared understanding of difficulties and guide treatment planning, but diagnosis is not an outcome. Equally, access to a service does not guarantee access to evidence-based treatment or improved outcomes.

Improving outcomes for people with mental health and neurodevelopmental conditions depends on timely recognition and diagnosis, consistent delivery of effective treatments, and, most importantly, prevention and early support, particularly for children and young people. Effective care also depends on integrated mental health and primary care services that recognise people’s physical, cognitive and social stage of development, together with their wider social and cultural context.

Throughout this chapter, evidence-based care refers not only to interventions supported by research evidence, but also to their delivery through the integration of best available evidence, clinical expertise, shared decision-making and informed patient choice[789].

The evidence reviewed in this chapter suggests that a substantial proportion of unmet need arises not because effective interventions are unavailable, but because they are not delivered consistently or at sufficient scale. A substantial implementation gap therefore remains between recommended and delivered care. Although the precise nature of this challenge differs across services, a consistent pattern emerges. Services have gradually become organised around assessment and case management rather than the delivery of evidence-based interventions, leading to diagnosis without timely treatment. Access to a service is therefore not the same as access to evidence-based care, and pathways designed primarily to increase access may not improve outcomes.

While there are important exceptions, several recurring themes emerge across conditions and services. Stratified, or matched, care has become established for common mental disorders in adults but is much less developed elsewhere. Guidance on prevention, early intervention and the management of complexity remains comparatively limited. Similar gaps exist in guidance relating to autism and ADHD, particularly where people present with multiple co-occurring needs.

Access to evidence-based psychological therapies falls substantially short of need, especially for people with severe mental illness, eating disorders, substance misuse and self-harm. Medication review and deprescribing receive comparatively little attention. Outside NHS Talking Therapies, routine outcome measurement also remains inconsistent.

Closing this implementation gap requires changes in service design and delivery. Effective care depends not only on identifying need but on delivering evidence-based treatment consistently and at scale.

15.1 Delivery of evidence-based treatment: fidelity, uptake and outcomes

Diagnosis, where it occurs, does not in itself guarantee access to treatment or support. Many people experience delays, and even after assessment or diagnosis, interventions recommended in national guidance are not delivered consistently. This gap is particularly marked for neurodevelopmental conditions and, perhaps most strikingly, for people with severe mental illness, where unmet need may culminate in acute crisis rather than timely intervention.

A substantial programme of primary research undertaken for the Review, among the largest of its kind, drew on 364 professional interviews conducted across children’s mental health services between 2015 and 2021, followed by a further 33 interviews, seven focus groups and structural mapping across ten local systems during 2024–

  • Despite widespread commitment to evidence-based practice, the study found

that service leaders were often unable to demonstrate reliably what treatments had been delivered, to whom, with what fidelity or with what outcomes. Estimates of the proportion of children receiving evidence-based interventions ranged from around 15% to over 90%, in some cases varying by that amount within a single organisation.

These findings illustrate a broader implementation challenge identified throughout this chapter. Services frequently described themselves as delivering NICE- concordant care, yet often lacked the data needed to demonstrate what had actually been delivered and with what effect. A recurring finding of the Review was that systems were better able to demonstrate access, assessment and activity than whether people received the treatments recommended in national guidance.

Three interacting mechanisms help explain this persistent implementation gap: i. Complexity. NICE guidance is organised around primary presenting problems to support clinical decision-making, whereas real-world presentations are complex. As one participant observed, “people don’t present in those kind of neat packaged ways”[745].

ii. Assessment has become disconnected from intervention. Professionals repeatedly described pathways in which substantial effort was devoted to assessment while access to evidence-based intervention remained limited.

iii. Capacity pressures. Workforce shortages, infrastructure constraints and historic underinvestment limit services’ ability to meet need. As the Darzi investigation of the NHS in England observed, mental health accounts for more than 20% of the burden of disease but receives less than 10% of NHS

expenditure[790]. Referrals for children and young people have also risen sharply – from around 40,000 to almost 120,000 per month between 2016 and 2024, an average annual increase of around 11.7%[790].

Waiting-time targets, workforce capacity and limits on the number of available sessions increasingly shape provision more than NICE guidance. Psychological therapies become harder to access, while medication increasingly fills the resulting gap – not necessarily because it is preferred, but because it remains available when other interventions are not. As one participant reflected, “as they’re waiting, they will get more distressed and then there will be requests for medication… because what else is there?”[745]

These gaps have measurable consequences. Among those who eventually receive treatment, longer waits are associated with worsening mental health (Figure 15.1). In children’s and young people’s mental health services, this challenge is compounded by the absence of an enforceable national waiting-time standard.

Figure 15.1 Change in mental health while waiting between assessment and first treatment, by wait length.

Source: CQC, The State of Health Care and Adult Social Care in England 2024/25 (n = 12,319; response rate 19%; adults aged 16 years and over).

The scale of the treatment gap can also be quantified. National Audit Office analyses suggest that only around one quarter of children with a probable mental health condition were accessing NHS-funded services, with subsequent national policy aiming to increase this to around two fifths of those in need[747, 791]. Even among children who are referred, access does not consistently translate into treatment.

The Children’s Commissioner reports that, among almost one million children referred to NHS-funded mental health services during 2023–24, only 36% entered

treatment, while almost one third were discharged before treatment began. Active referrals exceeded one million for the first time during 2024–25[748, 757].

Among adults, NHS Talking Therapies represents one of the strongest examples of evidence-based implementation within the NHS, yet it still reaches only a minority of people who meet criteria for anxiety disorders or depression. Economic evaluations of psychological therapies report favourable returns on investment. A WHO analysis estimated returns of up to around 5.7:1 when health gains are included, and 2.3–3:1 on productivity gains alone, alongside meaningful improvements in employment[792].

National access and waiting-time standards are measured to a person’s first treatment appointment, which is often a lower-intensity guided self-help intervention appropriate for milder presentations. Waiting times to step up to high-intensity therapy, the qualified-therapist interventions recommended for more severe or non- responding presentations, are typically much longer, around two to three times those for low-intensity support in one detailed pathway analysis. It is at this stage that commissioning and workforce shortages have their greatest impact[793]. Figure 15.2 therefore illustrates the point in the pathway where delays are greatest: a person may meet the headline waiting-time standard while still waiting many months for the treatment their needs actually require.

Figure 15.2 Median wait from assessment to high-intensity treatment in NHS Talking Therapies, by integrated care board, November 2025.

Integrated Care Board Source: NHS England, Monthly NHS Talking Therapies Programme Development Tool, Future NHS Platform.

For children and young people, common mental health conditions account for the greatest volume of need across the population, yet most never reach specialist services, and reaching a service does not guarantee that the underlying difficulty will

be recognised or treated. The Stadia trial illustrates this. Only around one in eight children referred to community mental health services in England received a clinician-recorded diagnosis within a year, even though around two thirds of those assessed screened at very high probability of a common mental disorder, and fewer than half of referrals were accepted[794].

A different picture emerges in adult services, where substantial investment in NHS Talking Therapies has created one of the clearest pathways to evidence-based treatment within the NHS. Although coverage remains below estimated need, the programme represents an important example of large-scale implementation of evidence-based care. Significant gaps remain elsewhere in the system, particularly for people with severe mental illness, complex and co-occurring needs, and those described in this and Chapter 14 as the ‘missing middle’.

The treatment gap becomes particularly marked for people with severe and complex needs. Only around 2.7% of people receiving adult community mental health services for psychosis, bipolar disorder or complex emotional needs received a NICE-recommended psychological therapy during the year to June 2026[795]. Similar gaps exist for people with self-harm, eating disorders and co-occurring substance misuse[796, 797]. For eating disorders, only around 30–32% receive formal treatment while eligibility thresholds may disadvantage people who are not underweight despite clear clinical need. Several factors appear to contribute to this treatment gap. Historically, many services have prioritised medication management, risk management and care coordination over the delivery of structured psychological therapies, particularly where workforce capacity is constrained. National investment and policy attention have also focused disproportionately on common mental health conditions, while access to specialist therapies for people with severe mental illness has expanded more slowly.

Where national guidance and standards are accompanied by dedicated commissioning, sustained investment, and transparent reporting of access, treatment and outcomes, however, implementation improves.

The Review also identified important limitations within national guidance itself. Guidance remains comparatively underdeveloped in relation to prevention, early intervention in the first years of life, complexity, co-occurrence and deprescribing. Guidance relating to active monitoring, the management of multiple co-occurring conditions and the systematic review of long-term medication also remains limited. Although guidance for common mental health conditions in children and young people has been updated in recent years, important gaps remain and the evidence base is less well developed across several areas of childhood mental healthcare.

Nevertheless, evidence-based guidance exists for a wide range of mental health conditions. Where treatment options are limited, services may become organised around the interventions they are able to provide rather than those most appropriate to the individual’s needs.

“Talking therapies is not always adapted to the individual persons need. I am autistic with mental illness and I needed to talk to someone about trauma. I self-referred as needed talking therapy but was offered course of telephone CBT as the form I filled in said I was anxious and that’s what they offer anxious people. CBT didn’t work for me. So after the course finished I was told to wait three months then self-refer again.”

Lived Experience Partner Consequently, many children and young people with common mental health conditions continue to be supported primarily by general practice, schools, families and community services, despite difficulties that would ideally be addressed through timely access to evidence-based specialist intervention. With co-occurrence being common (Chapter 9), services that do not reflect the complexity of people’s presentations are unlikely to provide effective care. This affects not only children and young people, but also the parents and carers who support them. Parental stress may be exacerbated by fragmented and poorly coordinated services, particularly where families feel disempowered[783]. Important gaps in provision also remain. For example, specialist early-years mental health services are available to only around 4% of the infants and families estimated to need them[798].

Figure 15.3 summarises the psychological and physical interventions currently recommended by NICE for children and young people’s mental health conditions. Although important gaps remain in areas such as prevention, complexity and co- occurrence, the figure illustrates that the principal challenge is often implementation of existing guidance rather than its complete absence.

Figure 15.3 Psychological and physical therapies recommended by NICE for children’s and young people’s mental health conditions. Neurodevelopmental pathways (ADHD and autism) differ in nature and are considered separately in Chapters 7 and 8.

ABFT – Attachment-based family therapy; AFP-AN – Adolescent-focused psychotherapy for anorexia nervosa; BPI – Brief psychological intervention; CBT-ED – Eating-disorder-focused CBT; CBT-TF – Trauma-focused Cbt; Da – Dietary advice; FB-IPT – Family-based interpersonal psychotherapy; FLX – Fluoxetine; FT – Family therapy; G-SH – Guided self-help; MED – Medication; MST – Multisystemic therapy; NDST – Non-directive supportive therapy; PP – Psychodynamic psychotherapy; PS – Problem-solving therapy; Psych-ED – Psychoeducation; PT – Parent training; WW – Watchful waiting.

The Review concludes that improving diagnostic pathways alone will not be sufficient. Recognition of need, assessment, diagnosis and access to treatment must connect reliably into a coherent pathway of care. This has important implications for the future organisation and delivery of mental health and neurodevelopmental services.

15.2 Stepped care, matched care and needs- led care

If access alone does not guarantee effective treatment, then the way services organise care becomes critically important. This section examines the principal models currently used and considers what the evidence suggests about their strengths, limitations and implementation.

In stepped care, individuals generally begin with the least intensive intervention and progress to more intensive support only if they do not respond. Matched, or stratified, care instead seeks to direct people from the outset towards the level of intervention most appropriate to the nature and severity of their difficulties. Needs- led care goes further still by determining support according not only to diagnosis or symptom severity, but also to functioning, context, risk, preferences and wider circumstances.

The evidence suggests that purely sequential stepped care too often becomes a system in which people must deteriorate or fail to improve before they qualify for additional help. Professionals working across children’s and young people’s services between 2015 and 2025 repeatedly described a deterioration-gated system. Children whose needs fell below service thresholds frequently received little or no support, their difficulties worsened while they waited, and only then did they become eligible for intervention. As one participant observed, young people are “having to get more unwell … to be able to use a Tier 3 service because there’s nowhere else to go”[745]. More recent evidence supports this observation: as Figure 15.1 shows, the proportion of people whose mental health deteriorates while waiting rises steeply with increasing waiting time[19]. The Review also heard that, where specialist intervention was unavailable, families were often directed towards emergency departments, crisis services and voluntary-sector helplines. Although these services play an important role, contributors did not regard them as substitutes for timely access to evidence-based assessment, treatment and continuing care. The consistency of these observations across professional groups, multiple sources of evidence and ten years of service experience makes this one of the strongest qualitative findings presented in the chapter[799].

“Our 13-year-old daughter started to self-harm and was referred to CAMHS who turned her down as she didn’t meet the criteria and was not considered ill enough so was referred to a lower-threshold support charity, but the waiting list was 6 months.

Within the next month our daughter’s mental health got worse and we ended up in A & E after she took an overdose. She was then accepted

by CAMHS! Why couldn’t we have got support earlier that might have prevented a crisis?”

Lived Experience Partner There is also direct experimental evidence that outcomes depend not only on how quickly treatment is accessed, but on how appropriately it is matched to need. A cluster randomised controlled trial in adult psychological therapy services in England compared stepped care, in which patients begin with low-intensity support and step up only if they do not respond, with stratified care, in which patients are matched at assessment to either low- or high-intensity treatment. Stratified care produced significantly higher rates of reliable and clinically significant improvement in depression (52% versus 45%; odds ratio 1.40), at a modest additional cost reflecting more people being directed to high-intensity treatment from the outset, and was judged likely to be cost-effective[800].

Within children’s and young people’s mental health services, there has similarly been a gradual shift away from traditional age- and severity-based tiers towards models organised around identified need, most notably Thrive[801]. However, the Review found that changes in language have often outpaced changes in practice. Professionals continued to describe services operating according to tiers beneath a new vocabulary, while families reported that allocation to different needs groups often remained opaque and difficult to understand. This work was commissioned specifically for the Review. Although not yet peer-reviewed, it represents one of the largest primary studies of English children’s and young people’s mental health services undertaken for the Review and will be published in full alongside it. National quasi-experimental evidence indicates that, where implemented, needs-based approaches improved accessibility and shared decision-making[802]. However, they produced no statistically significant overall change in service characteristics, with improvements concentrated in systems that already demonstrated strong multi- agency collaboration[803].

A needs-led approach therefore requires not only access to evidence-based interventions, but also shared decision-making, flexibility and recognition of individual goals, preferences and circumstances.

“The services never saw me as an individual, a person with passions, interests, a personality, rational thoughts and opinions. Instead everything became ‘the eating disorder’. I lost myself and I think they lost me. The text-book model of approach didn’t work for me and there was no room or curiosity to think outside of the box. They simply saw the illness, but not the years of struggles leading up to this point, and when I tried to tell them they couldn’t or wouldn’t hear me.”

Lived Experience Partner Needs-led care also depends on sufficient capacity. Without it, the model becomes unsustainable. Some services that moved to lower-threshold, needs-led access achieved short waiting times, while others became overwhelmed and reverted to severity-based eligibility, with one participant reflecting that “we went threshold- less… and it failed”[745]. Aspirations cannot substitute for infrastructure. Contributors consistently emphasised that any new approach should simplify rather than complicate care, support professional judgement rather than create additional bureaucracy, and demonstrably improve outcomes.

15.3 Workforce and implementation: right people, right place, right time

Every service model described in this chapter ultimately encounters the same constraint: delivery depends on people. Workforce is therefore the principal limiting factor in implementation. The National Audit Office identified workforce shortages as one of the major risks to successful implementation[791]. However, this is only part of the picture. The challenge is not simply that there are too few practitioners, but that those who are trained are often not funded or deployed to deliver the evidence- based interventions for which they were trained[804].

The Review also heard that effective care depends on a broader workforce than is typically captured within traditional professional groups. Peer support workers, family support workers, youth workers and other non-clinical practitioners can play an important role in engagement, trust, participation and ongoing support, complementing rather than replacing formal clinical interventions.

The Review identifies three broad workforce challenges.

First, there are too few staff to meet current levels of need. These shortages reflect not only rising demand but also longstanding challenges in workforce planning, training and investment.

Second, investment in training clinicians to deliver evidence-based interventions does not automatically translate into greater treatment capacity. Across children’s and young people’s services, substantial investment has been made in training clinicians to deliver specialist psychological therapies. Yet many subsequently work in generic roles that leave little opportunity to provide the interventions for which they were trained. The same pattern is evident in adult services.

Third, as co-occurring neurodevelopmental, mental health and social difficulties become increasingly common, services require both specialist intervention skills and broader capabilities in formulation, consultation and integrated care. One of the principal developments during the past decade has therefore been the growth of a wider workforce surrounding a relatively unchanged specialist core. Children’s Wellbeing Practitioners, Education Mental Health Practitioners (EMHPs) and Mental Health Support Teams (MHSTs) have expanded access to lower-intensity interventions across schools, primary care and community settings, supported by consultation and supervision models in which specialist expertise enables a wider workforce. This represents an important and promising development. Where training is fully funded, posts are protected and career pathways are clear, these roles can mature into a sustainable workforce pipeline.

The Review identified two broader workforce design principles that recur across services.

First, scarce specialist expertise is most effectively deployed through consultation, supervision and shared decision-making rather than exclusively through direct case- holding. Specialist knowledge has greatest impact when embedded within multidisciplinary teams. Seniority need not be medical: experience, a broad understanding of local services, and familiarity with the evidence base may matter more than professional discipline. One trust that substantially reduced waiting times over two years attributed this largely to placing senior, experienced staff at the front door, enabling rapid direction to the most appropriate support. Second, pathways organised around single diagnoses risk progressively narrowing professional expertise, encouraging clinicians to determine whether someone meets criteria for a particular condition rather than developing, with the wider team, a formulation and management plan for the whole person.

The Review also identified a shift in the skills required for effective practice. This includes skills in engagement, co-production and partnership working with parents, carers and families, which are increasingly important in needs-led approaches to care and support. Joint decision-making, the ability to work across organisational boundaries and effective care coordination become increasingly important in guarding against fragmented care and confusion for individuals and families, particularly during periods of significant distress.

The Review recommends that need, rather than diagnosis, should be the starting point, leading to a coordinated plan for support and treatment. In systems where diagnosis becomes the endpoint, support for need does not necessarily follow.

Another important workforce lesson concerns how access is organised. The Review repeatedly heard that triage and access functions work best when experienced senior clinicians are available at the front door rather than later in the pathway.

Effective delivery also requires a workforce extending well beyond specialist mental health services. Primary care, neighbourhood and universal services, together with the voluntary, community, faith and social enterprise (VCFSE) sector, all make important contributions. Health visiting, a statutory component of the Healthy Child Programme and a cornerstone of infant mental health, experienced approximately a 21% reduction in staffing between 2016 and 2021, accompanied by substantial reductions in funding and increasing caseloads[805].

General practice also plays a central, though often overlooked, role. For many adults, primary care is the first and most consistent point of contact for mental health and neurodevelopmental concerns, providing assessment, treatment, medication management, referral, shared care and continuity across the life course. Strengthening the interface between primary care and specialist services through advice, consultation and shared care is therefore central to any workforce strategy.

Schools have likewise become an increasingly important part of the workforce supporting children’s mental health, as discussed in Chapter 13. Survey evidence reviewed there showed that informal and semi-formal support provided by teachers and other school staff was one of the most frequently accessed sources of help and was rated as helpful by 82% of those receiving it. By comparison, fewer than 6% reported accessing CYPMHS, and 55% described that support as helpful[14, 696]. Young people are therefore already seeking support from trusted adults within schools – not only teachers, but also counsellors, family support workers, parent support advisers, inclusion staff and other pastoral practitioners. The question is not whether schools play this role, but whether staff have sufficient training, time, supervision and access to specialist consultation to undertake it confidently and sustainably. Supervision, in particular, is not yet an established part of school culture, leaving staff to manage emotionally demanding situations with limited support.

The central lesson is that workforce effectiveness depends on more than headcount. It depends equally on how expertise is organised, supported and connected across systems. Training increases access to evidence-based treatment only when jobs are designed to allow clinicians to deliver the interventions for which they have been trained, workforce planning explicitly protects treatment capacity, and supervision and governance sustain high-quality practice. Where these conditions are absent, specialist expertise is gradually absorbed into generic activities such as assessment,

care coordination and risk management. The system may successfully create capability, but the treatment capacity it intended to build never fully materialises.

15.4 Commissioning, contracting and accountability

How services are commissioned strongly influences what they deliver. Evidence reviewed for this chapter describes a system in which contracts often purchase activity rather than outcomes. Although organisational responsibility shifted from clinical commissioning groups to integrated care boards, many contributors felt that incentives changed much less than organisational structures. Contracts commonly specified volumes of activity that were only partly delivered, while effective commissioning depended on more than contractual reporting. It also required direct engagement with providers, service users and families alongside routine performance data.

This is particularly evident in the expanding market for neurodevelopmental assessment. Under the combined influence of Right to Choose, discussed in Chapter 14, and independent-sector provision, a substantial proportion of NHS- funded assessment now takes place through providers operating beyond direct local service management. This makes it more difficult for local systems to understand activity, follow-up care and outcomes. Evidence gathered by the Review suggests that a relatively small number of large independent providers completed around 220,000 NHS-funded assessments within a single year. As discussed in Chapter 14, this represented more than half of NHS-funded ADHD assessments and a substantial proportion of autism assessments, with activity varying enormously between providers, from fewer than 100 assessments annually to more than 75,000. Growth in diagnostic activity has not been matched by investment in the capacity needed to respond to identified need, including post-diagnostic support, treatment and workforce development. Contributors repeatedly described assessment and diagnosis expanding more rapidly than services able to provide continuing intervention.

Within NHS-commissioned services, these pressures fall unevenly. The median wait from referral to first contact is longer for neurodevelopmental pathways – principally autism and ADHD assessment – than for other secondary mental health services, and varies considerably between regions. Many people have therefore used Right to Choose to obtain assessment from independent providers, where waiting times are shorter. The concern is not assessment itself, but that rapid growth in assessment activity across both NHS and independent providers has not been matched by investment in the treatment and support that should follow.

Right to Choose, as a statutory entitlement rather than a commissioned service, also creates an accountability challenge. Providers are generally not required to submit the routine data that commissioners can require through conventional contractual arrangements. Among approximately 120 ADHD service providers, only around 10 submitted data. As Chapter 14 describes, Right to Choose expanded access without corresponding implementation funding or data infrastructure, so growth in activity outpaced arrangements to monitor and assure quality.

The Review also identified substantial variation in prices for apparently equivalent assessments, with tariffs differing by around fourfold between integrated care boards[18]. This appears to reflect the absence of a consistently evidence-based tariff rather than systematic differences in clinical complexity. Current non-mandatory tariffs may unintentionally incentivise combined ADHD and autism assessments because reimbursement is higher irrespective of clinical indication. Financial analyses also showed marked variation in provider operating margins, with the highest approaching one third and exceeding those typically reported by NHS trusts operating close to break-even.

GP acceptance of shared-care arrangements for medication initiated by independent providers has fallen to an average of around 62% and continues to decline because of concerns about assessment quality and the availability of ongoing specialist support. For many people, assessment and treatment recommendations do not reliably lead to continuing care. Contributors described uncertainty about responsibility for ongoing management when different parts of the pathway were delivered by different organisations, and raised concerns about continuity of treatment where shared care could not be established or where people moved between systems. Some people resort to private prescriptions in order to maintain treatment, creating inequalities between those able to pay and those whose treatment is disrupted or discontinued. Diagnosis can therefore become the beginning of another wait rather than the start of treatment.

The concern is not increased access to assessment, but whether commissioning arrangements adequately support accountability for assessment quality and for the continuity of care that follows. The Review therefore recommends that provision funded through Right to Choose and the independent sector should be held to the same standards as directly commissioned services, including a common minimum dataset, consistent pricing, agreed service standards and models of care, and quality and outcome measures as conditions of funding and registration. Qualitative evidence and structural mapping indicate that commissioning linked to clearly specified service models, accompanied by dedicated funding and explicit performance expectations, changes what services deliver. Contributors consistently argued that commissioning should focus on the needs of whole populations rather

than only those meeting service thresholds and should measure outcomes and processes of care across pathways rather than activity within individual organisations.

Children’s and young people’s eating disorder services, supported by a national access and waiting-time standard, together with Early Intervention in Psychosis services, were repeatedly cited as examples of pathways that had benefited from being “paid for and contracted to do a particular job”, with integrated care boards attaching resources and measurable expectations. By contrast, adult eating disorder services lack an equivalent standard and continue to show a substantial treatment gap. These examples demonstrate that, where evidence-based service models are supported by dedicated funding, clear specifications and transparent expectations, implementation is much more likely. The implication is that commissioning should focus on the delivery and outcomes of clearly specified, evidence-based interventions rather than activity alone.

15.5 Community and neighbourhood models

Three broad directions emerge from the evidence reviewed in this chapter. The first, and most important, is strengthening core NHS services through the consistent delivery of evidence-based treatment, supported by the workforce and commissioning needed to provide it. The second, considered here, is to strengthen and expand support beyond clinics into neighbourhoods, schools and communities. The third, discussed in the following section, explores how digital and other non- clinical approaches can extend the reach of support. None is a substitute for the others. Community, digital and non-clinical approaches should complement rather than replace effective core services.

15.5.1 Schools, youth services and the erosion of community capacity

Over the past decade, schools have become the principal community platform for supporting children’s mental health. However, this shift raises important equity concerns, as more affluent areas and better-resourced schools are often able to secure additional support beyond that available elsewhere. At the same time, much of the wider community infrastructure surrounding schools has weakened. Contributors repeatedly linked this to reductions in youth services and the youth work workforce, which have traditionally provided trusted relationships and early support for young people less likely to engage with formal mental health services. As these forms of community support have diminished, increasing pressure has fallen on schools, primary care and specialist services. The reduction of Sure Start, youth services and community centres was consistently identified as contributing to

growing demand for specialist services, with the greatest impact falling on the most disadvantaged communities. The Review heard a consistent view that strengthening specialist services alone cannot compensate for the loss of earlier community-based support.

15.5.2 Open-access hubs and voluntary sector models

The most consistently valued model identified across the evidence was the open- access, self-referral youth hub, frequently delivered by voluntary-sector organisations or through partnerships between the NHS and VCFSE organisations. Value, however, should not be confused with demonstrated effectiveness. The formal evidence base for these models remains limited and, where evaluations exist, they are generally cautious.

The Review also heard concerns about the sustainability and governance of some community-based services. Contributors described increasing numbers of young people presenting with needs that exceeded what these services had been designed to provide. Funding insecurity was another recurring concern, with some successful community services continuing to depend on short-term funding despite becoming established components of local pathways. This picture is not universal, however. Some hubs – including well-established early-support services and Young Futures models – are securely funded, co-produced with young people and supported by clinical expertise. Contributors also highlighted the personal toll on those delivering such services, many of whom are volunteers or people with lived experience who may themselves be exposed to high levels of distress without adequate supervision or support.

15.5.3 Neighbourhood health and multidisciplinary models

The emergence of what is now described as ‘neighbourhood health’ reflects an attempt to rebuild this local infrastructure. The Review’s mapping demonstrates that many of these developments were already underway before neighbourhood health became part of national policy. Family hubs, early-help hubs, prevention hubs, open- access sanctuaries and community-based multi-agency services all expanded substantially during the period reviewed, with community-located multidisciplinary provision almost doubling. Many of these developments share common features, including community-based support, easier access to advice and intervention, and greater integration between mental health, physical health, social care, education and voluntary-sector provision. NHS England has sought to embed these principles through guidance on neighbourhood multidisciplinary teams for children and young people, emphasising integrated working, timely access to specialist advice – including paediatric and mental health expertise – and primary care-led coordination

of support[806]. This model is distinct from the 24-hour neighbourhood mental health centres being developed for adults.

The Review examined a range of emerging models in the commissioned study described earlier, including enhanced primary care in which GPs work alongside youth workers and mental health practitioners with specialist support; neighbourhood multidisciplinary teams integrating physical health, mental health and neurodevelopmental expertise; and youth social prescribing, for which early evidence from pilot evaluations suggests a social return on investment of around £5 for every £1 invested, alongside improvements in wellbeing and reductions in loneliness among participating young people[807]. The Review did not conduct a systematic survey of community provision because no locality was able to demonstrate robust outcome data for these models. The approaches described here are therefore those that recurred most consistently across the ten local systems mapped for the Review and in evidence submitted to it.

Collectively, these approaches offer an important opportunity to address the ‘missing middle’ identified earlier in the chapter, while providing continuity across the transitions where support is often lost. In many areas, general practice is expected to provide the organisational centre of these neighbourhood arrangements, bringing together mental health, physical health, social care and voluntary-sector support. Their effectiveness will therefore depend heavily on whether primary care receives the workforce, training and infrastructure needed to fulfil this expanded role.

However, the creation of neighbourhood teams should not be confused with increased access to treatment. These models may improve access to support and coordination, but their contribution to reducing unmet need will ultimately depend on whether they expand access to evidence-based interventions rather than simply creating additional points of contact within the system.

15.5.4 Structural limitations and fragmentation

Community services have expanded largely as additional layers within the system rather than as integrated components of a redesigned pathway. Hubs, voluntary- sector organisations, sanctuaries and digital platforms have frequently developed their own routes of entry, yet few systems have established routine mechanisms for coordinating them with NHS services.

Too often, these connections depend on referral pathways, professional goodwill and local relationships rather than shared allocation processes, integrated records or common accountability. The direction of travel therefore points towards whole- system approaches in which community provision is planned, resourced and held accountable as an integral part of the local system rather than as a parallel set of

services[808]. The fragmentation described in Chapter 14 has therefore not disappeared; it has been reproduced at neighbourhood level.

15.5.5 Evidence base, inequalities and evaluation

The evidence base remains comparatively limited. School-based provision has become embedded within national policy but has been evaluated only to a limited extent in terms of outcomes (see Chapter 13). Across the ten local systems mapped for the Review, all serving children and young people, no locality was able to demonstrate robust outcome data for community models as a whole, and referrals into these services were rarely tracked systematically after signposting.

Important inequalities also remain and may have been reinforced by the uneven development of community provision. Embedding provision within schools during a period of sustained budgetary pressure risks widening inequalities, since better- resourced schools are often able to offer more than those serving the most disadvantaged communities. The erosion of youth services and community centres has likewise fallen hardest on poorer areas. Children who are home educated or out of school may lose access to school-based support altogether, an issue considered further in Chapter 13.

For community and neighbourhood models to function effectively, they are likely to require access to psychologically skilled and experienced practitioners who can support non-specialist mental health practitioners and peer support workers. Access to such expertise is itself unevenly distributed. Independent and well-resourced schools are often able to purchase additional services directly, whereas provision within the state sector continues to vary substantially according to local resources, geography and school type.

MHSTs were introduced to reduce precisely this inequality and, by Spring 2026 were working in 48% of schools and colleges in England, covering around 60% of pupils and learners, with full national coverage intended by 2029/30[809]. Coverage, however, remains uneven. Around four in five pupils in state secondary schools are covered, compared with fewer than half in special schools, alternative provision and post-16 settings. Consequently, the settings serving children and young people with the most complex needs are the least likely to have access to an MHST. Completing coverage in these settings with an enhanced MHST offer should therefore be the first priority for the remaining expansion rather than the last: they serve the smallest number of learners, which makes them the least costly part of the rollout to finish, and the highest proportion of those at risk of leaving education altogether. Although Chapter 13 reports encouraging observational outcome data for children receiving MHST interventions, national evidence of programme-level effectiveness remains

limited because rollout has been staged and incomplete and early sites were selected rather than allocated.

National guidance has concentrated primarily on which interventions work rather than also on how services should be organised to deliver them. As NICE increasingly moves away from service-level recommendations, the Review suggests that greater attention should be given to evaluating and guiding the service models on which successful implementation ultimately depends.

15.5.6 Conclusion on community and neighbourhood models

Community and neighbourhood models represent perhaps the strongest convergence of professional opinion identified by the Review, particularly for children and young people, although the underlying principles apply across the life course. They align closely with the needs-led approach developed throughout this Report and offer a credible means of reducing fragmentation and strengthening continuity of care. Trusted relationships and continuity emerged consistently throughout the Review as among the strongest determinants of whether support was experienced as effective.

At the same time, these models remain among the least systematically evaluated, most unevenly distributed and most organisationally fragile components of the current system. They do not in themselves guarantee increased access to evidence- based treatment; this must be an explicit feature of their design. Their future success is therefore likely to depend less on creating entirely new models than on connecting those that already exist through shared allocation processes, interoperable information systems, clearer collective accountability and sustainable funding.

They also require the evaluation infrastructure needed to allow future policy to be guided by evidence as well as professional judgement. That infrastructure must support integrated working rather than obstruct it. Under current waiting-time rules, a young person’s referral clock stops only when a care contact, a baseline outcome measure, and an agreed care plan or coded intervention have all been recorded within the national dataset – something that voluntary-sector or digital partners generally cannot do, and in some cases are explicitly prevented from doing. Support delivered outside the NHS therefore does not register as treatment having begun, and a provider’s reported performance may actually worsen because of the partnership that reached the young person first. The Review heard of one trust considering withdrawal from an effective voluntary-sector partnership for this reason alone. Systems that cannot record what their partners contribute are unlikely to sustain working with them.

Their success will also depend on their ability to reduce, rather than inadvertently reinforce, existing inequalities in access to support, specialist expertise and evidence-based treatment.

15.6 Digital technologies, AI and non-clinical support

The evidence reviewed in this chapter suggests that the most helpful way of understanding digital provision is to distinguish three distinct functions. Digital tools may support wellbeing directly through information, self-management, peer support and symptom monitoring; they may improve access by helping people navigate services, self-refer and be triaged; or they may contribute directly to clinical care by supporting assessment, diagnosis, treatment and ongoing monitoring. These roles bring different opportunities and different risks. The greatest clinical and regulatory challenges arise where digital systems influence diagnosis, treatment or safety.

Across all three functions, however, the evidence points to the same conclusion: digital technologies have considerable potential, but successful implementation depends not on the technology alone, but on the systems within which it is deployed. This includes robust governance, workforce capability, data infrastructure and clear arrangements for people who are unable or unwilling to access digital forms of support.

15.6.1 Recent developments and a changing evidence base

Earlier digital initiatives largely sat alongside existing services and focused on alternative ways of delivering traditional care, including online counselling, video consultations, text-based support and chat services. Although these approaches increased convenience and access, they largely replicated existing clinical interactions through digital channels rather than becoming integrated components of care pathways.

More recent developments take a different approach. Increasingly, digital technologies are becoming part of the pathway itself. This shift is supported by a growing evidence base. Regulated digital therapeutics now carry conditional NICE recommendations, and emerging evidence suggests that, when carefully implemented, digital technologies can improve both efficiency and equity. At the same time, digital innovation risks reproducing wider social inequalities. Access to devices, connectivity, private space and digital literacy remains unevenly distributed, meaning that those most in need of support are not always best placed to benefit from digitally enabled care.

15.6.2 Risks and implementation challenges

The same evidence also illustrates why implementation requires caution. For example, where machine-learning models are trained on historical clinical data that reflect existing inequalities, those inequalities may be reproduced in the model’s recommendations. Digital systems therefore have the potential either to reduce or to amplify existing inequalities, depending on how they are designed, trained and governed. The same principle applies to digital triage more generally.

One evaluated triage system demonstrated genuine technical strengths, yet its operational value was substantially limited because triage thresholds were set so narrowly around crisis that most people with significant functional impairment were directed into the same undifferentiated waiting pool. The effectiveness of digital tools therefore depends not only on their technical performance but also on the service rules within which they operate.

This raises an important question. If effective digital tools increasingly exist, why have so few become established at scale? The evidence reviewed suggests that the principal barriers lie not in the technology itself but in its implementation. Contributors described regulatory processes that slowed innovation, fragmented procurement across multiple integrated care boards, the absence of funding for implementation and long-term maintenance, insufficient investment in implementation, variable digital confidence across the workforce, and inconsistent evidential standards.

The increasing use of AI-enabled systems also raises important questions about data governance, transparency, privacy and accountability. Contributors described insufficient clarity around commissioning, evidential standards and clinical governance as the use of these technologies accelerates. As AI-supported approaches become more widespread, clear national policy and governance frameworks are likely to become increasingly important.

15.6.3 Non-clinical interventions

Alongside clinical services and digital technologies, the evidence points to the importance of a wider group of non-clinical interventions that contribute to mental health and wellbeing. For many children, young people and families, these may represent the most appropriate first response, particularly where distress is closely linked to social circumstances, isolation, practical difficulties or reduced participation in education, employment or community life. They are not substitutes for specialist treatment where clinical intervention is required, but provide proportionate support that can enhance wellbeing, strengthen resilience and, in some cases, prevent escalation of needs.

The evidence supporting these approaches is stronger than is sometimes recognised. Reviews consistently identify positive associations between mental health and participation in arts and creative activities, regular physical activity, access to green and blue spaces, and organised youth activities and volunteering. Within this broader ecosystem, digital technologies may prove particularly valuable in connecting people to appropriate clinical and non-clinical support and helping to coordinate care across them.

15.7 Secure settings, children’s homes and complex needs

This section focuses on the individuals the system most often loses (see also Chapter 17). The Review found that, although provision for these groups existed in almost every area, it expanded very little over the period studied. While overall mental health provision across the mapped systems increased by around one quarter, services for looked-after children, children with learning disabilities and those in youth justice remained largely unchanged, while provision for young people with substance misuse difficulties contracted. Growth occurred predominantly at the front end of the system rather than for those with the greatest and most enduring needs. This section therefore focuses on children and young people in secure and residential settings.

15.7.1 The scale and nature of need

Single-diagnosis prevalence figures do not adequately describe the complexity of need in these settings. Children in the secure estate experience substantially higher rates of mental health and neurodevelopmental conditions than the general population, and those identified typically present with multiple, persistent and significant difficulties shaped by adversity, family disruption and inequality. Co- occurrence is the norm rather than the exception. Within neurodevelopmental conditions this is well established[810], and overlapping needs are particularly pronounced in justice-involved populations, where children frequently present with combinations of ADHD, communication difficulties, learning disabilities and traumatic brain injury[312, 406]. Of children in the secure estate surveyed in 2024–25, 65% reported they had previously been in local authority care[811], and most have experienced extensive adverse childhood experiences alongside significant instability in care, relationships and accommodation (see Chapter 17).

Entry into secure settings rarely marks the beginning of a child’s difficulties. More commonly, it represents the outcome of a prolonged developmental trajectory in which needs have been overlooked, misunderstood or responded to punitively. Early difficulties frequently emerge in schools, health services, family support or wider

community settings, and subsequent pathways may be shaped as much by how those needs are interpreted and responded to as by the needs themselves. The school-to-justice pathway illustrates this clearly. Children with SEND make up approximately 15% of the school population yet account for 46.7% of permanent exclusions, while exclusion is associated with at least a doubling in the risk of later convictions for serious violence compared with matched peers[812, 813].

A substantial body of evidence demonstrates that neurodivergent children and young people are disproportionately represented within youth justice and secure settings. This reflects not only underlying conditions, but also the interaction between neurodevelopmental needs and wider social, educational and environmental disadvantage. The children at greatest risk of entering the secure estate are those who have accumulated disadvantage across health, education and social care over many years, and whose needs have not been recognised or responded to adequately. Historically, those needs have too often been addressed through single- agency or single-diagnosis responses. This can produce what has been described as a ‘cumulative continuity of disadvantage’, in which children are repeatedly misunderstood and then inadequately accommodated within the systems intended to support them.

Family support, including the role of parents, carers and corporate parents, may be particularly important in interrupting these trajectories. Evidence reviewed elsewhere suggests that effective parenting and family support interventions improve behavioural, developmental and attachment-related outcomes and may represent an important protective factor across the pathway into secure care.

Disadvantage accumulates as children move through systems, and at each stage there are opportunities either to interrupt or reinforce these trajectories. Fewer than 1% of children are in local authority care at any one time, yet care experience becomes increasingly common further along these pathways. Around one quarter of adult prisoners report having been in care as children[814], and care-experienced children are substantially over-represented in pupil referral units and alternative provision. The consequences of responding late rather than early are considerable. Across children’s services in England and Wales, the wider societal costs of late intervention have been estimated at around £17 billion annually[815].

15.7.2 Whole-system approaches

Whole-system approaches to secure care represent a response to this complexity and accumulation of need. The most developed example in England is the Framework for Integrated Care (FfIC; delivered through Secure Stairs), developed jointly by NHS England, HM Prison and Probation Service, and the Department for Education. The framework is grounded in the principle that children’s

needs should be understood in the context of their life experiences, relationships, adversity, strengths and wider system context, rather than through mental health or neurodevelopmental lenses alone. Rather than focusing on isolated assessment and treatment, it seeks to strengthen the entire network of relationships surrounding the child. A co-produced formulation, developed through understanding the child’s background story (‘My Story’), sits at the centre of the framework and informs shared care planning across the workforce.

Research suggests that effective models for children in contact with the justice system are informed by attachment-based, trauma-informed and social-pedagogic approaches, with the quality of everyday relationships acting as a central mechanism of change rather than discrete therapeutic programmes[816-819]. Independent evaluation suggests that successful implementation depends less on introducing new interventions than on achieving cultural change across services. Such approaches appear easier to implement in secure children’s homes than in young offender institutions, with consistent evidence of improvements in staff knowledge and confidence, although evidence for reductions in violence remains preliminary[820].

These approaches share another important characteristic: they recognise that children do not experience care through professional or organisational silos. Children consistently describe their strongest and safest relationships as being with the staff they encounter every day, and emphasise the importance of healthcare working with, rather than separately from, those relationships. They are critical of systems in which care and control are treated as opposing functions, or where healthcare is delivered independently of the relational context in which they live.

15.7.3 Identification, data and health inequalities

The Review also identified opportunities to improve how need is recognised. The Healthcare Standards for Children and Young People in Secure Settings[821] are intended to ensure that children receive healthcare matched to their individual needs, and every child entering the secure estate begins a health assessment within the first ten days of admission using the Comprehensive Health Assessment Tool (CHAT), a validated screening instrument[406].

For children with multiple and co-occurring needs, diagnosis-led approaches may obscure the cumulative impact of adversity, neurodevelopmental differences, educational difficulties and mental health needs. This reinforces the importance of formulation-based approaches that seek to understand the combined impact of multiple needs rather than focusing on individual diagnoses in isolation.

The resulting data give cause for concern. CHAT data collected at admission to the secure estate, presented in the Chief Medical Officer’s 2025 report, recorded autism

in only 0.1% of children entering secure settings – far below general population prevalence and inconsistent with research evidence – while 35% were identified as having ADHD[251]. Despite this level of identified need, and despite the availability of effective treatments[260], provision varies widely. Waiting times for diagnosis can be prolonged[251, 822], and some sites report limited support or intervention following diagnosis, including access to ADHD medication[251].

Health inequalities are also evident in recognition itself. Children from minoritised ethnic communities are less likely to have their needs recognised early and remain over-represented within the justice system[823, 824]. Girls, although far fewer in number within secure settings, are similarly less likely to have their needs recognised early. Their difficulties may be overlooked or misinterpreted, so that support, where it is eventually provided, often arrives later and through different pathways.

Perhaps the clearest evidence considered by the Review came directly from children and young people themselves through a standing advisory group established over several years. Their accounts provided a striking contrast with systems organised primarily around monitoring and management. They described needing time to feel safe before meaningful engagement could begin and emphasised that behaviour should be understood as communication rather than defiance.

The implications for the wider chapter are clear. Children and young people with the greatest complexity require sustained, specialist and relational support. They should remain central, rather than peripheral, to future service reform.

15.8 Outcome transparency and learning systems

This chapter concludes where improvement itself begins: with measurement, learning and quality improvement. Without reliable information about the safety, effectiveness and experience of care, none of the arguments developed in the preceding sections can be tested. Over the past decade, services have become considerably better at collecting outcome measures than at using them. Compared with many areas of acute and primary care, mental health services have generally operated with less mature data infrastructure, fewer standardised outcome frameworks, and fewer incentives for routine data capture, analysis and reporting.

15.8.1 The scale of the problem

Paper-based systems, electronic records poorly designed for outcome measurement, incomplete paired observations and non-interoperable information systems have all limited the ability of data to inform clinical care.

Nationally, only around one quarter of closed cases in children’s and young people’s services have paired outcome measures, a figure reported at the beginning of the period under review and independently confirmed a decade later[825]. Some services have demonstrated that substantially higher levels are achievable within the current system. Provider organisations that invested in structured outcome monitoring achieved paired outcome completion rates well above the national average, demonstrating that the principal barriers are infrastructure and culture rather than feasibility.

This weakness has been compounded by performance management approaches that increasingly rewarded activity rather than outcomes, creating incentives, as one participant observed, focused largely on “churning kids through”. The Review’s mapping identified 134 new mental health services established across ten local systems over four years alongside the closure of 42 others yet found no consistent mechanism linking service performance to decisions about whether services expanded, adapted or closed. Public transparency was similarly limited. Children’s and young people’s services rarely reported outcomes routinely by ethnicity, socioeconomic deprivation or other protected characteristics, representing an important national gap in understanding equity as well as effectiveness.

15.8.2 Completing the feedback loop

The value of measurement depends not on collection alone but on completing the feedback loop. Where outcome data were actively returned to clinicians, children, young people and families – for example, “you said this and this is how we’ve improved it” – measurement became part of care itself. Where information travelled only upwards to commissioners or national datasets, it was often experienced as an administrative burden and a source of scrutiny rather than learning.

NHS Talking Therapies demonstrates that a different approach is possible. Through routine collection and publication of patient-reported outcomes at both local and national level, it has shown that systematic measurement – with paired outcomes collected for more than 98% of patients – can drive service improvement and inform investment in interventions that demonstrably improve outcomes.

Appropriate validated outcome measures already exist for secondary mental health services for both children and adults and should now be implemented more consistently[826, 827]. The Review’s proposed learning system, described more fully in Annex A, illustrates how these principles could operate at national scale. It envisages a federated data architecture in which detailed clinical records remain within local services while a nationally agreed minimum dataset supports continuous learning across the system. Importantly, the model is digital-first. Structured information generated routinely through pathway tools – including symptom

measures, stratification decisions, treatment engagement, adverse effects and patient-reported outcomes – becomes actionable intelligence rather than remaining trapped within individual products or organisations.

Existing precedents demonstrate that this approach is achievable. National diabetes registries covering more than 99% of people with diabetes in one UK nation have supported substantial improvements in quality of care while creating an invaluable research resource, and NHS Talking Therapies provides a well-established English example of outcome measurement that is routinely collected, interpreted and acted upon.

The broader implication extends well beyond data collection itself. A genuinely needs-led system cannot determine whether support has been successfully matched to need unless it routinely measures outcomes and learns from them. Learning systems also require information about harms, safety incidents, inequalities, access, engagement, and the experiences of children, young people and families. Measurement is therefore not an additional administrative requirement but one of the foundations of effective care.

15.9 Prevention

Everything to this point in the chapter has focused on how the system responds once mental health has begun to deteriorate. A system that only ever responds downstream, however, will remain permanently under strain. The volume of need described in Chapters 6 to 9, together with the environmental drivers discussed in Chapter 10, cannot be addressed by expanding treatment capacity alone. Prevention is therefore not an optional addition to the Review’s recommendations but the upstream complement to its central argument: that the system should organise itself around need, functioning and participation rather than diagnosis alone, and should act on the causes of need as well as its consequences.

Prevention aims to reduce the likelihood of mental ill health by addressing risk and protective factors across the life course before significant difficulties emerge, through action at population, community and individual level, including on the social, economic and environmental factors that shape mental health[828]. This section sets out the case for prevention as supported by the evidence. Two major recent syntheses in World Psychiatry – a review of the social determinants of mental health and their prevention[829], and a blueprint for preventive psychiatry in young people[830] – summarise the current evidence and are drawn on throughout (referred to here as the social determinants review and the preventive psychiatry review). In addition, The Lancet Psychiatry has recently published a consensus statement on youth

mental health[831]. Where the evidence is limited, we say so, because overclaiming would ultimately weaken the parts of the argument that are genuinely robust.

15.9.1 Why prevention matters

The scale of the burden. The Centre for Mental Health estimates the annual economic and social cost of mental ill health in England at around £300 billion, comprising approximately £130 billion in human costs, £110 billion in lost economic output and £60 billion in health and care expenditure[1]. This is around twice the annual NHS England budget and, in economic terms, comparable to absorbing a pandemic every year[1]. The largest share of these costs falls on individuals and their families, but the remainder is distributed across education, employment, social security, social care, criminal justice and the wider economy. This is therefore a cross-government argument for prevention: the costs do not sit solely within the health system, and neither do many of the policy levers capable of reducing them.

The timing of onset. Mental ill health is overwhelmingly a condition of early life. The median age of onset of mental disorders is approximately 14.5 years, and almost two-thirds first emerge before the age of 25[830]. The greatest returns are therefore likely to come from action during childhood, adolescence and the transition to adulthood.

The role of the environment. Economic insecurity, educational and employment conditions, the erosion of community and social connection, the effects of the pandemic, and the digital and commercial environment all shape the prevalence of distress (see Chapter 10). Both World Psychiatry reviews conclude that the social determinants represent “the most modifiable set of targets currently available” for preventing mental ill health[829]. Prevention therefore requires action across the wider conditions in which people live, learn and work.

The economic case. Economic modelling identifies a number of specific interventions that generate returns exceeding their costs over time, including workplace mental health promotion, parenting programmes for children with conduct problems, interventions for anxiety disorders, early intervention in psychosis, school- based social and emotional learning, debt advice and health-visitor-delivered support for postnatal depression[832]. These are genuine returns, but they relate to well- defined interventions delivered to clearly specified populations. They should not be taken as evidence that prevention in general necessarily pays for itself. The more defensible conclusion is that prevention is a portfolio in which some interventions are demonstrably cost-effective, some remain unproven, and some are ineffective. The economic case therefore needs to be made intervention by intervention rather than in the aggregate.

15.9.2 What prevention addresses

Chapter 10 described the social and developmental drivers of mental ill health in terms of four recurring mechanisms: uncertainty, evaluation, belonging and recovery. The classic public health distinction between universal approaches, selective approaches targeted at groups with elevated risk, and indicated approaches for people already showing early subthreshold difficulties cuts across all four[830]. The recurring finding is that selective and indicated interventions consistently outperform universal ones, while the universal interventions with the greatest population impact are structural actions that address the underlying determinants themselves rather than universal psychological programmes.

15.9.2.1 Reducing uncertainty: economic and material security

The strongest causal evidence across the prevention field concerns material security, corresponding closely to the mechanism of uncertainty described in Chapter 10. A systematic review of 136 studies found that increases in individual or household income improved mental health and wellbeing, while decreases worsened them, with the largest effects seen when people were lifted out of poverty[829, 833]. A series of natural experiments conducted in the United States, Canada and Kenya similarly found that policies aimed at poverty reduction, income security and material support were associated with improved mental health outcomes[834, 835]. Under Mexico’s Oportunidades conditional cash transfer programme, children whose families enrolled before the age of two showed fewer behavioural problems at ages eight to ten than those enrolled later ([833], as reviewed in [829]). A further randomised study of unconditional cash transfers improved wellbeing and reduced depressive symptoms[836], while increases in the minimum wage have been associated with reductions in suicide among lower-educated workers, with no corresponding effect among college-educated workers – a specificity that strengthens the causal interpretation[837].

Two findings are particularly important for policy. First, the design of the benefits system is itself a mental health intervention, for better or worse. The rollout of Universal Credit in England was associated with a 6.6 percentage-point increase in psychological distress among those affected, equivalent to tens of thousands of additional people crossing the threshold for clinically significant distress[838]. Second, structural interventions often improve the intermediate outcomes they directly target – such as housing stability or income – more reliably than they reduce psychiatric symptoms. For example, the Housing First approach substantially improves housing stability but has much smaller effects on psychiatric symptoms[839]. Their benefits therefore appear first in functioning, security and participation, with broader effects on mental health emerging more gradually. The social determinants review

concludes that poverty is “inextricably linked to most social determinants of mental health, and could be considered a root cause”, and that no public health approach to mental health that fails to address it is likely to achieve its aims[829].

15.9.2.2 Reducing harmful evaluation: bullying, comparison and the online environment

Chapter 10 identified exposure to harmful evaluation and unfavourable comparison as a second mechanism through which environments influence mental health. Two well-evidenced school-based programmes act through this pathway and illustrate the value of behavioural rather than awareness-based interventions. The KiVa anti- bullying programme, developed and evaluated in Finland, was replicated in a UK cluster-randomised trial that found a reduction in victimisation of around 13% at a cost of less than £2 per pupil per year[840], with systematic reviews reporting reductions in both victimisation and perpetration of around 15–19%[841]. The evidence is less clear, however, on whether these programmes improve mental health directly, at least in the short term. The Good Behaviour Game reliably reduces conduct problems, and children who participated at around six years of age were less likely in early adulthood to report suicidal thoughts or suicide attempts[842]. Both programmes are relatively inexpensive, replicable and capable of being delivered at scale.

School-based cognitive behavioural programmes produce small reductions in depression and anxiety, with stronger effects when targeted at children and young people at increased risk rather than delivered universally[829]. One important exception is the universal SEYLE suicide prevention programme, which approximately halved severe suicidal thoughts and suicide attempts at 12 months[843]. Evidence concerning social media remains more contested. Average associations with adolescent wellbeing are generally small, but there is a precautionary case for regulating specific harms, including algorithms that promote self-harm, suicide and eating disorder content to vulnerable young people[844-846].

Mental health awareness and literacy programmes reliably improve knowledge and reduce stigma, but have weaker and less consistent effects on help-seeking and little demonstrated effect on symptoms[847]. Chapter 11 examines this evidence in greater detail.

15.9.2.3 Strengthening belonging: community, connection and place

Community and place-based interventions are often evaluated against broader outcomes such as wellbeing, educational attainment, social participation and inequality rather than diagnosed mental illness. Differences in the evidence base should therefore not be interpreted as evidence of ineffectiveness.

The closure of youth clubs in London was followed by measurable increases in youth offending and reductions in educational attainment, concentrated among disadvantaged young people[848]. Long-term evaluation of Sure Start similarly demonstrated the prevention of thousands of hospital admissions each year, with the greatest benefits occurring in the most deprived communities[486]. Both provide strong evidence for the value of community infrastructure, although in both cases the mental health benefits are inferred rather than measured directly.

Evidence regarding organised activities, volunteering and civic participation is more mixed[849], suggesting that the nature of the activity is important. A large epidemiological analysis identified five characteristics consistently associated with benefit: social connectedness, opportunities to develop skills, increased self- confidence, enjoyment and environments characterised by safety and trust[850]. There is moderate evidence supporting arts engagement and improvements in adolescent mental health, particularly among girls[851], alongside strong evidence for the preventive benefits of sport, exercise and physical activity[852], with consistent positive effects on depressive symptoms and anxiety[853], including through organised sport[854].

Evidence relating to place is stronger still. Access to green space during childhood is robustly associated with a lower risk of psychiatric disorder in very large cohort studies[855], while a cluster randomised trial in Philadelphia found that greening vacant urban land reduced depressive symptoms and improved self-worth among residents in the poorest neighbourhoods[856]. This is one of the few experimental neighbourhood-level studies and illustrates the potential value of changing environments rather than individuals.

By contrast, the evidence supporting social prescribing remains relatively weak. Systematic reviews conclude that many UK evaluations are at high risk of bias, with little robust evidence of effectiveness, while people from minoritised ethnic communities remain under-represented[829, 857]. Interventions addressing loneliness also show promise, particularly cognitively focused approaches, but rest on a limited trial base with few disorder-level outcomes[858]. The Review therefore supports investment in community infrastructure and social connection on the strength of their demonstrated effects on offending, educational attainment and hospitalisation, while recognising that direct evidence that they prevent diagnosable mental illness remains limited.

15.9.2.4 Protecting recovery: activity, rest and restoration

The fourth mechanism identified in Chapter 10 is recovery – the opportunity for rest, restoration and psychological replenishment. The clearest evidence concerns physical activity. A large prospective cohort study found that physically inactive

people had substantially higher odds of subsequently developing depression, although no comparable association was observed for anxiety[859]. Meta-analysis involving more than a quarter of a million participants similarly demonstrates a protective association[860]. Universal and selective exercise interventions may reduce the incidence of anxiety disorders and lessen the severity of depressive symptoms among young people at increased risk, although some apparent benefits may reflect selective reporting in unblinded trials[830]. The effect therefore appears genuine but modest. Physical activity remains one of the few individual-level protective factors supported by prospective evidence.

The wider recovery mechanisms identified in Chapter 10, including adequate sleep, protected downtime and restorative contact with nature, are mechanistically coherent and consistent with the green-space findings described above, although direct evidence from prevention trials remains comparatively limited.

15.9.2.5 Early years and the family: foundational, cross-cutting and enduring The evidence on early-years family support is encouraging in many areas and equally instructive in its failures. Parenting programmes represent the most developed strand. A systematic review of 24 trials of short-term group-based parenting programmes for children under four found benefits for child mental health, behaviour and parent-child interaction[829]. The two most widely used programmes, Triple P and Incredible Years, reduce disruptive behaviour, and one review estimated that around ten children would need to receive a parenting programme to prevent one case of adolescent anxiety disorder[829]. Two important qualifications consistently emerge. Effects are greater when programmes are targeted towards families at higher risk, and larger for conduct problems than for emotional difficulties[861, 862]. They also become smaller when evaluated independently of programme developers using blinded outcome measures[863].

There is also increasingly strong evidence supporting early intervention for families of autistic children. For example, the Preschool Autism Communication Trial (PACT) demonstrated sustained reductions in autism symptoms at a mean follow-up age of 10.5 years following a parent-mediated social communication intervention delivered when children were aged two to four years[864]. Economic evaluations have also reported substantial savings, including an estimated £46,000 reduction in societal costs over six years following PACT[784], and around a 3:1 return on investment over eight years following a video-feedback parenting intervention[785].

Evidence for early-life home visiting follows a similar pattern. A systematic review of 34 randomised and quasi-experimental studies found that home-visiting programmes improved the home environment, particularly parenting behaviour[829]. The strongest evidence comes from nurse home-visiting programmes. In David Olds’ US trials,

children of first-time mothers who received nurse visits during pregnancy and infancy experienced fewer behavioural problems during childhood and, by age fifteen, lower levels of alcohol use and offending, with programme costs recovered by approximately the child’s fourth birthday[829]. Perinatal mental health support also carries some of the strongest evidence in the prevention field. Structured psychological interventions for women at risk of perinatal depression reduce new episodes of depression, with counselling-based approaches receiving a US Preventive Services Task Force Grade B recommendation[865].

The prevention literature also contains an important cautionary example. The Nurse- Family Partnership, which produced substantial benefits for high-risk families in the United States, was introduced into England as the Family Nurse Partnership and evaluated through the large randomised Building Blocks trial. Unlike the US findings, the English evaluation found no effect on any of its four primary outcomes and no evidence of cost-effectiveness[866]. Prevention programmes therefore need to be judged individually rather than by category, and tested in pragmatic trials that reflect the settings in which they will actually be delivered.

As discussed in Chapter 14, proactive, targeted and flexible support for families must continue beyond the early years as children encounter new developmental challenges and different mental health difficulties emerge across development[867]. Randomised controlled trials demonstrate the effectiveness of parent-mediated interventions across a wide range of needs, including social communication[864], sleep[868], feeding[869], anxiety[870] and disruptive behaviour[871]. Benefits have also been demonstrated among socially disadvantaged families[872], where parents experience mental health difficulties[873], and across different developmental stages, including adolescence[874]. Although economic evaluations remain comparatively limited, those available consistently indicate cost-effectiveness[784, 875], including for therapist-supported digital parent-mediated interventions[e.g. 876].

Accessible and effective support for parents’ own mental health is also an important preventive intervention. Poor parental mental health can directly affect parents’ ability to care for their children as they would wish[877]. National primary care data indicate that around one in four children aged 0–16 years are exposed to maternal mental illness, with the highest prevalence in the most deprived communities[425]. Meta-analytic evidence also indicates that psychological treatment for maternal depression improves not only maternal mental health, but also mother-child interactions, parenting and marital functioning, together with children’s subsequent mental health outcomes[878].

15.9.2.6 The workplace

Work is where much adult distress arises and where much of its economic cost is borne, yet the evidence on workplace mental health interventions delivers an uncomfortable conclusion. A study involving tens of thousands of workers across more than two hundred organisations found that individual wellbeing initiatives – including mindfulness sessions, resilience training and wellbeing apps – produced no measurable benefit compared with non-participation. Only volunteering was associated with improved wellbeing[879].

The emerging consensus is therefore to change the workplace rather than the worker. Organisational and job-design interventions appear more promising than interventions aimed solely at individuals, although their pooled effects remain modest and highly dependent on implementation. The clearest exception is Individual Placement and Support (IPS) for people with severe mental illness, which approximately doubles rates of competitive employment across dozens of trials and demonstrates similar benefits across both high- and low-income settings[829, 880]. This is an intervention that changes people’s opportunities and circumstances rather than simply encouraging them to cope better (see Chapter 16).

15.9.3 Reviews and assessments of impact, area by area

The evidence on prevention points consistently in the same direction. Preventive interventions are real, but their effects are generally small to moderate; they tend to attenuate when programmes are scaled up and independently evaluated; and targeted (selective and indicated) approaches consistently outperform universal approaches at the individual level. The central finding is that the strongest and most consistently causal evidence is for structural and upstream interventions acting on the material and commercial environment, which have the greatest potential for population-level impact. The preventive psychiatry review reaches a parallel conclusion from the literature on children and young people: selective and indicated psychological or psychoeducational programmes may improve symptoms, but their effectiveness in reducing the incidence of disorder remains largely unproven, whereas “universal public health approaches targeting social determinants have the greatest potential for reducing the risk profile of the population as a whole”[830]. This is not an either/or choice, however. Selective interventions delivered at different stages of development can also be effective[881].

15.9.4 Prevention that works at scale

The interventions with the strongest evidence share a common feature: they change the environment rather than the individual. The clearest example comes from suicide prevention, where means restriction remains the best-supported intervention.

Legislation limiting the pack size of over-the-counter paracetamol in England and Wales was followed by a reduction of around 43% in deaths from paracetamol poisoning over the following decade, together with a comparable reduction in liver transplants resulting from overdose[882]. Physical barriers at suicide hotspots reduce deaths at those sites by around 90%[883], while the removal of carbon monoxide from the domestic gas supply produced one of the largest falls in the national suicide rate ever recorded[884]. None of these interventions acts directly on an individual’s psychology; each changes the environment in which a moment of crisis either does or does not become fatal.

The same structural logic underpins other successful population-level interventions. Tobacco control reduced adult smoking from around 46% in the mid-1970s to under 13% by 2022[885]. The preventive psychiatry review draws a direct lesson for mental health, arguing that “experience from smoking prevention suggests that similar public-health, population-level interventions are far more effective than individual- level approaches”[830]. Minimum unit pricing for alcohol in Scotland reduced alcohol- attributable deaths by around 13%, with the largest reductions occurring in the most deprived communities[886]. The Icelandic Prevention Model transformed adolescent substance use across an entire population over two decades by changing family and community environments rather than attempting to persuade young people individually[887].

15.9.5 Integrating the case for prevention

The prevention evidence therefore points in the same direction as the Review’s central argument. The interventions with the strongest evidence act upstream, do not require diagnosis to reach people, and show their earliest benefits in functioning, security, behaviour and participation. Individual-level indicated prevention remains valuable but cannot by itself change population incidence. Population-level action on the social determinants offers the greater opportunity[830]. One important qualification applies throughout. Population-level prevention should mean structural action on the determinants of mental ill health rather than universal psychological programmes delivered indiscriminately across whole populations. It is the latter that carry the clearest evidence of potential harm.

15.9.6 The risks of prevention

Prevention is not inherently benign. Three forms of risk deserve particular attention.

First, awareness-raising is not psychologically neutral. Chapter 11 shows that greater awareness has brought substantial benefits while also shaping how experiences are recognised and interpreted, so that the same explanatory framework that helps one person may encourage another to interpret ordinary

distress through a diagnostic lens[888]. Awareness and literacy programmes should therefore be treated as active interventions that are evaluated for their effects rather than assumed to be beneficial because their intentions are positive. This point is reinforced by the Aware trial, in which a universal school-based awareness intervention was followed by higher levels of emotional difficulty at long-term follow- up[620].

Second, some universal programmes may actively cause harm. The largest UK trial of universal school-based mindfulness, Myriad, found no benefit on its primary outcomes and a signal of harm among pupils already at risk[702, 703]. A trial of universal school-based dialectical behaviour therapy skills similarly reported increases in anxiety and depression together with poorer emotional regulation among participants[889]. Beyond schools, psychological debriefing following trauma was once widely practised until trials demonstrated that it increased, rather than reduced, subsequent post-traumatic stress[890], while deterrence programmes such as Scared Straight increase rather than reduce offending[891]. Universal, one-size- fits-all psychological interventions delivered irrespective of need therefore carry genuine iatrogenic risk. Preventive programmes should be held to the same evidential standards, and the same vigilance regarding harm, as therapeutic interventions.

Third, it is important not to overcorrect into a blanket fear of assessment or labelling. A randomised trial found that screening adolescents for suicide risk did not increase distress or suicidal ideation[892]. The distinction that matters is between well-targeted assessment and support offered to those who need it – which appears safe – and universal psychological interventions delivered irrespective of need, where the evidence of harm is concentrated. This aligns directly with the central conclusion running throughout this section: targeted and indicated prevention consistently outperforms universal prevention, in terms of both effectiveness and safety.

Finally, prevention is often promised but not delivered. The preventive psychiatry review is explicit about the tendency within parts of the policy literature to overstate the evidence, and about a research field in which very few risk-prediction tools have been validated and almost none implemented[830]. At the same time, opportunities for prevention are frequently missed because population-level interventions are held to evidential standards that can be difficult to achieve for complex, cross-sector and long-term programmes. The history of prevention policy is therefore one of repeated ambition undermined by the diversion of resources towards acute pressures and by the absence of clear accountability for upstream action. Both risks need to be addressed in the design of future policy and are reflected in the recommendations that follow.

15.10 Chapter synthesis

This chapter has examined prevention and what happens once a person enters the system.

The central challenge identified throughout is an implementation gap: the gap between knowing what works and delivering it consistently, equitably and at scale.

A system that only responds once distress has crystallised will remain permanently under strain because the queue reforms behind every person taken out of it. The prevention evidence consistently indicates that the interventions with the strongest causal support act structurally: on income and the design of the benefits system, on the commercial determinants of health, on the means of suicide, on the school and early-years environment, and on supporting families rather than through psychological programmes delivered to whole populations. They do not require diagnosis to reach people, and their earliest benefits are seen in functioning, security and participation.

The Review also concludes that some universal psychological programmes delivered to whole populations are not merely ineffective but carry a genuine risk of harm. Where prevention aims to reduce the incidence of enduring mental health conditions, interventions should generally be targeted towards people at increased risk (selective or indicated prevention) rather than delivered universally.

The evidence points consistently towards a model of care that extends from upstream prevention through to specialist treatment. Support should be organised around need rather than diagnosis alone; delivered early and proportionately through stratified pathways; supported by a well-organised workforce; commissioned to deliver clearly specified evidence-based interventions; and held accountable for the outcomes it achieves. Those with the greatest complexity require sustained, specialist and relational support, while the system as a whole should demonstrate improvement through outcomes that are routinely measured, transparently reported and actively used to guide care.

None of these developments can be assumed to occur automatically. Prevention is particularly vulnerable. Across successive governments, the recurring problem has not been disagreement about its importance, but the persistent diversion of resources towards acute and emergency care together with the absence of clear accountability for upstream action. The challenge for future reform is therefore not simply to redesign services, but to create the conditions, including clear national and local accountability for prevention, in which evidence-based delivery becomes routine practice.

The success of the system should ultimately be judged not by how many people reach assessment or receive a diagnosis, but by whether people and their families receive timely, evidence-based support that improves functioning, enhances safety and enables fuller participation in everyday life and, upstream of all this, by how much distress is prevented from arising in the first place.

Chapter 16. Work, adulthood and participation

The participation gap is widening, reflected in rising absence from education and employment. Current systems are better at certifying absence than enabling participation.

Participation and health influence one another. Good education, employment, relationships and community connection support wellbeing and improve long- term outcomes, while exclusion can worsen health and reduce future opportunities. Once disengagement becomes established, returning to education or employment becomes progressively more difficult.

Participation should be treated as both an early warning indicator and a core outcome. Timely, needs-led support – based on functioning, context and the factors driving difficulty – should aim to sustain participation and prevent future exclusion.

The preceding chapters examined recognition, diagnosis, access and treatment. This chapter asks what all of that is for. For most people, the end purpose is not diagnosis or treatment itself, but the opportunity to live a full life: to learn, to work, to sustain relationships and to participate in the ordinary activities of family and community life. Participation is therefore a core measure of whether systems are improving people’s lives.

Rising numbers of young people, particularly those aged 16 to 24, are outside education, employment and training, while among working-age adults economic inactivity associated with ill health has increased substantially. Mental health and neurodevelopmental conditions account for a growing proportion of both. These developments provide the context for the evidence considered throughout this chapter.

Adult participation trajectories are often established much earlier. For that reason, the discussion begins with adolescence and the transition into adulthood before considering participation across the wider working-age population.

Participation is understood broadly throughout this chapter, encompassing education, employment, training, relationships, community life and wider social

inclusion, while recognising that participation will look different for different people. The strongest available evidence relates to participation in education, employment and economic inactivity. As reviewers and lived-experience partners emphasised, an employment-only perspective would both overlook the wider benefits of participation and exclude those for whom employment is not the relevant outcome. The chapter therefore also considers wider forms of participation that contribute to health and wellbeing (Section 16.9).

This chapter focuses on systems rather than individual motivation. Most people want to work, study and contribute, but participation is shaped by educational opportunity, recruitment, workplace design, social circumstances and access to support. ‘The system’ therefore includes health services, education, employers, welfare, and voluntary and community organisations.

Consistent with the wider Review, diagnosis is not a measure of work capacity. Participation depends on functioning, context and the demands people face, and support should respond accordingly. Across health, education, employment and welfare, current systems remain better at managing absence than enabling participation.

16.1 The scale and shape of the participation gap

The participation gap has widened across the working-age population, particularly among younger adults. No single measure captures the whole picture: self-reported symptoms, sickness absence, benefit claims and formal diagnoses each describe overlapping but distinct populations. Many people experience psychological distress that substantially affects their capacity to study, work or participate without ever meeting formal diagnostic criteria[893, 894].

Between 2015 and 2024, the number of working-age adults reporting a long-term mental health condition that limited their ability to work increased by approximately 936,000. More than half of that increase (around 533,000 people, or 57%) occurred among those aged 16 to 34 years[11]. Mental ill health has also become the largest single contributor to work-related sickness. During 2024/25, work-related stress, depression and anxiety affected an estimated 964,000 workers and accounted for 22.1 million working days lost, representing 52% of all work-related ill health and 62% of all working days lost through work-related ill health[895].

In 2024/25, around 34% of working-age autistic adults were in employment, compared with approximately 55% of disabled people overall and around 82% of people who were not disabled[896]. Around 77% of unemployed autistic adults report

wanting to work[896]. Autistic people experience one of the widest disability pay gaps and are around twice as likely as non-disabled graduates to remain unemployed fifteen months after graduation, with only around 36% entering full-time employment during that period[311].

A School Census analysis in 2024 found that around 9% of identified autistic pupils had a recorded co-occurring intellectual disability, and around 5% had a moderate or severe intellectual disability[11]. Although intellectual disability is under-recorded in English data, and administrative records cannot identify autism with severe or profound intellectual disability, these figures suggest that those likely to require the most intensive lifelong support represent a minority of the identified autistic population, even though that group cannot yet be estimated precisely.

The Review’s findings in Chapter 9 suggest that the poorest outcomes are often experienced by people living with both neurodevelopmental and mental health conditions. Across the life course, these combinations are associated with lower educational attainment, more disrupted transitions into adulthood and poorer employment outcomes.

Participation is a systems outcome rather than a direct consequence of diagnosis. It can change, and early warning signs such as falling attendance or prolonged absence create opportunities for intervention before exclusion becomes entrenched.

16.2 A life-course view: NEET young people and the transition to adulthood

The participation gap often opens before adulthood. We regard absence from education, employment or training as both an early warning sign of unmet need and a predictor of later outcomes. Once disengagement becomes established, the likelihood of remaining outside education or employment increases.

In the three months to December 2025, 957,000 people aged 16–24 were NEET (not in education, employment or training), representing 12.8% of that age group and around 200,000 more than in 2021[16]. In 2025, 44% reported a work-limiting health condition, compared with 26% in 2015 – an increase of around 70% over the decade. Among those reporting a work-limiting condition, around 68% identified a mental health condition or autism as their primary condition[16]. Over a similar period, the proportion of 16–24-year-olds screening positive for a common mental health condition increased from around 17% in 2014 to approximately 23% in 2023/24 (see also Chapter 6)[16].

Impetus estimates that each additional young person who becomes NEET costs the state around £244,000 over their lifetime, including lost tax and National Insurance (£102,000), additional welfare payments (£68,000), NHS costs (£65,000) and criminal justice costs (£9,000)[897, 898]. Applied to the current NEET population, this implies an aggregate lifetime cost of well over £200 billion. These are modelled estimates rather than observed expenditure, but they illustrate the potential value of preventing disengagement.

Identified autism among English school pupils more than doubled between 2016 and 2024, from 1.5% to 3.3%, growing faster than any other category of special educational need[11]. Identification now peaks in the early secondary years, with roughly 45,000 more autistic pupils expected in English secondary schools by 2027/28 even if incidence remains unchanged. This indicates future demand for transition, employment and adult support[11].

Participation should therefore be understood in two ways: as an outcome that matters in its own right and as an early warning signal. Falling attendance, withdrawal from school or college, repeated job loss or prolonged sickness absence should prompt enquiry into what is driving the change, rather than being interpreted as evidence of mental disorder or treated solely as an educational or employment problem.

Figure 16.1 Number of NEET young adults aged 16–24 reporting a mental health condition or autism, 2015–2025

Source: Labour Force Survey analysis, Health Foundation[16]. Long-term health conditions refer to the main condition reported by respondents. Shaded bands show 95% confidence intervals.

Figure 16.1 shows that the number of NEET young people reporting a mental health condition or autism has increased as the NEET population itself has grown, while

their share of the NEET population has remained broadly stable at around one third. The rise is therefore better understood as an increase in the absolute number of young people experiencing significant difficulties rather than as a growing proportion of NEET status being attributable to health. More importantly, the likelihood of being NEET among young people reporting a mental health condition has widened relative to their peers: recent analysis suggests they are now around three times more likely to be NEET than those without a mental health condition[16].

Understanding Society shows a similar pattern. The proportion of NEET young people describing themselves as long-term sick or disabled almost doubled over little more than a decade, while the prevalence of health conditions or disability among young people generally rose from approximately 14% in 2017 to around 20% by 2023. Over the same period, the relative risk of being NEET among those with a health condition increased from around 1.9 to 2.3[899].

The rising number of young people who are NEET cannot be explained simply by increasing individual disadvantage. During the early 2010s, NEET rates declined despite increasing socioeconomic disadvantage, before rising during and after the pandemic. The relationship between psychological distress and later disengagement remained broadly stable over time. What changed was the number of young people experiencing significant difficulties in the first place[900]. This suggests that rising disengagement is associated primarily with an increase in the prevalence of psychological distress rather than a change in the extent to which distress predicts later participation outcomes.

Longitudinal evidence shows how quickly early disengagement can become persistent. In the Millennium Cohort Study, around 13% of more than 8,300 young people were NEET at age 23; those who were NEET at 17 were more than four times as likely to remain NEET at 23. A long-term mental health condition at 17 was associated with a 2.4-fold higher likelihood of being NEET at 23; recorded autism with a 3.6-fold increase and ADHD with a 3.2-fold increase[17].

These findings should not collapse very different groups into a single employment target. A smaller group of young people have lifelong disabilities or very high support needs for whom paid employment may never be a realistic outcome, particularly those with severe or profound intellectual disability, with or without autism. For these individuals, the question is not how to secure employment but how to ensure high- quality support throughout life. Unfortunately, the cohort data cannot reliably distinguish those who could participate in employment with better support from those who will require lifelong provision of other kinds. This remains an important evidence gap (see Chapter 8).

Co-occurrence increases risk further. In the Millennium Cohort Study, the likelihood of later NEET status increased from around 1.6-fold with one mental health difficulty to more than four-fold with five, and almost six-fold with six or more[901]. Figure 16.2 illustrates this accumulation of risk. It reinforces the importance of responding to the whole pattern of difficulties rather than waiting for a single diagnosis or crisis.

Higher education shows the same increase in visible need. The proportion of UK students disclosing a mental health condition rose from below 1% in 2010/11 to around 5.8% in 2023/24. Disclosure is not equivalent to prevalence and reflects changing awareness and reporting as well as underlying need; the educational implications are considered in Chapter 13[902].

Figure 16.2 Relative risk of being NEET at age 23 according to the number of co-occurring mental health difficulties at age 17

Source: Millennium Cohort Study. Relative risks greater than 1 indicate an increased likelihood of being NEET at age 23. Error bars show 95% confidence intervals. Reference groups: binary indicators are compared with absence of the condition or behaviour; BMI category = not overweight; number of mental health conditions = no mental health conditions; obesity and a mental health condition = not reporting both.

Health is not the only influence on participation. Labour-market opportunities matter too: by spring 2026, overall UK hiring was around 14% below its level a year earlier, with entry-level recruitment also falling[903]. Young people’s expectations of work are also shaped by families, schools, peers, social media and other digital environments (see Chapter 10). Support therefore needs to address both health and the practical routes into education, training and employment.

“I have two daughters, aged 21 and 23, who are both NEET. Neither followed the university route and neither received support with applying for jobs or understanding where to start. My youngest daughter has been unemployed for two years. My eldest, who is autistic, has ADHD and severe mental illness, has been unemployed for eighteen months since leaving college, when her EHCP (Education, Health and Care Plan) ended.

When they left college, no one followed up to see how job-seeking was progressing. It felt as though they became invisible because they were not going on to university. Apprenticeships attract hundreds of applicants and even shop or seasonal jobs often require previous experience. Both of my daughters struggle with customer-facing roles because of anxiety and their neurodevelopmental difficulties.

Every unanswered application and every CV that receives no response affects their mental health. Sometimes it feels as though they just need someone to give them a chance.”

Lived Experience Partner Another concern is young people’s preparation for the transition into employment itself. Stakeholders highlighted problems navigating workplaces and employment- related expectations. Some suggested that opportunities to build confidence, independence and familiarity with workplace environments may have been reduced following disruption by the pandemic. Others suggested that young people’s understanding, expectations and interpretations of work and their own difficulties are increasingly shaped not only by families, schools, employers and clinicians, but also by social media and peer networks, the digital environments discussed in Chapter

  • These expectations about employment shape how emotional distress and

neurodevelopmental differences are interpreted and affect beliefs both positively and negatively about what support is needed and what futures are possible.

It is clear that young people benefit most when support reaches them early and when health and education respond together rather than separately. Preventing later exclusion from education, training and employment depends less upon intervening

after disengagement has become established and more upon preventing that disengagement from becoming entrenched in the first place.

Unfortunately, the evidence base for re-engaging NEET young people is much weaker than deserved by the scale of the problem. Reviews find small and often methodologically weak studies, with promising signals for psychological support, nature-based activity, peer approaches, social prescribing and holistic coaching[904,

905]. For autistic young people, one meta-analysis found peer-mediated approaches and authentic real-world settings more effective than didactic alternatives[906]. Participation itself should be a primary outcome in future evaluations.

16.3 A bidirectional relationship: mental health, neurodevelopment and economic inactivity

Across working-age adults, mental health, neurodevelopmental conditions and economic inactivity are closely connected. Common mental health conditions, autism and learning difficulties such as dyslexia together account for around one quarter of the recent increase in long-term sickness among people aged 16–49, with common mental health conditions contributing the largest share[907].

The relationship runs in both directions. Mental health difficulties and neurodevelopmental conditions can reduce a person’s ability to participate in work or education, while prolonged absence from work or education can worsen mental and physical health, undermine confidence, weaken occupational identity and make subsequent return increasingly difficult[908, 909]. Early support may interrupt this cycle, whereas delays in recognition and support can allow difficulties to become progressively more entrenched.

Workplace context also matters. Mental ill health is more common in some sectors, including healthcare, hospitality, self-employment and insecure work[910-913]. Remaining in work depends partly on whether tasks, hours, communication, sensory environments and expectations can be adapted to changing capacities[914, 915]. Participation is therefore shaped by the interaction between health, needs and capabilities and the environments in which people work.

The nature of work is also changing. Many jobs place greater demands on cognitive, social and emotional skills and digital communication, while non-standard and insecure work may provide less access to protections, occupational health and adjustments[913]. Remote and hybrid work can increase access for some people while reducing workplace connection for others. These shifts are unlikely to explain

rising inactivity by themselves, but they may change which work environments are manageable for people with fluctuating mental health or neurodevelopmental needs.

16.4 Staying in work: recovery, timing and the return-to-work curve

The likelihood of returning successfully to work declines as absence lengthens.

The Keep Britain Working review reports that employees absent from work for four to six weeks have a 96% probability of returning, falling below 50% once absence extends beyond one year[22]. Earlier clinical studies show the same pattern, with return-to-work probability declining after relatively short periods of absence for emotional, stress-related and mood disorders[916, 917]. The precise estimates vary, but the direction is consistent: the longer disconnection lasts, the harder it becomes to reverse.

Low participation should not be interpreted as lack of ability, motivation or aspiration. Work is not invariably beneficial, and an early return is not always appropriate. Secure, well-designed and supported work can promote recovery, while insecure or unsupported work can worsen mental health[918, 919].

This emphasis on making participation possible is deliberate. For common mental health conditions, interventions combining clinical treatment with active employment support – for example, NHS Talking Therapies alongside employment advisers – can reduce sickness absence and improve return-to-work outcomes more effectively than treatment alone[920-922]. The strongest evidence in this field concerns Individual Placement and Support, which places people directly into competitive employment with ongoing support rather than requiring a period of preparation first. Across dozens of randomised trials it roughly doubles competitive employment rates, and the effect holds in both high- and low-income settings[829, 880]; adapted models for younger adults have achieved competitive employment rates of 58.3% against 32.4%, with additional benefits for education enrolment[923]. IPS works by changing a person’s circumstances rather than their outlook, but it is commissioned mainly for people with severe mental illness, which is why the Review recommends adapting it for young people and for others whose participation is at risk. The broader principle is that health and employment are better supported together than sequentially.

Current systems do not consistently provide this early, integrated support. Occupational health referrals are often delayed, uptake of the former Fit for Work service was low, and graduated ‘may be fit for work’ advice remains uncommon for people experiencing common mental disorders[924].

Access to occupational health and workplace health support remains strongly associated with employer size. Around 86% of large employers provide some occupational health support, compared with about 30% of small and medium-sized enterprises[22]. Occupational health is often weakest in hospitality, retail, gig and platform work and self-employment – sectors where mental ill health is common. Workers in more deprived areas are less likely to receive return-to-work advice and more likely to experience prolonged absence[925, 926].

The practical shift is from managing absence once it has become established towards early workplace engagement, adjustments and supported return. The Keep Britain Working review argues for shared responsibility between employers, employees and the NHS, with employers playing a central role in prevention through workplace adjustments, treatment flexibility and phased return to work[22].

Those at greatest risk of becoming detached from employment are often those least likely to receive timely support. Work-and-health support therefore needs to be available early enough to act within weeks rather than months of absence, particularly for small employers, insecure workers and people in sectors with little occupational-health infrastructure.

16.5 The workplace: needs-led support, adjustments and inclusive design

16.5.1 Supporting participation in work: adjustments, managers and functional need

The support people require to participate successfully at work depends on their functioning, the demands of the workplace and the context in which they work. Diagnosis can help explain difficulties and guide treatment, but it does not determine work capacity or specify which adjustment will help a particular person in a particular role.

Workplace adjustments can be effective when they are provided. Introducing two or more workplace accommodations has been associated with elimination of the approximately ten percentage-point employment gap between workers with and without mental impairments, while flexible scheduling and opportunities for remote working appear to confer greater benefit than reduced-hours arrangements alone[927]. Yet workers with mental health conditions and related impairments remain less likely to receive workplace accommodations than those with physical or pain- related conditions[928].

Government funds some of this directly. Access to Work provides grants for specialist equipment, support workers, travel and workplace changes for disabled people who are in or about to start work. It does not pay for the reasonable adjustments an employer is legally required to make; it sits alongside them, a distinction that employers and employees frequently misunderstand. Mental health and learning conditions now account for around half of the people the scheme supports – 37,900 in 2024–25, against 11,200 six years earlier – but autism and ADHD are not recorded separately, so the scheme cannot say how many of its users have needs related to neurodevelopmental conditions[929].

Its difficulty is delay. Outstanding applications rose from 21,700 in March 2022 to 62,100 in March 2025, and average processing time from 28 working days to 66 against a target of 25; applicants have been told they could wait up to 37 weeks[929]. The Buckland Review found that 61% of disabled applicants waited more than three months and one in five more than six, and that only around one in five disabled adults has ever received Access to Work support at all[311]. The Public Accounts Committee has noted the particular difficulty this creates for applicants with neurodevelopmental conditions, for whom the process demands exactly the administrative capacity they are applying for support to obtain.

Many effective adjustments concern how work is organised rather than expensive equipment: clearer expectations and written communication, predictable supervision, protected rest periods where feasible, flexibility in hours or location, workload management and attention to sensory environments.

As for any employee, the relationship with a line manager is often central. People are more likely to disclose needs to an approachable and empathic manager, who in turn is better placed to ensure that agreed adjustments work in practice. In one healthcare programme, structured active-listening training increased managers’ confidence in recognising and discussing psychological distress from under half to more than 80% one month later[930]. Even the best managers, however, will sometimes require access to occupational-health or specialist advice.

Much of the evidence comes from larger, office-based organisations with defined line-management structures. In sectors where mental ill health is common – hospitality, retail, construction, social care, gig and platform work, and small businesses – the ‘line manager’ may be a shift supervisor with limited time and resources. Relationships and psychological safety matter, but implementation needs to reflect the realities of different workplaces.

“Workplace support is difficult to obtain in many of the fields in which neurodivergent people work: the gig economy, self-employment, zero- hours contracts, small enterprises without occupational health services,

and industries where there are no clearly outlined examples of reasonable adjustments.”

Lived Experience Partner These principles therefore need to work outside office-based settings, where opportunities for flexibility may be more constrained[931-936]. This may be particularly important for mental health conditions, where effective support often depends on flexibility, supervision, communication and workload management rather than physical adaptations or specialist equipment.

These findings relate primarily to people already in work, or able to enter work with appropriate support. They do not imply that workplace adjustments alone can address every barrier to participation, which may also be shaped by the severity or fluctuation of a person’s condition, family circumstances, co-occurring difficulties and access to wider health and social support.

As ADHD UK’s submission put it, adjustments need to be person-led: “There is no such thing as a standard ADHD adjustment, because there is no standard person with ADHD and no standard job”

Henry Shelford, ADHD UK, 2026 The four adjustments most commonly recommended for ADHD illustrate this challenge. They are flexible working, noise-cancelling headphones, written instructions and a quiet workspace. Respectively, these may be impossible for many of the roughly 1.5 million people working in adult social care, unsafe for many of the 1.8 million working in hospitality, impractical for the more than one million people who drive for a living, and effectively unavailable for the 2.7 million people employed in retail[931, 937-941].

Effective support therefore cannot be reduced to a standard list of adjustments. What is reasonable and effective depends on the interaction between the individual, the role, the workplace and the wider team.

Functional assessment provides a more useful basis for identifying appropriate workplace support than diagnosis alone because people with the same diagnosis may have very different difficulties, while the same adjustment may help people with different diagnoses. Adjustments should also be reasonable, proportionate and clearly related to the demands of the role. Employer confidence matters: support is more likely to be offered when expectations are clear and practical, particularly for smaller organisations with limited HR or occupational health capacity.

This needs-led approach is consistent with the Equality Act 2010, under which the duty to make reasonable adjustments depends on the substantial impact of an impairment rather than on diagnosis alone. In practice, however, lived-experience partners noted that a diagnosis can still be the evidence that persuades an employer to act. The important point is that support should not be delayed until a diagnosis has been secured. Support offered while people wait should be a bridge to timely assessment rather than a substitute for it. Adjustments help people to function; assessment helps clarify needs and guide support and treatment; specific treatment can be transformative; and all three belong together[942].

A needs-led approach must also avoid narrowing access to assessment. ADHD UK and the autism charities cautioned that functional triage could become a new barrier in the absence of robust measures of severity. The Review records this concern because support based on need should complement, not ration, access to appropriate diagnosis and treatment.

Evidence on which adjustments work best for which people and in which roles remains limited, particularly for neurodevelopmental conditions. This is a research gap rather than a reason to withhold reasonable adjustments. Employees themselves are often the best source of information about what helps, and future evaluation should build a stronger evidence base around sustained participation and job retention.

16.5.2 Inclusive workplaces by design: from adjustment to participation

Inclusive, or universal, design seeks to organise recruitment, jobs, communication, working practices and physical environments for the widest possible range of needs from the outset. Clear communication, structured work, predictable routines, accessible recruitment processes and attention to sensory environments can benefit many employees without requiring disclosure or diagnosis. Individual adjustments will still be needed, but inclusive design can make them less exceptional and less stigmatising.

Only around one third of autistic employees report being fully open about being autistic, around one third feel unable to discuss adjustment needs at all, and more than one quarter of those requesting adjustments report having had them refused[311]. One qualification should be kept in mind: much of this evidence comes from autistic people already in work, who may already have navigated the barriers that keep others out of employment. Adjustments therefore will not be sufficient for everyone.

Around one third of employers report concerns that autistic people may not fit easily into existing teams, while approximately 60% worry about getting support wrong or do not know where to obtain appropriate advice[943]. Better recruitment design, manager capability and access to specialist advice can remove barriers before they become reasons for exclusion.

The autistic population moving through education and into employment is also changing. Identification is increasing, the sex balance is becoming more even, and a larger proportion of identified autistic people do not have intellectual disability[11]. This reinforces both the heterogeneity at the heart of the needs-led argument and the conclusion that employment outcomes reflect structural barriers rather than limited ability.

Participation should also be judged by the quality of work. Employment that is insecure, poorly paid, unstable or damaging to health cannot automatically be regarded as a successful outcome. Retention, progression, financial security and wellbeing matter alongside whether someone is technically in work.

Recent employment reviews converge on the same direction. The Buckland Review promotes ‘inclusion by design’; the Department for Work and Pensions Expert Academic Panel on Neurodiversity emphasises functional need rather than diagnostic category; and Keep Britain Working places health-related economic inactivity within a framework of shared responsibility between employers, employees and health services[22, 311]. The Review supports this direction: participation should be regarded as a legitimate health outcome, and health, employment and skills support should work together around the person.

16.6 The health–work interface: primary care and fit notes

General practice and fit notes form an important part of the health-work interface. They can either help people remain connected to work or, unintentionally, make full absence the default when workplace information and occupational health support are limited.

General practice sits at the intersection of health, employment and welfare. GPs provide clinical care and referral, issue fit notes, and often help people navigate wider support. This makes primary care highly influential in determining whether illness leads to temporary adjustment or longer-term disconnection, but it also requires clinicians to make decisions that depend on information lying outside the consultation itself.

GPs are frequently asked to make judgements about work participation without detailed knowledge of an individual’s workplace, the adjustments available or the practical realities of their role. Particularly for common mental health conditions, the consultation can therefore become as much an administrative gateway as a clinical encounter, with decisions about participation being made in the absence of the information needed to support them confidently.

During 2024/25, approximately 11.2 million fit notes were issued across England. Around 76% (8.54 million) contained no recorded diagnosis, while of the 2.69 million that did, around 36% (0.96 million) related to mental and behavioural disorders, including the common mental health conditions – such as anxiety, depression and stress-related disorders – that account for much of the current burden of work- related ill health[944].

The fit note was designed to support continued participation through options such as amended duties, altered hours and phased return. Yet ‘may be fit for work’ appeared on only 6.8% of fit notes issued during 2024/25[944]. Fit notes issued for mental and behavioural disorders were also substantially more likely than those issued for other conditions to recommend absences exceeding two weeks (84%, compared with 75%)[944].

Figure 16.3 Trends in fit notes issued in England by diagnosis status, April 2021–September 2025.

Source: NHS England / NHS Digital fit note statistics.

Figure 16.4 Trends in fit notes issued without a diagnosis and with ‘may be fit for work’ advice, England, April 2021–September 2025.

Source: NHS England / NHS Digital.

Figure 16.5 Percentage distribution of fit notes issued in England: mental health and behavioural disorders compared with all diagnoses, April 2021– September 2025.

Source: NHS England / NHS Digital.

Figures 16.3 to 16.5 reinforce this pattern, showing the scale of fit-note activity, the predominance of certificates issued without a recorded diagnosis, the persistently low use of ‘may be fit for work’ advice and the substantial contribution of mental and behavioural disorders. The data suggest a system that is better at certifying absence than enabling participation. This is not a criticism of individual GPs: without workplace information or a practical route to negotiate adjustments, ‘not fit for work’ may be the safest available option.

For mental health and neurodevelopmental conditions, where functioning can fluctuate, binary certification can reinforce an all-or-nothing model of work. Graduated hours, modified duties or temporary changes in communication or workload may sometimes preserve connection with employment and make subsequent return easier.

The present system frequently asks primary care clinicians to make decisions about work participation without access to occupational-health expertise, detailed workplace information or practical mechanisms for engaging employers. The problem therefore lies less with individual clinical decisions than with how responsibility is distributed across the system.

For this reason, the Review suggests that the role of primary care should evolve from one centred predominantly on certification towards one that more actively connects people to support. Within such a model, participation becomes a legitimate health outcome alongside symptom improvement. The fit note becomes less an authorisation of absence and more a prompt to consider what adjustments or support might enable continued engagement with work where this is safe and appropriate.

As lived-experience partners noted, a ten-minute consultation is unlikely to provide sufficient time to broker workplace adjustments. Faster access to occupational- health advice, dedicated work-and-health services, better information-sharing and clearer referral pathways would instead allow primary care to connect people to support rather than carry sole responsibility for providing it.

16.7 Disability, benefits and access to support

The benefits system plays an essential role for people whose health limits participation, protecting many from poverty and further deterioration. Detailed welfare-policy design lies outside the remit of this Review. Its contribution is to clarify the clinical evidence on which fair and effective policy should rest, treating the interaction between health and welfare as important clinical context rather than as a subject of welfare reform in its own right. The focus here is therefore on the relationship between health, diagnosis and access to support, and on how systems for evidencing need shape how participation is understood.

The relationship between mental health and the benefits system is substantial and continues to grow. Mental health conditions accounted for around 21% of long-term sickness and disability benefit claims in 1995, rising to approximately 47% by 2014. Using a different measure, by 2020/21 they were the primary disabling condition for around 44% of disability benefit recipients[945, 946]. Expenditure on benefits associated

with mental health conditions was estimated at £25.9 billion in 2023/24[947, 948]. Much of the growth in health-related benefit caseloads relates to common mental disorders – particularly anxiety and depression – rather than severe mental illness.

Two considerations follow.

i) Diagnosis remains important evidence of need but should not automatically be interpreted as evidence of fixed or permanent incapacity. As the Review has noted, diagnosis can open doors, but it can also close them. With the exception of people with severe or profound intellectual disability, with or without autism, describing someone as having fixed and permanent incapacity may have serious unintended consequences. A correct diagnosis can be transformative, but used poorly it can also become limiting, defining people by what they cannot do and reducing several complex difficulties to a single label.

ii) Equally, ‘need’ must not become an unregulated replacement gateway. Functional assessment requires clear standards and safeguards against inconsistency and inequity. For many mental health and neurodevelopmental conditions, functioning changes over time. Financial protection should therefore sit alongside timely treatment, adjustment and employment support where participation is realistic, so that systems protect people without creating avoidable barriers to future re-engagement.

16.8 Identity, expectations and participation

Participation is shaped not only by systems and services, but also by the ways in which people understand their own difficulties. Diagnosis can support participation by validating difficulties, improving self-understanding and opening access to treatment, adjustments and communities of shared experience[96, 127] (see Chapter 5). This can be particularly important for people diagnosed later in life, for whom diagnosis may bring relief and a more coherent understanding of earlier experiences. As several lived-experience partners observed, recognising that certain difficulties are not personal failings can itself help people pursue the settings in which they are most likely to thrive.

The Review’s concern is to avoid an interpretation – reinforced at times by systems, professional practice and public narratives – in which diagnosis becomes associated, intentionally or otherwise, with an assumption that future participation, employment or independence will necessarily be limited. This is not an argument against

diagnosis, nor a denial that some conditions and associated difficulties are enduring. It is an argument about the meaning attached to diagnosis. The role of health and wider public systems should be to reinforce its enabling rather than disabling potential: using diagnosis to understand need, guide treatment and support, and open routes to participation.

16.9 Participation beyond employment

Participation extends beyond education, training and paid employment. Relationships, community involvement, volunteering, physical activity and meaningful roles all contribute to a full life and, for people with autism with severe or profound learning disability, or the most severe and enduring mental illness, these may be the central participation outcomes. An employment-only measure would exclude people with the greatest support needs.

The health benefits of wider participation are substantial. The WHO estimates that loneliness and social isolation are associated with around 871,000 deaths each year worldwide[949]. Volunteering has been associated with a 22% lower risk of mortality, lower depression and higher life satisfaction and wellbeing, although the evidence is observational and causal mechanisms remain uncertain[950].

Everyday activities also matter. An overview of nearly one hundred systematic reviews found that physical activity reduced symptoms of depression and anxiety, with effects around one and a half times those reported for counselling or medication in the studies reviewed[951]. In a small randomised trial of adults with major depression, dietary improvement was associated with remission in around 32%, compared with around 8% among those receiving social support[952]. Spending at least two hours a week in green space is associated with better health and wellbeing across age, illness and disability groups[953]. Enabling social connection, activity and participation therefore contributes to health, but is not a substitute for effective treatment (see Chapter 10 on environmental drivers and Chapter 15 on structural approaches to prevention).

For people with lifelong high support needs, social inclusion, valued relationships, meaningful activity and quality of life should themselves be core outcomes. Social inclusion remains one of the largest unmet needs for people with severe mental illness[954]. Unpaid carers also need to be included: caring can restrict their own employment, relationships and participation, particularly where high support needs within a family require a parent or other carer to leave work[955]. Other groups – including young people who are NEET because of long-term sickness or disability, people in insecure or gig work, and those affected by homelessness, substance use

or the criminal justice system – may need routes into participation that begin with stability, connection and practical support rather than job search alone.

These outcomes require continuity across systems. Support provided through education, including an Education, Health and Care Plan, can end as young people move into adult life, just when coordination between education, employment, adult social care, housing and community support becomes most important. Social prescribing provides one existing route for connecting people with volunteering, physical activity, arts, nature and community support[955, 956]. Its evidence base remains in development (see Chapter 15), but it illustrates how participation beyond employment can become a practical aim of services rather than an aspiration.

16.10 Chapter synthesis

Participation should be treated as a core outcome of health, education and employment systems alongside clinical outcomes. Falling participation should also be recognised as an early warning sign: declining attendance, withdrawal from education or prolonged work absence should prompt timely assessment of what is driving the difficulty and proportionate support before disengagement becomes entrenched.

Support should be organised around need, functioning and participation rather than diagnosis alone. This does not mean lowering diagnostic standards or restricting access to diagnosis. It means combining clinical assessment and treatment where required with educational and workplace adjustments, employment support, occupational health, practical and financial help, and opportunities for social and community participation. For people with lifelong high support needs, participation should be understood more broadly through relationships, meaningful activity, community life and quality of life.

Systems should shift from managing absence to enabling participation: intervening earlier, strengthening links between health, education, employment and welfare, making inclusive practice and reasonable adjustments easier to access, and measuring whether people remain connected or are able to re-engage. This direction is reflected in initiatives such as Get Greater Manchester Working, which brings health, employment, skills and wider social support together around participation rather than service boundaries.

Participation provides a practical test of whether support is improving people’s lives – not simply whether a diagnosis has been made or a service provided.

Chapter 17. Special populations and settings the Review has considered

Some of the populations with the greatest mental health need – including children in care, children in custody, young carers and unaccompanied asylum-seeking children – have limited or outdated prevalence data, greater difficulty accessing services and poorer continuity of care.

Administrative data often reflect the behaviour of systems rather than underlying need. Rising self-reported eating disorders, increasing rates of self-harm and violence within the secure estate despite falling numbers of children in custody, and persistent gaps between recorded and estimated prevalence all illustrate that recorded demand is not the same as population need.

System design and service thresholds shape outcomes. People with multiple or overlapping needs are often excluded by services organised around single conditions or sectors. This chapter highlights the importance of integrated, cross- sector support, better measurement of need, and service models that reflect the complexity of the populations they are intended to serve.

Three considerations governed what is included in this chapter: the concentration of need; distinct service architecture; and the settings in which the Review’s distinction between population prevalence and administrative prevalence is particularly important. Where the evidence does not exist, we have said so, because the absence of data for some of the populations with the greatest need is itself one of this chapter’s central findings. Emergency and acute settings are considered only briefly; the Review’s substantive discussion of crisis presentation, paediatric admission and inpatient care is presented in Chapter 19.

17.1 Children in contact with children’s social care

On 31 March 2025, around 81,770 children in England were looked after by local authorities, around 400,000 were classified as children in need, and approximately 49,000 were subject to a child protection plan[957]. Almost all had experienced abuse,

neglect or serious adversity. Exposure to parental mental ill health, parental substance use, domestic abuse and extra-familial harm is common. By any measure, these are among the children in England most likely to develop mental health difficulties, and they are also among those about whom the system holds the most comprehensive administrative information.

It knows remarkably little, however, about their mental health. The most authoritative national estimate remains a survey conducted in 2002, which found that around 45% of looked-after children aged 5–17 met criteria for a mental disorder, rising to around 72% among those living in residential care, compared with approximately 10% in the general child population[20]. Reanalysis found conduct disorder in 26.7%, compared with 1.9% in non-disadvantaged private households, together with emotional disorders (11.1%), hyperkinetic disorder (8.7%) and autistic spectrum disorder (2.6%)[195]. No comparable national survey has been undertaken since.

A systematic review of subsequent UK studies identified 39 reports estimating the prevalence of mental health and neurodevelopmental conditions among care- experienced young people, with estimates ranging from 1% to 82%[958]. That range does not represent genuine variation in prevalence; it reflects the absence of a common method of measurement.

More recent cohort evidence has begun to fill this gap, although it cannot close it. A study recruiting through thirteen English and Welsh local authorities and adoption agencies found that 54.3% of care-experienced children aged 10–13 years, and 71.0% of those aged 16–17 years, scored above threshold for post-traumatic stress. Among the older group, anxiety affected 21.9%, depression 27.1%, and the proportion reporting low wellbeing increased from 14% to 42.6% across the two age groups[959]. This is a large multi-authority sample rather than a national probability survey, and it cannot substitute for one. Nevertheless, it provides the best current indication that the situation has not improved substantially since 2002. It also suggests something the earlier survey could not: that much of the burden within this population is post-traumatic, and that approaches focusing primarily on emotional and behavioural symptoms may fail to detect it adequately.

Routine information is collected using a screening instrument. Local authorities complete the SDQ as part of the annual review for looked-after children. The most recent returns record 42% of eligible 5–16-year-olds scoring in the range giving cause for concern and a further 13% in the borderline range, compared with approximately 10% and 7%, respectively, among children of similar age in the general population[957]. These figures point in the same direction as the earlier survey but are not directly comparable. The SDQ is a brief behavioural screening measure completed by carers rather than a diagnostic assessment; coverage is incomplete,

and it is administered only to children continuously looked after for at least twelve months.

The adequacy of this approach has now been tested directly, and the results are concerning. Compared with standardised measures in a care-experienced sample, the carer-completed SDQ failed to identify between 44% and 48% of children with clinically elevated post-traumatic stress, anxiety or depressive symptoms[960]. The much larger population of children in need or those subject to child protection plans is not covered at all. We therefore have a population that is comprehensively counted administratively, but whose psychiatric morbidity was last properly measured more than two decades ago, before the changes in recognition and diagnosis described elsewhere in this Review.

The position is even weaker for neurodevelopmental conditions. There is no recent national estimate of ADHD or autism prevalence among children in care. The nearest available indicators come from education. Around 61% of children looked after continuously for one year have an identified special educational need, compared with 19% of all pupils, with social, emotional and mental health recorded as the most common primary need. Even disability recording is inconsistent. One study found local authorities recording disability for 13% of care leavers, whereas 27% reported a disability or long-term condition themselves[961]. Children in care also move frequently, and each move may effectively reset their place on an assessment waiting list. It is difficult to assess conditions that require longitudinal observation when the people observing the child repeatedly change.

Access to treatment does not compensate for these shortcomings. Analysis of more than 71,000 records from one large mental health provider found children subject to a child protection plan to be around twice as likely to have a referral declined as children without social work involvement, while children with other forms of social work involvement were more than three times as likely[962]. The same study found that children currently looked after were less likely to have referrals rejected, which the authors attributed to the trust’s dedicated service for children in care. The Review regards this as particularly instructive because it suggests that disadvantage reflects service organisation rather than characteristics of the children themselves. Although this is a single-site study and should not be interpreted as a national estimate, its findings are consistent with evidence heard repeatedly by the Review: difficulties understood to arise primarily from adversity or placement instability are often judged not to constitute treatable mental health conditions, while services may be reluctant to begin work with children whose placements are unlikely to remain stable.

Analysis of social care records found that 81% of children with elevated mental health difficulties were referred to a service within a year. Of those referred, 13% had

their referral declined, while 30% of those who began support experienced premature discharge. Most strikingly, access to services was not associated with better mental health one or two years later[963]. A systematic review of 64 interventions for care-experienced children and young people similarly found that although short-term benefits can be detected, evidence of sustained improvement is lacking[964]. The Review therefore concludes that better outcomes depend not only on whether support is reached, but on whether it is the right support, delivered for long enough and with sufficient continuity to be effective. That is a more difficult problem than reducing waiting lists, and it cannot be solved by referral targets alone.

At the extreme, a child presenting with very high levels of risk may be assessed as not meeting criteria for detention under the Mental Health Act, or as someone for whom admission would not be in their best interests because their behaviour reflects trauma rather than a treatable mental disorder. The same child may also be declined by children’s homes because of the risks associated with self-harm or suicide. In such circumstances, application to the High Court for a placement of last resort may become the only available mechanism.

Applications for orders authorising the deprivation of liberty of a child rose to 1,439 in 2025, a 12.5% increase on the previous year and up from around 100 annually in 2017/18[965, 966]. Placements in accommodation not registered to provide care increased from 147 in 2020/21 to 982 in 2023/24[967]. Evidence examined by the Review from a September 2025 snapshot of children in unregistered placements found that over one third (36%) were receiving support from child and adolescent mental health services and a further 7% were waiting for support. Among those also subject to a deprivation of liberty order, the proportion receiving support rose to 55%, compared with 27% among those without such an order, although local authorities did not provide this information for 16% of children[968]. These are, in other words, the children with the greatest mental health needs, living in settings not registered to provide care, under court orders made because no part of the ordinary system was able to accept them. The Review also notes the Government’s announcement in April 2026 of a review into the deaths of care leavers following 91 notifications of death during 2024/25, most involving young people aged between 16 and 21[969].

Whether better assessment would alter these outcomes is now beginning to be tested empirically rather than assumed. A randomised controlled trial is underway across eight to ten local authorities, allocating young people in care to usual mental health consultation with or without a standardised comprehensive assessment, and following them for twelve months to determine whether this changes decisions made by social care and mental health services, and ultimately improves outcomes for the young people themselves[970].

17.2 Eating disorders

Eating disorders provide a particularly clear example of the interaction between changing definitions, rising demand and service organisation. Twenty years ago they were encountered largely within specialist services. They are now recognised as a major public health concern, and services across England report increasing difficulty meeting demand.

Population evidence points towards increasing prevalence. In the Adult Psychiatric Morbidity Survey, the proportion of adults screening positive on the Scoff questionnaire increased from 6.4% in 2007 to 9.1% in 2023/24, and among women from 9.2% to 11.9%. In 2023/24, 2.2% of adults screened positive and also reported that concerns about food interfered with work, personal responsibilities or social life, while screening prevalence reached 20.1% among 16- to 24-year-olds. Applying a clinical assessment alongside the screening measure produced an estimated past- year prevalence of 1.3%, although this estimate rests on only 26 identified cases and is therefore underpowered for subgroup analyses. Findings from surveys of children and young people appear more dramatic but require considerably greater caution. Clinician-rated assessments suggested prevalence among 11- to 16-year-olds increased from 0.5% in 2017 to 2.6% in 2023, and among 17- to 19-year-olds from 0.8% to 12.5%. The Review does not interpret these figures as evidence of a fifteen- fold increase in eating disorders over six years. The survey authors themselves advise against direct comparison. Three important aspects of the survey changed between waves: data collection moved from face-to-face to online, the 2023 data were collected through a second-stage study rather than within the main survey, and only the eating disorder module of the DAWBA was administered in 2023 rather than the complete instrument. The report explicitly notes that mode effects are likely to be greatest ‘for sensitive questions such as eating problems and disorders’[535]. The younger estimates rely largely on parent report, whereas the older estimates rely primarily on self-report, and the two approaches identify different aspects of eating disorders. Restrictive eating tends to be recognised more reliably by parents and carers, whereas binge eating and body image concerns are more readily identified by young people themselves[971]. Something real is clearly being detected, and the direction of change is supported by independent evidence. The magnitude of that change, however, cannot yet be estimated confidently.

Administrative data present a steadier picture. New eating disorder presentations recorded in primary care increased from 32.3 per 100,000 population in 2000 to 37.2 per 100,000 in 2009[972], with a further sharp rise, concentrated among girls and young women, after March 2020[973]. Recorded incidence is, however, consistently higher and rising faster in the least deprived areas than in the most deprived, and people from minoritised ethnic groups have lower rates of primary care contact and

lower odds of referral to specialist services than White patients[974]. The Review reads this as a pattern of recognition rather than a pattern of illness. Eating disorders have historically been characterised, incorrectly, as conditions affecting affluent White girls, and the administrative data still follow that assumption rather than correct it.

Growth has been concentrated not in anorexia nervosa or bulimia nervosa, but in the residual category of other eating disorders, including atypical presentations, binge eating disorder and avoidant/restrictive food intake disorder (ARFID). This partly reflects changing definitions. Successive revisions of diagnostic criteria removed the amenorrhoea criterion, relaxed frequency requirements for bingeing and purging, introduced binge eating disorder as a separate diagnosis, and added ARFID. In clinical series these changes increased the proportion of children and young people meeting criteria for anorexia nervosa from 29.3% to 40%, while reducing the proportion classified as unspecified eating disorders from 62.3% to 32.6%[975]. Under ICD-11, anorexia nervosa may now be diagnosed at any weight. Growth is therefore occurring precisely at the diagnostic boundary of the category.

Admissions of children and young people to acute medical wards for mental health concerns show the largest increase of any diagnostic category reported for eating disorders, rising from 478 in 2012/13 to 2,938 in 2021/22 – an increase of 515% – alongside a 113% rise in admissions for mental health concerns among girls aged 11 to 15[760]. Linked primary care and hospital data show that all-cause mortality is more than doubled among people with an eating disorder diagnosis, and more than tripled among those with anorexia nervosa. Mortality and admission rate ratios are higher in males, a population that remains substantially under-recognised.

Eating disorders offer the Review something no other condition does: a direct comparison between two different service models applied to the same illness. Children and young people have had an access and waiting-time standard since 2016 – treatment beginning within one week for urgent referrals and within four weeks for routine referrals – supported by dedicated community services and, more recently, around £54 million a year of additional investment[976]. Adults have no equivalent access standard and have received no comparable investment. The same condition, in the same country, is therefore being managed under two different policy approaches.

National service mapping in 2025 found a median wait from referral to treatment of four days for children and young people, compared with 42 days for adults. Around 62% of children’s community eating disorder teams accepted self-referrals, compared with 15% of adult services. Adult community eating disorder services had 3.8 whole-time equivalent staff per 100,000 population, compared with 10.5 per

100,000 in children’s services[977]. The average interval between onset and first specialist treatment for adults is estimated at 53 months for bulimia nervosa and 67 months for binge eating disorder[978], by which time severity and complexity are typically greater and treatment less effective.

An access standard, however, is not the same as sufficient capacity. The children’s standard has never been achieved nationally. In the most recently published quarter, 78.4% of urgent referrals and 81.7% of routine referrals began treatment within the target times, against a national standard of 95%, and performance has not recovered to pre-pandemic levels[979]. A change in data collection introduced in April 2023 also means that figures collected before and after that date should not be interpreted as a continuous time series. The future direction is similarly uncertain. An analysis published in February 2026 identified sixteen integrated care boards planning combined real-terms reductions of £2.1 million to children’s and young people’s eating disorder services[980].

Further lessons come from the rapid early intervention model for 16- to 25-year-olds with illness of less than three years’ duration. This approach reduced time to treatment and the need for inpatient care and was associated with estimated savings of £10.9 million across 2,430 patients during national rollout[981]. Although the model was extended to all 54 eligible trusts, only around half of teams responding to the 2025 national audit reported an active pathway. Implementation is not the same as adoption, and adoption is not the same as sustained delivery.

Eating disorders are also increasingly characterised by co-occurrence. Screening positive in the Adult Psychiatric Morbidity Survey was strongly associated with borderline personality disorder, common mental disorders, self-harm, ADHD, bipolar disorder and post-traumatic stress disorder. The association with ADHD strengthened between 2007 and 2023/24, whereas those with anxiety and depression remained broadly unchanged. The evidence reviewed also suggests that around one quarter of people with anorexia nervosa or bulimia nervosa may be autistic, although this estimate remains uncertain. ARFID, for which the only UK incidence estimate is 2.79 per 100,000 children and young people[982], commonly co- occurs with both. Diagnostic error can therefore occur in both directions: restrictive eating may be attributed to autism without adequate assessment of its nutritional consequences, while eating disorders may be treated without recognising the neurodevelopmental profile that shapes presentation and treatment needs.

The growth in eating disorders is occurring at the boundaries of the diagnostic category and among age groups not fully covered by current service design. Both observations reflect changes in the way services recognise and respond to need, as well as changes in the conditions themselves.

17.3 Children and young people in the secure estate

The number of children held in custody in England and Wales has fallen sharply. There were 2,881 children in custody in 2008/09. First-time entrants to the youth justice system fell by 68% over the following decade. The average custodial population in the year to March 2025 was 418, the lowest on record, and by May 2026 had fallen below 400[983, 984]. This represents one of the more substantial achievements of public policy during the period covered by this Review.

What has happened to the children who remain is a different matter and illustrates the difference between a count and a rate. As the population has fallen, its composition has become increasingly concentrated. The proportion held on remand rather than under sentence increased from 23% to 44% over ten years, while 62% of children remanded did not subsequently receive a custodial sentence. Children from minoritised ethnic backgrounds now account for around 52% of the custodial population, with the proportion of children of Mixed ethnicity increasing from around 10% to 18%. Around 65% have at some point been in local authority care, and approximately 37% are looked after at the point of entry (see Chapter 15)[811, 985]. This links the present section directly to the previous one: the secure estate is, to a substantial degree, downstream of the care system. Girls account for only around 2% of the population but experience a disproportionate share of self-harm incidents.

Mental health and neurodevelopmental needs are high by every measure, although the estimates require careful interpretation. The most recent meta-analysis of adolescents in detention reported conduct disorder in around 60%, ADHD in 17.3% of boys and 17.5% of girls, major depression in 10.1% and 25.8%, post-traumatic stress disorder in 8.6% and 18.2%, and psychotic illness in 2.7% and 2.9%, respectively[399]. The Review presents these figures separately for boys and girls because the differences are substantial and would be obscured by combined estimates.

For neurodevelopmental conditions, the evidence is more variable and much of it is older. Speech, language and communication needs have been reported in 60–90% of children in custody, compared with general population estimates of 1–10%, depending on the definition used. Screening studies suggest learning disability in 23–32%, compared with 2–4% in the general population, while a history of traumatic brain injury involving loss of consciousness has been reported in 37–46%, rising to 65–76% when any self-reported head injury is included[312, 986]. Estimates of autism range from around 4%, based on diagnosed cases across English secure establishments, to 15% in a Swedish institutional sample[987]. The Review does not regard the higher estimate as an appropriate basis for planning services in England.

Only around 5% of children with speech, language and communication needs have those needs recognised before entering the youth justice system[988]. Around 80% of children cautioned or sentenced belong to the special educational needs population, rising to 86% among those convicted of serious violent offences[21]. These children were, in the overwhelming majority of cases, already known to schools, health services and often social care. Their needs were identifiable. Where they were recognised, they were too often interpreted primarily as behaviour.

Every child entering the secure estate receives a structured health assessment using the Comprehensive Health Assessment Tool (CHAT), covering physical health, mental health, substance use and neurodisability[406]. National returns identify ADHD in around 35% of children, consistent with the research literature, but autism in only 0.1%[251]. Against research estimates of 4–15%, this should not be interpreted as a prevalence finding but as evidence of a measurement problem.

The Review heard several plausible explanations. In some establishments CHAT may be used as though it were a diagnostic assessment rather than a screening instrument. Assessments undertaken within the first ten days of arrival may encourage children to minimise difficulties before trusting relationships have developed. Autism may also be interpreted primarily as behavioural disturbance within this population. The instrument itself is unlikely to be the principal problem.

Safety indicators point in the same direction. Violence and restraint have increased despite the continuing fall in the custodial population. During the year to March 2026 there were 1,926 assaults (423.9 per 100 children), an increase of 11%, and 4,329 uses of force (952.8 per 100 children), an increase of 5%. Self-harm incidents numbered 1,474 (324.4 per 100 children), only marginally lower than the previous year[989]. 43% of children surveyed reported feeling unsafe at some point, 61% had experienced bullying, violence or victimisation, and inspectors judged that 55% of previously identified safety concerns had shown insufficient or no meaningful progress[811]. Inspectors also reported in 2024 that, in one establishment, 90% of the population was subject to keep-apart restrictions and some children spent more than 22 hours a day locked in their cells, conditions they judged met the widely accepted definition of solitary confinement.

The principal service response has been the Framework for Integrated Care (Secure Stairs), a trauma-informed, formulation-led model described more fully in Chapter 15. Its independent evaluation[820] illustrates a well-implemented reform whose benefits are real but more limited than some claims made for it. Around 85% of children had a multidisciplinary formulation drawing on an average of five professional disciplines, staff-reported understanding of children’s histories increased from 52% to 67%, and 84% of children reported feeling listened to.

Implementation, however, was faster and more complete in secure children’s homes than in the larger young offender institutions where most children are held, and the evaluation itself concluded that evidence for reductions in violence remained preliminary[820]. Formulation changes how staff understand a child. Whether it changes what happens to the child depends on the setting.

The setting itself is under pressure. Places in secure children’s homes have fallen by around 27% since 2010, and in 2024/25 there were approximately 42 referrals for every available place[967]. The Review infers that many of the children who cannot be placed in a secure children’s home are those who subsequently enter unregistered accommodation under a deprivation of liberty order. The Youth Justice White Paper, published in May 2026[983], commits to a ten-year transition from large, centralised institutions towards smaller, regionally distributed therapeutic settings. The Review welcomes this direction while noting that the White Paper contains comparatively little detail on mental health and neurodevelopmental provision, and that its therapeutic ambitions depend on a secure children’s home sector that has been contracting for the past fifteen years.

Custody is the end of a pathway, not a population. Almost everything described in this section was visible much earlier in children’s lives, within systems that either did not recognise need or were unable to respond effectively.

17.4 Emergency and acute settings

The Review’s detailed account of crisis care and emergency pathways is presented in Chapter 14. Three findings are highlighted here because they bear directly on the populations described above.

Emergency departments and acute paediatric wards have increasingly become the route into care for children experiencing mental health crises. Since 2019 there have been more than half a million emergency attendances for mental health needs, while admissions to acute medical wards for mental health concerns increased by 65% between 2012/13 and 2021/22, compared with 10.1% for admissions overall. More than half of these admissions followed self-harm[760]. Children may then spend days or even weeks in emergency departments or paediatric wards because no appropriate placement is available.

Around 30% of admissions of under-18s to specialist mental health inpatient wards in 2025/26 involved young people with a learning disability, autism, or both. Increasing numbers involve autistic young people without learning disability, and the great majority of these admissions occur under the Mental Health Act[795].

17.5 Other groups the Review has considered

Five further groups warrant brief consideration, some extending beyond children and young people. In each, the Review is satisfied that mental health need is substantially elevated. In each, however, the available evidence remains too limited to support more detailed conclusions.

On 31 March 2025, around 6,540 unaccompanied asylum-seeking children were looked after by English local authorities, a 12% decrease on the previous year but still substantially above the level of a decade earlier. Around 94% were male and 90% were aged 16 years or over. No UK meta-analysis has isolated unaccompanied status as an independent risk factor. European systematic review evidence reports prevalence estimates among refugee and asylum-seeking children ranging from 19% to 53% for post-traumatic stress disorder and 10% to 33% for depression, several times higher than in the general population[990]. The studies are heterogeneous, however, and mostly do not distinguish unaccompanied from accompanied children.

The distinctive challenges are often structural rather than clinical. Between 2021 and 2024, more than 5,400 children were placed in hotel accommodation outside the statutory child protection framework, a practice subsequently found unlawful by the High Court, and more than 50 remain missing. A disputed age assessment may determine not simply the service response, but whether a young person is treated as a child at all.

The 2021 Census identified around 120,000 young carers aged 5–17 in England. A young carer indicator was introduced into the School Census in January 2023 and identified 38,983 pupils in its first year, increasing to approximately 64,500 by January 2025. Survey estimates suggest the true number may exceed 800,000 among secondary school pupils alone[991]. The discrepancy may reflect the fact that caring is disclosed where young people expect support and concealed where they fear intervention. Evidence linking caring to poorer mental health is consistent in direction but generally low to moderate in quality and largely cross-sectional, making it difficult to disentangle caring itself from the effects of parental illness and poverty[992]. The educational consequences are clearer: 51% of young carers achieved grade 4 or above in English and mathematics in 2024/25 compared with 65% of their peers, while 35% were persistently absent compared with 18%.

Youth homelessness has also increased. Between April 2024 and March 2025, 123,934 people aged 16–24 across the UK experienced homelessness, and more than one third of those approaching English local authorities received no initial statutory assessment. Among homeless young people using specialist services, 54% had an identified mental health condition, rising to 68% among those with experience

of rough sleeping[993]. Care leavers are heavily over-represented, with homelessness among 18- to 20-year-old care leavers increasing by 37% over five years. Practical barriers remain substantial. Registration with a GP without a fixed address, although a legal entitlement, continues to be refused often enough to constitute a recognised obstacle, while transitions between children’s and adult mental health services coincide with the age at which homelessness is most common. Separate organisation of mental health and substance use services can also leave those with both needs excluded from each.

The Review examined post-traumatic stress disorder among armed forces veterans expecting to find a public narrative of epidemic running ahead of a relatively stable epidemiology. That is not what the evidence shows, and the correction is instructive.

The King’s Centre for Military Health Research has followed the same cohort of UK armed forces personnel since 2003[994]. Across successive waves of follow-up, probable PTSD increased from around 4% in 2004–06 to approximately 6% in 2014– 16, and to 9–10% in 2022–23, with the increase concentrated among those who had left service and those who had served in combat roles. The increase is therefore real, and it is concentrated in identifiable subgroups. Alcohol misuse was, for much of the same period, the larger problem within this population and remains an important concern.

At the same time, another widely quoted finding is also true, but applies to a different generation. Veterans of an earlier era, including those who completed National Service, experience mental health outcomes that are comparable with, or better than, those of the age-matched general population. Both statements are correct. The only thing that separates them is a cohort followed for twenty years.

That is the lesson, and it is not really about veterans. It is that sustained longitudinal measurement can resolve contested public narratives in both directions at once – confirming a genuine increase where one exists while also refuting an exaggerated version of the same claim. None of the populations described in Sections 17.1 to 17.5 has anything comparable.

The Review also notes that, although the dedicated veterans’ mental health service established in 2017 has substantially increased referrals, the evidence evaluating its effectiveness remains comparatively limited. Building a longitudinal evidence base and building a service are complementary, but they are not the same task.

We draw four observations from this chapter.

First, need is measured least well where it is most concentrated. There has been no national survey of the mental health of looked-after children for more than twenty years. There is no published national aggregation of the health screen administered to every child entering the secure estate. Young carers are identified administratively at only around one tenth of the numbers suggested by survey evidence. There is no UK evidence base specifically examining the mental health of unaccompanied children. This is the reverse of what a rational allocation of measurement effort would produce. The consequence extends beyond statistics: policy for these groups is shaped largely by advocacy and inspection rather than epidemiology, making it episodic and reactive. The veterans cohort illustrates the alternative. Twenty years of following the same people made it possible to establish that PTSD had genuinely increased within one part of that population. Measurement infrastructure is not an administrative overhead. It is what allows contested questions to be answered and provides a secure foundation for policy.

Second, these settings illustrate more clearly than anywhere else that administrative prevalence reflects system behaviour rather than population need. The youth custody population fell by around 85%, while rates of self-harm, assault and use of force increased. Referrals to children’s social care rose over a comparable period while the number of children formally recorded as being in need fell. Eating disorder diagnoses in adults changed relatively little while self-reported eating difficulties increased substantially. In each case, the measure that changed most was the one describing what the system did, not what was happening to people.

The dementia example discussed in Chapter 5 (5.8) carries this point even further. Here, administrative and population prevalence did not simply diverge – they moved in opposite directions. The reason is documented, together with the policy that accompanied it. A payment of £55 for each recorded dementia diagnosis was introduced in October 2014 to increase diagnosis rates, while the best epidemiological evidence suggested that age-specific prevalence of dementia itself was declining. The Review draws no conclusion about whether the scheme was right or wrong. Under-recognition was a genuine problem, and improving diagnosis was a legitimate policy objective. The point is narrower, but more generally applicable: administrative prevalence is not simply an observation of the world. It is shaped by the systems that generate it. Anyone interpreting rising diagnosis rates as direct evidence of rising illness risks confusing the instrument with what it measures.

Third, thresholds designed around single conditions exclude people whose needs span several conditions or systems. The clearest example in this Review is the child assessed as not detainable under the Mental Health Act because their distress is attributed primarily to adversity rather than a treatable disorder, while simultaneously being declined by children’s homes because the level of risk is judged too great. The only remaining route is an application to the High Court. This is not primarily a failure of individual clinical judgement. It reflects a system in which every threshold is organised around categories rather than around the child.

Fourth, service design changes outcomes. Eating disorders provide the clearest comparison identified by the Review: one condition, one country, an access and waiting-time standard with dedicated investment for children and young people, and neither for adults. The resulting differences in waiting times, workforce and routes into care are substantial. This is not a controlled experiment – adult and adolescent services differ in case mix, duration of illness and complexity – but those differences are unlikely to account for the scale of the observed gap. A standard that specifies when support should begin, rather than what must first be established before it begins, changes behaviour throughout the pathway. The qualification remains important: the children’s standard has never been achieved nationally, performance has not returned to pre-pandemic levels, and some commissioners are now planning service reductions.

Taken together, these populations are not an appendix to the Review’s argument but one of its strongest tests. They are the groups for whom diagnosis is least informative about need, functioning is most unstable, and support most obviously has to be organised across health, education, social care and justice simultaneously. A needs-led model that cannot accommodate a child in unregistered accommodation under a deprivation of liberty order, a 17-year-old on remand with unrecognised language needs, or a 22-year-old with binge eating disorder five years after onset is a model that serves only those already well served by the system.

The Review’s recommendations on earlier and more proportionate support, matching support to need rather than category, organising services around people, closing the treatment gap, protecting those with the greatest need, strengthening cross- government accountability, and improving surveillance and data collection are tested here first.

Section Five: A fairer and more effective response

Chapter 18. Understanding need: beyond diagnosis alone

This chapter argues for a shift from a system in which support depends largely on diagnosis to one in which diagnosis and assessment of need work together for people with mental health and neurodevelopmental conditions.

Support – including educational and workplace adjustments, family and peer support, psychoeducation, evidence-based interventions, and ongoing monitoring – should be based on assessed need wherever possible, rather than waiting for a formal diagnosis.

Diagnosis remains fundamental. It helps people understand their difficulties, guides treatment and, in some cases, determines access to specific interventions. But diagnosis alone does not describe the support an individual needs to function, participate and flourish.

Education and adult social care already assess support needs in relation to functioning and participation. The Review proposes developing a shared language for describing need across health, education, welfare and social care, building on the principles of the International Classification of Functioning (ICF) and aligning this work with current SEND reforms.

A needs-led approach has the potential to provide earlier, fairer and better coordinated support. However, it must be developed with strong safeguards so that it does not become simply another gateway to services or a new mechanism for rationing support.

The evidence reviewed throughout this report suggests that diagnosis is now being asked to do more than it can reasonably carry. It has become the principal gateway to specialist services, educational provision, workplace adjustments, welfare benefits and many other forms of practical support, performing both a clinical and an administrative function. Diagnosis remains essential: it helps explain a person’s difficulties and guides treatment. But it often provides only limited information about the support someone needs at a particular point in their life. People with the same diagnosis may differ greatly in their ability to participate in education, employment, relationships and everyday life. Co-occurrence is common, and people’s needs change as they develop and as their circumstances change.

Support therefore depends on answering a further question: what does this person need in order to participate more fully in everyday life? Across the evidence reviewed, a remarkably consistent set of factors emerges. These include current functioning, the risk of deterioration, the complexity of the presentation, the environments in which the person is living, learning and working, the interventions most likely to help, and the practical factors that determine whether support can be taken up. Together they provide a richer understanding of need than diagnosis or symptoms alone and a more proportionate basis for matching support to changing circumstances. Annex A summarises these six dimensions of need and proposes a way of recording them consistently across services.

The Review refers to this process as needs assessment. The difficulty is that different systems currently describe need in different ways. Health services emphasise diagnosis and symptoms; education focuses on barriers to learning and participation; employers on workplace functioning and reasonable adjustments; and social care on independence and daily living. Each approach is appropriate within its own setting, but together they create duplication, inconsistent thresholds and repeated assessments, requiring people and families to explain the same difficulties again and again. Education and adult social care already undertake detailed assessments of participation and support through Education, Health and Care Plans (EHCPs) and Care Act assessments, whether or not a diagnosis is present. Families told the Review that these assessments are often carried out thoughtfully by local authority staff, teachers and social workers, but that health services are too often absent from both their development and review.

The Review therefore proposes that health, education, employment and social care develop a shared language for describing need while preserving the essential role of diagnosis.

Timing is important. The Department for Education is revising arrangements for identifying and supporting special educational needs through its White Paper Every Child Achieving and Thriving[734] (Chapter 13), while the Department for Work and Pensions is reforming health-related assessment for welfare. If a shared language for describing need is to succeed, it cannot be developed by health services alone and then offered to education and welfare. It must be developed jointly, so that it aligns with the ways these systems determine and provide support. Otherwise, it risks recreating the same fragmentation that the Review is seeking to overcome.

18.1 Prioritising participation

Participation may be affected in many ways by mental health and neurodevelopmental conditions. A child may begin missing school, withdraw from

friendships or struggle with learning. A university student may continue attending lectures but become increasingly isolated or unable to manage independent study. An adult may find it progressively harder to sustain employment, maintain relationships or manage everyday responsibilities. In each case, the immediate question is not simply whether a diagnosis is present, but how current difficulties are affecting everyday life and what support is needed to prevent further deterioration.

Across ADHD (Chapter 7), autism (Chapter 8), education (Chapter 13) and health service planning (Chapters 14 and 15), the Review identifies the same broad determinants of support beyond diagnosis alone: functional impairment, persistence, complexity, risk and environmental context. Together these form the core components of a needs assessment.

Support needs may change even when diagnosis does not. Autism, ADHD and many long-term mental health conditions often persist across the lifespan, while participation, functioning, risk, distress and environmental demands change. Diagnosis provides continuity of clinical understanding; needs assessment describes the changing pattern of support required over time. Needs should therefore be reviewed at major transition points or when circumstances change significantly, rather than at fixed intervals alone, with a named professional responsible for continuity wherever possible. The transition from children’s to adult services remains the point at which this continuity most often breaks down. Chapter 14 shows how sharply service use falls despite continuing need.

18.2 Towards a shared language for need

If needs assessment is to work consistently across health, education, employment and social care, systems require a common language for describing functioning, participation and the environmental context in which they occur.

At present, no such language exists across public services. EHCPs describe barriers to learning and participation; universities use Study Needs Assessments and Disabled Students’ Allowance processes; employers undertake occupational health assessments; and adult social care uses Care Act assessments focused on wellbeing and independence. Information is difficult to transfer, assessments are duplicated, and people are repeatedly asked to describe the same difficulties as they move between services and life stages. Reviews of SEND, transitions into higher education and employment, and adult social care consistently identify this fragmentation as an important contributor to delay and discontinuity[995-999].

Earlier shared-assessment initiatives, including the Common Assessment Framework for children’s services, illustrate another risk: a shared language can become an administrative form rather than a meaningful conversation about need.

The International Classification of Functioning, Disability and Health (ICF)[29], developed by the World Health Organization, is the most developed attempt to provide such a common language. Unlike diagnostic classifications such as ICD, which describe disorders, the ICF describes how health conditions affect functioning in everyday life. Its important conceptual contribution is to understand functioning not simply as a consequence of diagnosis, but as the product of interactions between individuals and their environments, encompassing body functions, activities, participation and environmental factors.

The ICF complements rather than replaces diagnosis. Within the framework proposed in this chapter, it addresses two of the six dimensions of need: functioning and participation, and the environmental context that shapes them. It therefore provides an important foundation, but not the whole of the proposed approach.

The ICF is hierarchical, allowing both broad summaries and detailed descriptions. At population level it supports planning and commissioning; for individuals it allows richer descriptions of strengths and difficulties while maintaining consistency across settings. International experience suggests that ICF-informed approaches can improve communication between professional groups and coordination across sectors[1000-1002]. However, there is little evidence that the framework alone improves outcomes, and there are no national-scale comparative evaluations. For this reason, the Review does not recommend mandating its use.

Portugal provides the clearest national experience. Following legislation in 2008, ICF-CY descriptions were used in 93% of functioning profiles but in only 3% of the intervention plans that followed[1003]. Descriptions of functioning became more comprehensive, while the support children received changed relatively little[1004]. A survey of 551 professionals found a mean of 7.3 hours spent on each case, and training that improved conceptual understanding without improving perceived usability or procedural clarity; the presence of interdisciplinary working arrangements was the strongest predictor of whether professionals found the framework usable at all[1001]. The requirement was removed in 2018. The important lesson is not that the ICF failed, but that it became attached primarily to decisions about eligibility rather than to planning intervention – precisely the outcome this Review seeks to avoid.

Although often criticised as too large for routine use, with around 1,700 categories, the ICF is not intended to be applied in full. Generic descriptions of functioning can be used before diagnosis, while condition-specific Core Sets become relevant only once diagnosis has been established. Given how common co-occurrence is,

completing multiple Core Sets would be impractical. A single generic description of functioning, supplemented where necessary by condition-specific information, offers a more workable approach.

Mapping work undertaken for this Review compared the instruments currently used in autism and ADHD services with the corresponding ICF Core Sets. Depending on the instrument, between 8% and 41% of relevant domains were covered, while the Core Sets contained most of the areas already assessed. Although this comparison concerns coverage rather than accuracy, it suggests that many aspects of everyday functioning and environmental context are not routinely considered[1005-1008]. Published work reaches a similar conclusion. A content-validation study of 14 widely used measures found that they covered only one fifth to three fifths of the relevant Core Set domains, while 12 of the 14 contained no assessment of environmental factors at all[1009]. The gap therefore lies less in how carefully assessments are undertaken than in what they ask about. A worked example showing how a Core Set can be reduced to a practical set of specifiers, linked to existing education planning, is provided in Table 18.1.

Table 18.1 An illustrative ICF-aligned specifier set (autism example)

Profile areaWhat to describeHow it informs support
Learning, attention and executive functioningLearning abilities; attention and memory; energy and motivation; planning, flexibility, problem-solving and decision-making; and the ability to start, organise and complete tasks.Educational and workplace adjustments, workload planning, executive-function support, specialist educational advice and transition planning.
Communication and social interactionUnderstanding and using spoken, written, signed or nonverbal communication; conversation; language functions; fluency and prosody; reciprocal and formal or informal relationships.Accessible communication, speech and language strategies, support for interaction and adjustments across home, education, work and services.
Emotional, sensory and physical functioningEmotional regulation, anxiety and low mood; sleep; stress tolerance; sensory perception and sensitivity; pain; movement and coordination; eating, gastrointestinal or other relevant physical-health difficulties.Mental- and physical-health assessment, treatment and monitoring; sensory adaptations; support with sleep, eating, distress and emotional regulation.
Daily routines, self-care and independenceManaging routines, personal care, eating, safety, health, transport, money, household activities and the amount and consistency of support required.Practical and occupational-therapy support, socialcare assessment, independent-living provision and planning for transitions or lifelong care.
Participation in everyday lifeEngagement in play, relationships, education, training, employment, leisure and community life, including attendance, exclusion, withdrawal and sustainability of participation.Inclusion planning, education or employment support, access to community activities and action to prevent social or occupational exclusion.
Environmental facilitators and barriersFamily, friends, peers and professional support; attitudes, bullying and discrimination; light, sound and other sensory environments; communication technology, transport, housing, finances and access to services.Reasonable adjustments, assistive technology, family or carer support and coordinated changes across education, health, care and employment.
Co-occurring conditions, deterioration and riskDiagnosed, suspected or emerging neurodevelopmental, mental- and physical-health conditions; self-harm, suicidality, safeguarding concerns, loss of skills or patterns of crisis.Immediate treatment or safeguarding action, specialist assessment, escalation planning and coordinated neurodevelopmental, mental- and physical-health care.

Evidence on practical implementation is also beginning to emerge. In adults, generic measures such as the WHO Disability Assessment Schedule (WHODAS 2.0) have established psychometric properties and can be used alongside diagnosis[1010]. The evidence for children is much more limited. A systematic review of 143 psychometric studies identified only six using the child version, none conducted in the United Kingdom[1011]. The Review therefore concludes that any measure of functioning for children and young people should be developed and evaluated rather than adopted uncritically. Similarly, early digital implementations of autism and ADHD Core Sets report encouraging findings on feasibility, usability and communication[1012-1014], but these approaches also require larger independent evaluation.

Box 18.1 One description, four uses: an illustrative composite The following is a constructed example rather than an account of an identifiable person. It is drawn from situations the Review encountered repeatedly and is included because reviewers rightly observed that a shared description of need is easier to understand through an example than in the abstract.

A 17-year-old in the second year of a college course has begun missing classes. She has been waiting 14 months for an autism and ADHD assessment. A shared description of need is developed with her, her mother and her personal tutor. It records that she is managing the coursework but struggling with deadlines and unstructured group work; that she is sleeping poorly and describes a persistent anxiety she has not previously disclosed; that no diagnosis has yet been established and autism and ADHD remain under consideration; that her college is supportive but her part-time employer is not; that she has withdrawn from the friendships she had at school; and that she has twice disengaged from services after being asked to tell her story again from the beginning.

None of this required a diagnosis, and very little of it was new information. Each part was already known to someone; what was missing was a shared description that everyone could use. Her college introduces deadline support and access to a quiet room using arrangements it already has. With her consent, and seeing only the section relevant to workplace adjustments, her employer moves her from the front counter to stockroom duties. Her GP, recognising the difficulties with anxiety and sleep, begins treatment without waiting for the neurodevelopmental assessment to conclude. Her mother, reading the same description, understands for the first time that deadlines, rather than the work itself, are the main difficulty.

The neurodevelopmental assessment continues unchanged. If a diagnosis is made, it adds to this shared understanding rather than replacing it, and where diagnosis determines access to specific treatments or legal entitlements, those continue to follow. The important difference is that support has already begun.

The examples shown in Table 18.1 and Box 18.1 also illustrate the limits of the ICF. It is a descriptive framework rather than a decision-making one. Diagnosis explains the condition; the ICF describes functioning and participation; and needs assessment brings these together with information about risk, complexity, likely benefit and capacity to manage engagement. The three are complementary rather than competing.

The Review therefore recommends developing a concise shared language for describing need, using the ICF as its foundation but extending beyond description to support more consistent decisions. It should be developed jointly across health, education, welfare, social care and employment, with people with lived experience and families as partners throughout. It should align with the systems through which support is authorised, be independently evaluated, and have clear national ownership.

18.3 Applying needs assessment across different conditions

The same broad dimensions of need recur across mental health and neurodevelopmental conditions: participation, current functioning, risk, distress, complexity, environmental context, the purpose and likely benefit of support, and capacity to manage engagement. What differs is not the dimensions themselves, but the relative importance each assumes in understanding a person’s circumstances and determining the most appropriate response.

18.3.1 Autism: participation depends on the interaction between the individual and the environment

For autistic people, participation depends as much on the environment as on individual characteristics. The settings in which people live, learn and work profoundly influence everyday functioning. Sensory demands, communication styles, predictability, social expectations, workplace culture and educational environments all shape participation. Practical guidance on creating neuro-inclusive environments already exists but remains underused[1015].

Autistic people and their families consistently describe educational adjustments, workplace accommodations and peer support as being at least as important as diagnostic clarification itself[66, 305]. Unfortunately, support during waiting periods and after diagnosis remains inconsistent. In the clinician survey undertaken for this Review, only around half of services reported offering support while people waited for autism assessment, and post-diagnostic provision varied similarly[299]. The Review also heard from autistic adults that diagnostic reports often combine detailed personal histories with the relatively limited information schools or employers need in order to make adjustments. As a result, people may have to disclose much more than they wish, or seek a separate supporting letter. A shared description of need that stands alongside the diagnosis could help address this problem (see Section 18.4). Assessment should also consider what participation costs the individual rather than inferring need from attendance or achievement alone[321].

One group requires particular attention. Chapter 8 described autistic people whose difficulties emerge early and who require substantial, often lifelong, support. For this group, environmental adaptation remains important but cannot remove the need for sustained, skilled support throughout life. Needs assessment must therefore be capable of describing support at this level of intensity, including communication, dependency, behaviour that presents risk and physical health. Whether the levels proposed in Annex A achieve this adequately should be tested during implementation.

Autism illustrates one of the central principles of needs assessment: participation depends as much on the interaction between individuals and their environments as on diagnosis itself. Even in supportive settings, many people continue to struggle with administration, social demands, loneliness, independent living or becoming overwhelmed by everyday activities. Needs assessment must be broad enough to capture these experiences and be undertaken by people who understand them.

Environmental adaptations cannot remove every difficulty, but appropriate support enables many autistic people to participate and thrive.

18.3.2 ADHD: understanding variability rather than average performance

ADHD illustrates a different aspect of needs assessment. Whereas autism highlights the importance of environmental fit, ADHD highlights variability across time and settings.

A defining characteristic of ADHD is not simply impaired performance but inconsistency in the deployment of skills. People may function effectively under structured, motivating or externally supported conditions while experiencing substantial impairment in routine or unstructured situations. Functional assessment therefore requires information gathered across settings and over time rather than from a single clinical encounter[1016, 1017].

Symptom counts and DSM subtypes are relatively poor predictors of long-term outcomes. Functional impairment, co-occurring conditions, emotional dysregulation and environmental adversity are more consistent predictors of educational attainment, occupational outcomes, psychiatric morbidity and persistence into adulthood[72, 230, 234, 237, 240].

Different interventions influence different aspects of functioning. Medication is particularly effective for attention, concentration and productivity, whereas psychosocial interventions often have greater impact on relationships, educational participation and family functioning[256]. Where ADHD is suspected, psychosocial

interventions can provide valuable support while assessment is underway. Medication, however, requires detailed clinical assessment before it can be initiated. Consistent approaches to assessing impairment, both during waiting periods and at clinical assessment, are therefore important in identifying those most likely to benefit from medication without unnecessary delay. Where assessment demonstrates ADHD with moderate to severe impairment, NICE[260] recommends medication and psychosocial interventions together.

18.3.3 Mental health conditions: understanding dynamic change

Mental health conditions illustrate the importance of responding to change over time. Diagnosis may remain stable while functioning, risk and the support needed for recovery, education, employment and relationships change.

Anxiety and depression may fluctuate between distress, clinically significant disorder, crisis and recovery, requiring support proportionate to severity, duration, risk and impairment. Eating disorders, self-harm and emerging psychosis may carry substantial risk before outward functioning has collapsed; urgency therefore depends on trajectory and risk as well as diagnosis[1018].

18.3.4 Co-occurrence: the strongest test of needs assessment

The greatest challenge for diagnosis-dependent systems arises when people present with multiple interacting conditions – which, as Chapter 9 shows, is the rule rather than the exception. Mental health and neurodevelopmental conditions, communication difficulties, physical health problems and social adversity frequently occur together, and as they accumulate the likelihood of severe multi-domain impairment increases[237, 238]. High levels of neurodevelopmental and mental health traits also co-occur more frequently than recognised disorders[1019, 1020]. Systems organised solely around confirmed diagnoses therefore risk overlooking substantial need. Comprehensive profiles of strengths, difficulties, context and functioning allow support to begin before secondary difficulties become established (see Chapter 9). As Annex A outlines, complexity reflects the interaction between conditions rather than simply the number of diagnoses.

Physical health also needs explicit attention. Chronic illness, obesity, pain, sleep problems, gastrointestinal disease, epilepsy and medication effects are often under- recognised, particularly where communication is limited. People with neurodevelopmental conditions and severe mental illness should receive proactive physical health review rather than care only when problems are reported[351, 1021, 1022].

18.4 Matching support to need

The remaining question is how this approach would change the organisation of services.

First, meaningful support should begin before diagnosis wherever possible. Support available while people wait for assessment is often limited, inconsistently commissioned and rarely evidence based. As a result, diagnosis has become the point at which meaningful help begins for many families. Earlier chapters describe the consequences from different perspectives: long waits for neurodevelopmental assessment (Chapters 7 and 8), many children referred to mental health services never reaching treatment (Chapter 15), and increasing numbers of statutory education plans issued outside statutory timescales (Chapter 13). Together they describe the same underlying problem. Support has been placed behind assessment, and assessment has become the bottleneck.

A needs-led approach would begin when participation starts to break down, allowing proportionate support and review while diagnosis proceeds where appropriate. Practical support, reasonable adjustments, psychoeducation and active monitoring can all begin before diagnostic certainty is achieved.

Existing models offer useful lessons. Wales’s Nest/Nyth framework adopts a ‘no wrong door’ approach organised around family circumstances rather than diagnostic thresholds[1023], while the Welsh Mental Health and Wellbeing Strategy 2025–2035 extends a connected, needs-led approach across the life course[1024]. Evidence also shows that functioning often deteriorates during long waits[19, 72], whereas parenting support, psychoeducation, educational adjustments and active monitoring can begin much earlier[19, 260].

Second, support should be matched to need throughout the person’s journey, both before and after diagnosis. It should reflect current functioning, participation, risk, complexity, environmental circumstances and the interventions most likely to help, rather than assuming that everyone with the same diagnosis requires the same pathway.

Such a system has four defining features. Support is layered, offering different levels of intensity; it is matched to current need rather than diagnosis alone; it is available early, without waiting for diagnostic thresholds; and diagnosis is incorporated within a broader assessment rather than acting as its prerequisite (Recommendation 2).

In mental health, urgent care is already organised largely around presentation and risk rather than diagnosis. It is in neurodevelopmental pathways, education and

welfare that diagnosis most commonly determines access. Part of the change required needs no new legislation. The duty to make reasonable adjustments under the Equality Act 2010[50] is triggered by substantial long-term impairment in day-to- day activities, not by diagnosis. Schools, colleges, universities and employers can therefore act on a description of functioning and its consequences without waiting for diagnostic confirmation, and many already do. The widespread belief that diagnosis is required has itself contributed to diagnosis becoming the gateway. Making the legal basis clearer, and describing need in terms that align with it, would allow much of the support people currently wait for to begin immediately.

Third, needs assessment should not be a one-off event. Needs change with development, life transitions and response to intervention. Assessment should therefore be reviewed when circumstances change or at key transition points, with support adjusted accordingly. Reviews should be undertaken with the person, who should also be able to request them, while a named professional maintains continuity wherever possible.

Fourth, a shared description of need should improve continuity between systems. As discussed earlier, fragmentation generates duplication, delay and discontinuity. A shared description of need, reviewed over time and expressed in a common language, would allow information to travel with the person rather than being recreated repeatedly. Individual organisations would retain their own professional judgement and statutory responsibilities, but decisions would begin from a shared understanding. The Review proposes that the general practice record should remain the principal repository of health information, with the NHS App providing access rather than functioning as a separate record.

This raises important questions about confidentiality. Consent to share information should remain with the individual, or with parents where appropriate, and should be withdrawable at any time without affecting access to support. What should travel is the description of need and the adjustments it implies, not the underlying clinical history. Access should be restricted according to professional role and governed by the Caldicott principles. People should also be able to see, correct and contribute to their own description of need, because it belongs to them. The Review also supports greater portability of care, for example through an ADHD passport so that established medication can continue when people move location rather than requiring reassessment (Recommendation 7). Annex A sets out these operating principles.

Two further safeguards are essential. Non-digital routes must always remain available so that support is never limited by digital access or confidence. Equally,

people are less likely to disclose sensitive information if they believe it will be widely shared.

Finally, needs assessment is not intended to replace professional judgement with algorithms or numerical scores. Decisions about support require clinical interpretation and dialogue with individuals and families. Information collected in advance allows professionals to build on what is already known rather than repeatedly asking people to recount difficult experiences. Annex A sets out which parts of assessment require different levels of expertise. Structured assessment should inform decision-making, but responsibility for judgement, risk assessment and support decisions must remain with appropriately trained professionals.

Ultimately, a system organised around needs assessment would judge success less by the number of assessments completed or diagnoses recorded than by whether support improves participation, prevents deterioration and enables people to live the lives they wish to lead. Many such outcomes – including school attendance, exclusion, crisis presentations, sustained employment and the proportion of young people not in education, employment or training – are already measurable and could become part of routine system accountability. Chapter 20 returns to this question.

18.5 Accountability and safeguards for a needs-led system

Previous attempts to complement diagnosis with structured, earlier assessment of need have had mixed success. They provide important lessons for future implementation.

The first is that descriptions of need can easily become confused with decisions about entitlement. Frameworks developed to improve communication and understanding may gradually become mechanisms for determining eligibility, allocating resources or restricting access to services. In doing so, they become another gateway. Such systems encourage people to maximise scores, emphasise deficits and contest decisions, undermining trust in both assessment and services.

The experience of the Work Capability Assessment and Personal Independence Payment illustrates this risk. Both introduced structured descriptions of functional capability to improve consistency. Once assessment outcomes became directly linked to entitlement, however, they came to determine access to resources, undermining public confidence[1025]. Describing need and deciding entitlement are related, but they should remain distinct. A similar lesson emerged from the NHS Mental Health Care Clusters. Although they grouped people according to patterns of need, implementation was undermined by variable reliability, inconsistent application

and the growing influence of payment mechanisms on clinical practice[1026]. The difficulty lay not in describing need, but in turning a descriptive framework into a financial instrument before its reliability, validity and fairness had been established.

More recent experience in neurodevelopmental pathways illustrates a different risk. Structured digital profiling has shown promise but has also demonstrated how easily proportionality can be lost. In one pilot using the Do-IT Profiler, escalation criteria were so restrictive that only a small group of people in crisis received a different response, while most remained within the same undifferentiated waiting list[1027]. Similar patterns were reported in some ADHD pathways. Needs assessment therefore has to distinguish meaningful differences across the whole population, not simply identify those already in crisis.

A further lesson concerns equity. New approaches are often adopted first by those with the confidence, resources and skills to navigate them, reproducing the familiar pattern of the inverse care law. A shared description of need may favour those who can articulate their difficulties most easily, while digital systems favour those confident with technology. A needs-led system should therefore be judged not only by whether it describes need more accurately, but by whether it reduces inequalities in participation and access to support.

The Review therefore proposes a number of principles to guide implementation.

First, responsibility for decisions should remain with appropriately trained professionals. Structured descriptions of need should inform, but never replace, professional judgement. They should also be developed with people and their families rather than about them, and where professional and personal accounts differ, both should be recorded.

Second, needs assessment should remain descriptive rather than deterministic. It should describe participation, functioning, risk and context without automatically determining access to treatment or support.

Third, descriptions of need should be dynamic rather than fixed. They should be reviewed whenever significant changes occur.

Fourth, assessment should promote equity rather than rationing. Particular attention should be given to people who have historically experienced barriers to recognition or engagement. Judgements about the likely benefit of support should never direct resources only towards those most likely to improve, and capacity to manage engagement should never become a reason to offer less support.

Fifth, systems should distinguish clearly between describing need and allocating resources. Assessment should improve understanding rather than create new eligibility thresholds. Decisions about entitlement remain matters of professional judgement, statutory responsibility and public policy.

Finally, implementation should be accompanied by continuous evaluation and refinement, undertaken in partnership with the people who use the system. Evaluation should examine not only reliability but also effects on equity, participation, waiting times, user experience and outcomes, while remaining alert to unintended consequences.

Ultimately, the success of needs assessment should not be judged by how consistently people can be classified, but by whether it enables earlier support, improves participation, reduces unnecessary delay, strengthens trust and continues to guide care long after diagnosis.

18.6 From principle to practice

The Review’s overall conclusion is one of cautious optimism. Many of the components of a needs-led system already exist, but the evidence remains at an early stage, and implementation should proceed through careful evaluation rather than assumption.

Early work suggests that clinical complexity can be described separately from the likelihood of diagnosis, although the evidence is preliminary and should not yet be taken to demonstrate improved pathways. Any complexity-based model requires prospective evaluation of access, outcomes, equity and unintended gatekeeping before wider adoption.

Some local systems have introduced single points of access that organise referrals around presenting need rather than requiring people to choose a diagnostic pathway at the outset[1028]. Although promising, published evidence on integrated neurodevelopmental and mental health pathways remains limited, and these approaches require prospective evaluation rather than assumed benefit (Recommendation 8).

Evidence is also emerging for structured descriptions of functioning.

Autistica’s feasibility study of an ICF-informed digital strengths-and-needs toolkit found that it improved shared understanding and collaborative planning between families and education professionals[1029]. The tool focused on functioning, participation and environmental barriers rather than diagnosis. Its effectiveness, portability, equity and cost-effectiveness now require larger independent evaluation.

Digital technologies may strengthen these developments further. Structured information collected before appointments can reduce duplication, improve consistency and give people and families greater opportunity to describe their experiences. Digital platforms may also help update descriptions of need over time and improve portability across services.

The Review does not recommend adopting any existing model nationally at this stage. Rather, it observes that many of the elements of a needs-led system already exist but have developed independently for different purposes. Annex A brings these elements together into a single operational framework, offered for demonstration, testing and independent evaluation rather than as a national standard.

Ultimately, the purpose of needs assessment is not simply to describe need more accurately but to improve access to effective support. Evaluation should therefore examine whether it leads to earlier intervention, better matching of evidence-based support, improved participation and equity, reduced delay and duplication, and better long-term outcomes.

The approaches already emerging – ICF-informed assessment, integrated access pathways, integrated assessment and structured descriptions of complexity – provide valuable starting points. Their success should be judged not by whether they classify people more accurately, but by what follows: whether support begins earlier, participation improves, unnecessary delay and duplication are reduced, long-term outcomes improve, and inequalities in access narrow.

Chapter 19. A model of care matching support to need, functioning and participation

The current system too often follows a sequence of wait, assess, diagnose, then support. The pathway proposed in this Review starts from a different principle. It seeks first to recognise need, understand what is happening, begin appropriate support, establish diagnosis where it changes treatment or entitlement, and review and adapt support as circumstances change. Support should be matched to need, functioning and participation, not diagnosis alone.

Mental health and neurodevelopmental support should increasingly be organised through a single coordinated pathway, working closely with education, social care and other local services. Specialist expertise should function increasingly as an expert at hand, supporting neighbourhood teams through consultation, supervision, shared care and quality assurance, rather than requiring everyone to enter a specialist service.

Digital infrastructure should ensure that people have to tell their story only once. Used well, it can support navigation, improve coordination, reduce duplication, increase accessibility, and generate the information needed to improve services over time.

Ultimately, success should be judged not by referrals, contacts, diagnoses or waiting times alone, but by whether people are better able to participate in education, work, relationships, family life and their communities.

The preceding chapters have shown that support is no longer organised in a way that reflects how need is experienced. Diagnosis has become the route through which support is often reached, yet it is only a partial guide to what a child, young person or adult needs (Chapter 18). Care too often begins only after diagnostic thresholds have been crossed, even though functioning may deteriorate while people wait (Chapters 14 and 15). Mental health difficulties, neurodevelopmental conditions, communication needs and social adversity co-occur so consistently that services organised around single conditions struggle to reflect the people they are intended to serve (Chapter 9). The outcomes that matter most to people – participation in

education, relationships, family life, work and community – remain only loosely connected to how services are organised and judged.

This chapter describes what a better pathway could look like in practice. The detailed framework for assessing need is set out in Chapter 18 and Annex A; here the focus is on how that framework is applied. We propose a single integrated pathway for mental health and neurodevelopmental support from conception to 25 years, organised around need, functioning and participation rather than diagnosis alone. The pathway is grounded in the evidence reviewed throughout this report and illustrated by examples of emerging practice. Its successful implementation will depend on continuous learning, robust data, independent evaluation and regular review. The recommendations needed to achieve this are set out in Chapter 20.

19.1 Why one integrated pathway is needed

People rarely present with one clearly defined difficulty that fits neatly within a single service. A child referred because of attention difficulties may also be anxious, sleeping poorly and struggling with language. A teenager with low mood may also be autistic, missing school and in conflict at home. Adults likewise move between primary care, autism or ADHD assessment, talking therapies, occupational health, employment support, social care and specialist services. Co-occurrence is the rule rather than the exception, and it is often the interaction between difficulties, rather than any single diagnosis, that determines how people function (Chapter 9).

19.1.1 An effective response to a growing need

People and families experience these difficulties as one unfolding story, yet services have developed around separate conditions, professions and thresholds. National analyses of neurodevelopmental pathways identify separate autism and ADHD pathways as sources of duplication, inefficiency and longer waits[1030], a conclusion echoed by the Independent ADHD Taskforce[203]. More than one million children and young people had an active referral to mental health services within a single year, around one-third were still waiting at the end of that year, and around 60,000 had waited for more than two years[757]. Functioning often deteriorates while people wait. Forty-two per cent of community mental health service users reported that their mental health had worsened during the wait for care, rising to 71% among those waiting longer than six months (Chapter 14), although these findings should be interpreted cautiously given the survey response rate of 19%[19].

The scale of need underlying these waits is substantial and is projected to continue increasing (Chapter 6, Section 6.7). Among people aged 0–25 in England, an estimated 4.3–4.5 million experience clinically significant mental health need when

emerging and subthreshold difficulties are included; around 2.6 million meet criteria for a diagnosable disorder; 1.1–1.3 million experience severe or very severe difficulties; and between 140,000 and 200,000 have highly complex, severe or high- risk needs (Table 19.1).

Table 19.1 Illustrative modelling of the approximate distribution of mental health need across the four levels (England, ages 0–25)

Level of needTypical responseIndicative scale (0–25 years)
Level 1 – emerging / mildUniversal and community settings: parent and family support, early help, psychoeducation, adjustments, guided selfhelp~1.7–1.9 million
Level 2 – moderate (‘missing middle’)School-linked and neighbourhood teams, matched therapies, Mental Health Support Teams, early support hubs~1.2–1.4 million
Level 3 – severe / complexCoordinated specialist services organised around one plan~1.0–1.1 million
Level 4 – very severe / highriskSpecialist and crisis services, intensive coordination, inpatient care where necessary~140,000– 200,000

Source: for children aged 0–5 years where no representative survey exists, estimates are derived from limited English evidence together with Nordic preschool cohorts and should be treated as indicative. For ages 6–15 years the MHCYP 2017 and 2023 is used. For ages 16–25 years APMS 2023/24 is used and cross- referenced with 2023 MHCYP. These estimates describe mental health need only. Autism is estimated separately and ADHD only partly captured, so the overall level of mental health and neurodevelopmental need is higher than shown here. Population denominators are based on ONS mid-2023 estimates for England.

The four levels describe the intensity of response, not the number of people requiring a particular service. Although symptom severity is greatest among 16–25-year-olds, most would be expected to receive support at Levels 1 or 2 rather than specialist services, because the young-adult figures derive from a symptom-severity threshold rather than from a measure of service need. The estimates also combine different measures – symptom questionnaires, clinician-rated diagnoses and severity thresholds – which capture related but different aspects of need. Because symptom severity and functional impairment have increased over time, these estimates are, if anything, likely to be conservative. Diagnosable need corresponds broadly to levels 2 to 4 (around 2.6 million people) while all clinically significant mental health need,

including emerging and subthreshold difficulties, spans all four levels (around 4.3– 4.5 million people).

These figures should not be confused with the estimate in Chapter 6 that around four million children and young people experienced high psychological distress in 2024. The two are broadly consistent in scale but are not directly comparable. They should therefore be interpreted as illustrative estimates of the order of magnitude of need for planning purposes rather than precise prevalence estimates, a basis for determining individual eligibility or a commissioning entitlement.

Responding effectively to this level of need requires a different system. The pathway proposed in this Review spans prevention and public health, early non-clinical support, community-based care and specialist services. Its primary purpose is to improve people’s lives by enabling support to begin earlier, match need more closely and improve participation. The economic case is also compelling. The costs of unmet need – and the benefits of effective support – fall across education, employment, social care, health, welfare and the wider economy. UK analyses estimate the annual cost of late intervention at around £17 billion, while modelling suggests substantial returns from parenting and early years programmes[690, 701, 784,

785, 815]. These benefits, however, depend on effective implementation, sufficient capacity and sustained collaboration between services. The current system spends increasing resources managing delay, duplication and crisis. The proposed pathway aims to direct those resources towards earlier and more effective support.

The pathway improves public service productivity in four ways. It reduces repeated assessments and parallel waiting lists by collecting information once and using it across the pathway. It helps people reach the right support earlier, reducing avoidable deterioration. It enables scarce specialist expertise to support more people through consultation, supervision, shared care and digitally enabled multidisciplinary working rather than repeated assessments. Finally, it allows commissioners to organise resources around need and outcomes rather than activity alone.

19.2 Principles of the pathway

Many routes, one system

Need may first be recognised during pregnancy, in the early years, in nurseries, schools, colleges or universities, in primary care, youth and community services, through self-referral or within specialist care. People should be able to enter the support system wherever need is recognised, without having to understand how services are organised – the principle of ‘no wrong front door’. Where access

depends on families or individuals identifying the correct service in advance, those least able to navigate complex systems are least likely to receive timely support.

Organised around functioning and participation, not diagnosis alone

What usually brings a person to attention is not the diagnosis they may eventually receive but a change in everyday life: attendance falling, learning becoming harder, work becoming unsustainable, relationships coming under strain, distress increasing, family life becoming more difficult or independence reducing. Understanding need therefore requires attention to functioning, participation, risk, complexity, environmental context and the likely benefit of support, while retaining diagnosis as essential but not sufficient. Expertise in autism, ADHD, other neurodevelopmental conditions, mental health, communication, learning, sensory needs, trauma, risk, family context and work participation should be coordinated within a single pathway, rather than requiring people to move between separate systems organised around different conditions.

Matched to need, and responsive as need changes

Matched care means offering the type and intensity of support that best reflects a person’s current needs from the outset, rather than requiring progression through fixed service levels before more appropriate support becomes available. Difficulties emerge, improve, recur and change over time. Support should therefore be able to increase or reduce in intensity as need changes, without people repeatedly leaving and re-entering the system. This approach is supported by evidence that matched care can produce better recovery than stepped care[800], by emerging work matching support to need from the outset for young people with anxiety[1031], and by the Thrive framework’s needs-based organisation of support[801].

Person-led and relational

Children, young people, families and carers should be active partners throughout the pathway. Evidence-based support for parents and carers should be available at every level, recognising both the responsibilities they carry and the pressures they experience[782, 1032]. Support should be based on shared decision-making, peer support and, where evidence allows, genuine choice between interventions, with autonomy and consent increasing appropriately with age – particularly during the transition from adolescence to adulthood. It also depends on building trust and continuity over time, rather than focusing solely on assessment, diagnosis or treatment.

Participation and equity should be routine outcomes

Services should be judged by the outcomes they can reasonably influence, while recognising the wider circumstances affecting people’s lives. They should not retreat to measuring symptoms, activity and waiting times alone. Participation in education, employment, relationships, independent living and community life – including reductions in the number of young people not in education, employment or training (NEET) and successful transitions after age 16 – should become routine measures of success. The Patient and Carer Race Equality Framework should be used to strengthen anti-racist practice and reduce racial inequalities in access, experience and outcomes.

Figure 19.1 Schematic diagram of the shift from a diagnosis-dependent system to one that provides support in response to need alongside diagnosis

These principles depend on sufficient capacity, effective coordination across agencies, a workforce able to work across organisational boundaries, and meaningful measures of functioning and participation. Otherwise, even a needs-led system can drift back towards the very severity thresholds that deny early support

today[745]. Evaluation of the i-THRIVE programme suggests that needs-based approaches improve accessibility and shared decision-making when implemented well, but work best where strong multi-agency collaboration already exists[802, 803].

19.3 Recognising and understanding need

The pathway begins wherever concern is first recognised. A parent may speak to a health visitor. A teacher may notice attendance falling. A young person may seek help online. A GP may respond to a family’s concerns. A child may present through an Early Support Hub, youth service, voluntary organisation, crisis pathway or specialist service. Adults may similarly seek help through primary care, occupational health, employment support, social care, voluntary organisations or specialist services.

Because stigma, blame and other barriers continue to delay help-seeking[778, 1033], services should identify emerging difficulties proactively rather than waiting for crisis. This is particularly important during childhood and adolescence, when neurodevelopmental differences emerge and mental health trajectories are most responsive to early intervention. Examples include using information from the perinatal period, preschool surveillance, school-based developmental review, the iCATS-i2i approach[701], and integrated detection-and-care pathways for autism[1034] (see also Chapters 13 and 14). As throughout the Review, these approaches require careful attention to false positives and false negatives.

Detection at this stage should be matched and, where evidence supports it, condition-specific. A profile of functioning answers one question: what support does this person need now? Brief validated detection tools answer another: what condition-specific expertise might also be needed? The two questions are complementary. Validated instruments already exist for this purpose. The SDQ identifies children at increased likelihood of psychiatric disorder using information from multiple informants[1035]. A brief two-item parent-report measure developed in English primary schools identifies children with probable anxiety disorders[1036]. In Australia, a two-stage developmental surveillance (Social Attention and Communication Surveillance-Revised (SACS-R)) delivered by community child health nurses identified autistic children in their second and third years of life with a positive predictive value of 83% across more than 13,000 children[1037]. None of these requires specialist administration.

This distinction matters. A needs-led system is sometimes interpreted as one that sacrifices diagnostic specificity. It does not. Brief, validated, condition-specific detection complements rather than replaces a shared description of need. It can be delivered by trained community practitioners and helps ensure that people reach the

right expertise earlier. Detection is not diagnosis and should never be presented as though it were.

Three important constraints apply. First, brief detection tools deliberately identify more people than will ultimately receive a diagnosis. For example, in community samples the SDQ has a positive predictive value of around one in two for any disorder. A positive result should therefore open a conversation, not close one or become a label. Second, the strongest evidence for developmental surveillance currently comes from outside England. The autism pathway described above was developed in Australia, although tested widely internationally, and the UK National Screening Committee does not currently recommend population screening for autism in young children[1038]. Third, England’s existing universal review at 2–2½ years currently uses a general developmental assessment and contains no structured mental health or neurodevelopmental component. Extending it would therefore represent a genuine service development requiring careful evaluation, and for any autism component a further view from the National Screening Committee, rather than a simple modification of existing practice. An early detection-care pathway for autism including SACS-R[1034] is already, however, successfully piloted and evaluated in England (Greater Manchester, Midlands, Humber and North Yorkshire) with NHSE support. Chapter 13 discusses the wider evidence on the strengths and limitations of screening.

Detection is valuable only when it leads directly to support. Screening without a response simply identifies need that the system then fails to meet. The evidence that identification improves outcomes comes from studies in which identification and support were evaluated together[690]. Where families can complete brief screening themselves and be directed immediately to an appropriate first step of support, that digital route should be designed as part of the pathway rather than added afterwards.

A defining feature of the pathway is that support and diagnosis, where required, proceed in parallel. From first contact, the system should begin to build a shared description of need, capturing functioning, participation, risk, complexity and environmental context while enabling support to begin immediately. Initial information may be gathered by trained practitioners, supported where appropriate by digital tools.

This information should guide three distinct decisions:

  • What support should begin now?
  • How urgent is the response?
  • What level of expertise is needed for further assessment or treatment?

These questions should not be collapsed into a single score or level. Immediate risk may require urgent intervention without diagnostic complexity, while substantial diagnostic uncertainty may require specialist assessment without constituting an immediate crisis.

Experienced clinical oversight at the point of entry is essential. Decisions about urgency, significant risk, diagnostic complexity, the direction of assessment and pre- diagnostic support require senior professional judgement. Specialist expertise should therefore be concentrated where uncertainty or potential harm is greatest.

Three points about diagnosis should be clear.

First, the Review does not propose changing the role of diagnosis in determining access to NHS treatments or other services that depend upon it, nor does it propose changing diagnostic criteria. Diagnosis remains essential for treatment, specialist intervention, and the legal rights and protections that depend upon it.

Second, the proposed reform has two complementary strands working in parallel: an assessment of need that allows practical support to begin immediately, and a high- quality diagnostic assessment that informs treatment and specialist care. Far from reducing the importance of diagnosis, the Review places greater emphasis on accurate assessment of co-occurrence and differential diagnosis, undertaken in accordance with NICE guidance and established classification systems.

Third, needs-led support is additive. It should never become another hurdle before diagnostic assessment, nor a reason to delay or withhold diagnosis where diagnosis is needed.

Where support is provided without an adequate understanding of the underlying condition – for example, where an autistic young person’s difficulties are misunderstood – support can be not merely insufficient but actively harmful. Equally, poor-quality or inaccurate diagnosis can itself cause harm (Chapter 5).

Chapter 18 and Annex A describe the proposed framework for assessing need in detail. They distinguish six dimensions: current functioning and participation; risk of deterioration; complexity and co-occurrence; environmental context; purpose and likely benefit of support; and capacity to manage engagement. Together these guide and authorise support; they are not intended to ration it.

The pathway should operate on the principle of ‘no wrong front door’, with active support for navigation. Wherever possible, support should be available close to

home through neighbourhood services, including youth centres, early support hubs and education settings[1039]. Public service directories should support self-referral, while primary care should coordinate access to community, voluntary and social prescribing services rather than defaulting to specialist referral.

From first contact, each person should have a portable description of need, held digitally within the NHS, accessible to those providing care and held by the person or family so that it travels with them (Annex A). That shared description should authorise support from the outset and remain with the person as they move between services, improving continuity, particularly during transitions[802, 1040]. It should be reviewed at agreed intervals using validated measures of functioning together with structured, person-centred review, for example Dialog+ or WHODAS 2.0[1041]. People should be able to step up, step down or move between services without having to begin again.

Case study: Sutton Neurodevelopmental Needs One Stop Shop – support before diagnosis

This pilot was developed in response to rising demand for neurodevelopmental assessment and the impact of long waiting times on children, young people and their families.

It provides a drop-in service for families of children and young people aged 5–18 whose difficulties may relate to autism, ADHD or both. Presenting concerns include emotional difficulties, behaviour, sleep, friendships, sensory needs, attention and concentration, hyperactivity and impulsivity. Families do not need an appointment and no diagnosis is required to access support. The service brings together professionals from CAMHS, Mental Health Support Teams, parenting services, disability services, Sutton Information, Advice and Support Services, Playwise and Sutton Carers Centre.

The example illustrates how a needs-led approach can begin with the difficulties a child or family is experiencing, rather than waiting for diagnostic assessment to be completed. It is included to illustrate a service model rather than as evidence of effectiveness. The model has not yet been independently evaluated for outcomes.

19.4 Matching support to need

Support should be matched across a small number of broad levels of need, described in Chapter 18 and Annex A and summarised in Table 19.1. These levels

describe the intensity of response, not categories into which people are placed. As needs change, people should be able to move between levels while the shared understanding developed through assessment and formulation remains with them. Stepping down should not automatically mean discharge. It should mean a planned reduction in the intensity of support, with a clear route back should circumstances change. People should not have to repeat their story, restart assessment or re- establish eligibility each time their needs change. Planned transitions, continuity of relationships, gradual reductions in contact and clearly signposted routes back into care all help sustain progress[756].

19.4.1 Beginning early: proactive intervention

The greatest opportunities to improve mental health arise during pregnancy, infancy, the preschool years and the transition into, and through, school, when development is most responsive to support and difficulties are most amenable to change. The Review found that the largest group of children with mental health need are those experiencing emerging emotional, behavioural or relationship difficulties – early reductions in functioning that have not yet reached diagnostic thresholds but are beginning to affect everyday life. Timely support for this group has the greatest potential to improve life chances and may generate substantial social and economic benefit.

The aim is to prevent avoidable deterioration. In the earliest years, support should strengthen parent–child relationships, communication, emotional regulation, sleep, routines and family confidence before difficulties become entrenched. Neither this group, nor those with more severe needs who often join the same assessment queues, are well served by a system organised around high diagnostic thresholds.

As children grow older, support should focus on sustaining participation in education, friendships, family life and the wider community.

Working with parents and carers is central to this approach, but it is not a single intervention. It is a family of evidence-based approaches, each matched to different needs. Parent-mediated social communication interventions reduce autism symptom severity, with benefits persisting for around six years after treatment[864], and the proposed mechanism has been linked to increases in child-initiated communication during therapy[1042]. Pre-emptive intervention for infants showing early signs of autism reduced the proportion meeting diagnostic criteria at age three from around one in five to one in fifteen[1043]. Behavioural interventions are effective for infant sleep problems[868], while caregiver training improves feeding in children with paediatric feeding disorder[869]. For child anxiety, parent-led digital CBT with therapist support was non-inferior to usual specialist treatment while using around 40% less therapist time[706]. For conduct difficulties, individual participant data from fifteen

European trials found no evidence that parenting programmes were less effective for families experiencing poverty, lone parenthood, unemployment, lower educational attainment or minority ethnic status[1044], nor for children between two and eleven years of age[871]; benefits were broadly sustained at follow-up[1045]. Parenting support also has an important role where the parent’s own mental health is the presenting difficulty[873].

The evidence is not uniformly strong, and the Review should not present it as though it were. The largest synthesis of early autism interventions found considerable variation between approaches and outcome measures, while adverse effects remain inadequately monitored[1046]. The implication is not that one programme should be commissioned nationally, but that parenting and family support should be commissioned as a matched portfolio of evidence-based interventions, and evaluated as it develops.

Beginning early is not, however, the same as doing anything early. The evidence supports targeted, relational, family-focused and developmentally informed support. It does not support indiscriminate whole-school psychoeducation, where several trials have shown little benefit and, in some cases, harm (Chapters 13 and 15). The following evidence-based actions should therefore form part of the pathway:

  • Invest in the perinatal period and the first 1,001 days. Ensure that relevant

information – such as prenatal alcohol exposure and very preterm birth – can inform later assessment where appropriate. Develop integrated local pathways bringing together maternity services, health visiting, Best Start Family Hubs, early years mental health, parental mental health and specialist perinatal mental health services so that families receive timely support matched to need. (Evidence maturity: moderate.) (Recommendation 3)

  • Use existing universal contacts to identify emerging difficulties earlier.

Build brief, developmentally sensitive checks for children and parents into pregnancy and perinatal contacts, the 2–2½-year review, school readiness, the transition from primary to secondary school and post-16 transition. Deliver support through Best Start Family Hubs, health visiting, early years settings and schools. (Evidence maturity: moderate.) (Recommendation 1 and 3)

  • Strengthen evidence-based family support in the early years. Parent–

infant, parenting and whole-family interventions should be available early. Video-feedback approaches such as VIPP-SD improve parental sensitivity[1047], while behavioural parenting programmes such as the Incredible Years improve child behaviour[861, 862]. For school-aged children and young people, expand parenting and parent–child support through workforce

development and greater access to evidence-based digital self-help and therapist-supported interventions, such as OSI for child anxiety[706]. (Evidence maturity: strong.) (Recommendation 1)

  • Provide parent-mediated interventions for preschool children with

emerging autism or ADHD-related needs without waiting for diagnosis. Examples include PACT and iBASIS-VIPP[864, 1043]. Economic evaluations suggest that these interventions can repay their additional early costs over time, although estimates differ according to setting, time horizon and perspective[784, 785]. (Evidence maturity: strong.) (Recommendation 2)

  • Make family support continuous rather than episodic. Families should be

able to return for structured reviews at key developmental transitions without requiring a new referral, supported by a record that carries forward previous assessments and interventions. The factors influencing the onset of difficulties are not the same as those influencing their later course; children who benefit from early intervention may require further support as circumstances change. (Evidence maturity: moderate.) (Recommendation 3)

  • Deliver support where families already are, using the workforce already

there. Detection and brief family-focused support should be provided through Best Start Family Hubs, Mental Health Support Teams and education settings, delivered by trained community practitioners working with specialist supervision. Workforce development and supervision should be commissioned alongside the model, not added afterwards. (Evidence maturity: moderate.) (Recommendation 6)

19.4.2 Neighbourhood delivery and the missing middle

Most support should be provided not in specialist clinics but in the settings where people already live their lives. Perhaps the clearest service gap identified throughout the Review concerns children and young people whose needs are recognised by everyone but owned by no one. Their difficulties exceed what schools, primary care and universal services can reasonably provide, yet remain below the threshold for specialist services. The ‘missing middle’ is therefore more than a gap between services; it is a gap in responsibility. The aim is not simply to relocate care into the community, but to organise expertise differently so that support is matched to need.

The proposed model requires two important changes.

First, waiting should be understood as part of care rather than an interval before it. Timely support is defined not only by the speed with which diagnosis is reached but by the speed with which effective help begins. Waiting lists should

therefore be prioritised according to functional need, complexity and risk so that those with the greatest impairment receive proportionate support as early as possible (Evidence maturity: moderate.) At present, only around 55% of clinicians report routinely offering support while people wait for an autism assessment (Chapter 8). This should change. Waiting should include active monitoring, evidence- based psychoeducation, practical support, parenting support, peer support and planned review, supplemented where appropriate by evidence-assured digital tools. Where anxiety, depression, eating disorders, self-harm, trauma-related distress or psychosis are present, appropriate mental health assessment and treatment should begin without waiting for the neurodevelopmental assessment to be completed (Evidence maturity: moderate.)

Second, support should become more structured, multidisciplinary and clinically informed while remaining, wherever possible, embedded within everyday settings. In adult services, a comparable ‘missing middle’ contributes to NICE-recommended psychological therapies for eating disorders, psychosis, bipolar disorder and addictions remaining substantially under-delivered. Schools, family hubs, youth services, primary care, employers and community services should be able to draw on specialist advice, supervision and, where needed, direct assessment without requiring everyone to enter a separate specialist queue. This is the ‘expert at hand’ principle.

Earlier support must not become lower-skilled support. Neighbourhood models will succeed only if practitioners have ready access to specialist expertise. Task-sharing supported by specialist supervision has been shown to be effective across a range of settings[1048, 1049]. Child psychiatry consultation programmes increase the complexity of care that can safely be provided in primary care[1050], while collaborative care models consistently achieve better outcomes than usual care[1051].

Case study: Stort Valley & Villages CYP Social Prescribing and Neurodiversity Service – one clinical home, neighbourhood support

This service illustrates how primary care can provide an integrated neighbourhood response for young people with emotional, social and neurodevelopmental needs. Developed for young people aged 11–25, it was designed to provide earlier support, prevent avoidable escalation, and help young people while they wait for – or in some cases avoid the need for – specialist intervention. The model has since been adopted in other areas.

Young people enter through their GP practice, which remains their clinical home throughout. A multidisciplinary team – including a GP clinical lead, occupational therapist, social prescriber, health coach and care coordinator – works with the young person to understand what matters to them and what support they need, rather than beginning with diagnosis alone. Support may address emotional wellbeing, education, sleep, social participation and neurodevelopmental needs, while diagnostic assessment proceeds in parallel where appropriate.

Case study: The Well Centre – integrated youth health support in the community

The Well Centre provides a youth-friendly primary care and integrated health service for young people aged approximately 11–20 years. Originally developed in Lambeth and subsequently expanded to other areas, it brings together general practice, mental health support, social prescribing, sexual health services and youth work within a single accessible setting.

The model combines adolescent-focused general practice with holistic biopsychosocial assessment and short-term wellbeing support, enabling young people to access help for physical health, emotional wellbeing, relationships, self- esteem and neurodevelopmental needs through one coordinated service. It works closely with CAMHS, schools and community organisations to provide earlier support and reduce barriers to seeking help.

Evaluation found high levels of satisfaction, with 94% of young people reporting that they would recommend the service. It also reported statistically significant improvements in wellbeing, with more than half of young people showing clinically meaningful improvement following engagement.

Case study: Centre 33 – flexible support for young people at risk of disengagement

Centre 33 is a voluntary sector organisation working across Cambridgeshire and Peterborough, providing free and confidential support for young people with emotional and practical difficulties through local hubs, community venues and education settings. It’s Someone To Talk To pilot (2019–2022) was developed to provide more flexible and sustained support for young people who were less likely to engage with conventional models of care.

The evaluation included 2,286 young people seeking mental health support during the pilot period. Following an initial assessment by an Engagement and Support Worker, young people were offered either the enhanced pilot model or the traditional service. Rather than relying on rigid eligibility criteria, the type and intensity of support were matched to need. Support ranged from text messages, telephone calls and drop-ins to booked appointments and talking therapies, with the intensity of contact increasing or reducing over time according to circumstances.

Young people receiving the enhanced model presented with greater needs and higher levels of psychological distress than those receiving the traditional service, yet showed greater average improvement in psychological health and wellbeing. They particularly valued having choice over how support was provided, maintaining contact while waiting, regular check-ins, and the absence of a rigid distinction between being ‘open’ or ‘closed’ to the service.

  • Build neighbourhood multidisciplinary teams (MDTs) as a distribution of

expertise, not responsibility. Neighbourhood MDTs should bring together primary care, mental health, paediatrics, neurodevelopmental services, education, speech and language therapy, occupational therapy, educational psychology, specialist teachers, social care and the voluntary sector, with senior specialists providing consultation, supervision and shared care. Task- sharing supported by specialist supervision is endorsed by the World Health Organization and has been shown to extend specialist expertise into neighbourhood services (Evidence maturity: strong.) (Recommendation 6)

  • Make transitions a planned, shared responsibility. Transitions should be

coordinated across organisations, with responsibility shared until transfer is complete. Where separate services remain, structured transition planning and readiness tools should be used, while services increasingly move towards

integrated 0–25 pathways where feasible. (Evidence maturity: strong for the problem; moderate for the intervention.) (Recommendation 3)

  • Make education a core partner in the pathway. Every education setting

should have a Senior Mental Health Lead, appropriate mental health training for staff, and an inclusive culture for neurodevelopmental difference. Mental Health Support Teams and specialist expertise should be organised through neighbourhood hubs, working alongside speech and language therapy and other specialist services as ‘experts at hand’. (Evidence maturity: moderate.) (Recommendation 3)

  • Reform the health–work interface towards earlier support. Primary care

should move beyond passive certification towards active support for remaining in, or returning to, work, recognising that the likelihood of return declines as absence lengthens. Workplaces should adopt principles of universal design so that environments, tools and ways of working are accessible by design. Support should begin early rather than waiting for diagnosis. (Evidence maturity: strong.) (Recommendation 3)

  • Commission high-fidelity Individual Placement and Support (IPS)

adapted for young people. IPS for young adults should be delivered to recognised fidelity standards, embedded within services used by young people and targeted particularly towards 16–25-year-olds who are NEET and experiencing mental health difficulties. Meta-analysis shows competitive employment rates of 58.3% compared with 32.4%, together with improved educational participation[923]. (Evidence maturity: strong.) (Recommendation 3)

  • Establish open-access, self-referral early support hubs for 11–25-year-

olds. Building on existing early support hubs and Young Futures Hubs, these should bring together mental health, education, employment and physical health support, with independent evaluation built in from the outset. International models, including Australia’s Headspace, provide encouraging evidence for this approach. (Evidence maturity: moderate.) (Recommendation 3)

  • Match interventions for neurodevelopmental conditions to the person’s

functional profile. Medication should be used where the goal is to improve attention, concentration or productivity[257], while psychosocial and parent- mediated interventions should be used where the goal is to improve relationships, participation and family functioning[256]. (Evidence maturity: moderate.) (Recommendation 4)

Case study: Greater Manchester IPS – employment support as part of mental health care

Individual Placement and Support (IPS) is an internationally recognised, evidence-based approach that helps people with mental health conditions find, retain and progress in paid employment as part of their recovery. Across studies, just under half of people receiving IPS move into competitive employment, while evaluations suggest savings of around £6,000 per person through reduced hospital admissions, fewer relapses and lower rates of readmission.

Since April 2025, more than 4,300 people have received IPS support across Greater Manchester, exceeding the NHS target for the city region by more than 40%. Pennine Care NHS Foundation Trust delivers IPS across all ten Greater Manchester boroughs in partnership with Pure Innovations, PossAbilities and The Growth Company. This sits alongside a complementary primary care IPS service commissioned by the Greater Manchester Combined Authority. Around 70 employment specialists work alongside mental health services, supporting people from exploring career options and developing skills through to applications, interviews and settling into work.

This example illustrates the Review’s argument that participation in work, education and community life should be treated as part of recovery, not only as a later outcome of care. It also shows how employment support and mental health treatment can be delivered together through a multidisciplinary and partnership- based model.

19.4.3 Delivering effective treatment, adapted care and parity of esteem

Good access and dependable assessment matter only if they lead to effective treatment and support. The Review is not arguing that one therapeutic approach should replace another. Rather, people should have access to a broader, better matched range of evidence-based interventions, selected according to formulation, need, preference and evidence rather than the historical availability of a single local model.

Cognitive behavioural therapy (CBT) remains an important and effective treatment for many conditions. It should, however, sit alongside other evidence-based or evidence-informed approaches, including family and systemic therapy, interpersonal psychotherapy, psychodynamic therapies, dialectical behaviour therapy (DBT),

mentalization-based treatment (MBT), parent-supported interventions, and, where appropriate, play and creative therapies. Adapted therapies should also be available for neurodivergent people. The Impact trial found CBT, short-term psychoanalytic psychotherapy and brief psychosocial intervention to have broadly equivalent effectiveness for adolescent depression[1052]. Family and systemic therapy remains first-line treatment for adolescent eating disorders[1053], while interpersonal psychotherapy[1054], DBT[1055, 1056], MBT[1057] and Adaptive Mentalization-Based Integrative Treatment (AMBIT)[1058] each have an important place.

A neurodevelopmental condition should never become a reason to exclude someone from mainstream mental health care. Services should adapt communication, pacing, environments, formats and therapeutic materials so that neurodivergent people can access evidence-based treatment on an equal basis. Adapted CBT for autistic children with anxiety has evidence of benefit[421], with meta-analyses supporting effectiveness while also noting limitations in self-reported outcomes and maintenance of benefit over time[1059, 1060].

Two further points are important. First, co-occurrence is the rule rather than the exception. Around seven in ten autistic children experience at least one co-occurring mental health condition[79, 334], yet emotional distress is still too often attributed to autism or ADHD rather than recognised and treated in its own right. Diagnostic overshadowing remains an important source of inequitable access and, at its most serious, avoidable harm[354]. Second, matching works both ways. Family therapy, although first-line for adolescent eating disorders, did not reduce repetition of self- harm in a large trial[1061], while evidence for medication targeting the core features of autism remains limited. As throughout the pathway, interventions should be matched to need rather than applied by default.

  • Close the treatment gap by commissioning named interventions,

together with measures of fidelity and outcomes. Set explicit ambitions for coverage across children’s and young people’s mental health, eating disorders, tic disorders, self-harm, substance misuse and perinatal mental health, and monitor delivery, fidelity and outcomes. (Evidence maturity: strong.) (Recommendation 4)

  • Guarantee a minimum post-diagnostic offer. Diagnosis should mark the

beginning of continuing care, including explanation, psychoeducation, shared planning, peer support and a clear onward plan. (Evidence maturity: moderate.) (Recommendation 2)

  • Protect dedicated treatment capacity. Ensure that trained staff are able to deliver the interventions for which they were trained, supported by routine

outcome measurement, supervision and protected therapy time rather than being drawn into assessment and administrative pressures. (Evidence maturity: strong.) (Recommendation 3 and 6)

  • Adapt NHS Talking Therapies for young adults and neurodivergent

people. NHS Talking Therapies should be developmentally appropriate for 16–25-year-olds and adapted for people with ADHD or autism and other educational needs, including intellectual disabilities and language disorders, recognising that young adults and students achieve lower recovery rates within current services. (Evidence maturity: strong.) (Recommendation 3)

  • Commission autism-adapted psychological therapies as a defined

model of care. Adaptations should include concrete and visual materials, incorporation of interests, structured family involvement, longer pacing and maintenance sessions. A meta-analysis of nineteen randomised trials found a large clinician-rated effect, but pooled gains were not maintained at follow-up, so maintenance sessions matter[1060]. (Evidence maturity: strong.) (Recommendation 3)

  • Extend regulated digital therapies within matched care. Commission

guided, NICE-endorsed digital CBT as an early intervention for mild-to- moderate anxiety and low mood, always supported rather than offered as a stand-alone intervention. (Evidence maturity: strong.) (Recommendation 7)

  • Strengthen medication review and safe deprescribing. Continue

medication where benefit clearly outweighs harm, and review and reduce treatment safely where it does not, particularly in relation to long-term polypharmacy in adults. (Evidence maturity: moderate.) (Recommendation 4)

19.4.4 Crisis and the highest needs

A relatively small proportion of children, young people and adults experience the greatest levels of risk, complexity and impairment. Small in proportion does not mean small in importance: each represents profound, and sometimes life- threatening, need. They require highly specialised assessment, treatment and coordinated care, often over prolonged periods and across several agencies. Strengthening prevention and earlier intervention should reduce avoidable escalation, but a group requiring highly specialised care will always remain. Earlier support and specialist care are therefore complementary, not competing, priorities.

Crisis belongs within this pathway rather than outside it. It is an urgent response that should be accessible from every point because anyone may deteriorate suddenly. It is also the point the rest of the pathway exists, wherever possible, to prevent.

Emergency departments and acute paediatric wards are increasingly managing needs they were not designed to meet, including twelve-hour waits for children in mental health crisis that have more than tripled since 2019[1062] and mental-health- related acute medical admissions among children increasing by around 65% between 2012–13 and 2021–22[760].

A crisis response organised around the principles of this pathway should be delivered as close to home as possible, use the least restrictive approach consistent with safety, and be fully integrated across physical and mental health services. It should be understood as one stage of the pathway rather than its endpoint. Evidence supports intensive community treatment and day-patient models in some settings[1063-1065], structured interventions following self-harm[796, 1066-1068], and active follow-up after crisis[1069, 1070], recognising that the period immediately after discharge carries particularly high risk[1071]. Autistic children and young people with intellectual disability remain over-represented in restrictive settings[1072].

Admission is not, in itself, safety. Inpatient care can be therapeutic, but it can also be distressing and, at times, harmful. Describing hospital admission as inherently safe does not reflect the experience of many children, young people and families. Where inpatient care becomes necessary, it often signals that earlier opportunities to provide effective support have been missed.

This is also the point at which the pathway must be most explicit about groups it has too often reached last. People with autism and severe or profound intellectual disability require highly specialised, sustained and often lifelong support, planned across transitions into adult and later-life services and beyond the point at which family carers are able to continue providing care. Their needs – including physical health, communication, and future care when parents are no longer able to provide it – should be planned for proactively and reported transparently through data on numbers, placements, duration, outcomes and costs. The consequences for families are equally profound. Lost employment, financial hardship, reliance on benefits and strain on parents, carers and siblings are all part of the true cost of unmet need.

  • Extend keyworker support for children and young people with the most

complex needs. Build on the NHS England autism keyworker model so that a named professional holds responsibility for coordination beyond the current inpatient-risk cohort. (Evidence maturity: emerging.) (Recommendation 3)

  • Protect and strengthen specialist services for those with the highest

levels of need. Ring-fence and monitor specialist capacity for severe mental illness, autism with severe or profound intellectual disability or very high support needs, profound communication needs and complex co-occurrence.

Care should be organised around sustained relationships rather than episodes of care. (Evidence maturity: moderate.) (Recommendation 4)

  • Integrate physical and mental healthcare for those with the highest

needs. Deliver and act on annual physical health reviews for people with severe mental illness and learning disability, while ensuring access to adapted psychological therapies for co-occurring anxiety and depression. People with severe mental illness have a life expectancy 15–20 years shorter than the general population, largely because of preventable physical illness[1073]. (Evidence maturity: strong.) (Recommendation 4)

  • Recognise children in the secure estate as a priority population. Around

420 children are currently in the secure estate (Chapter 17). Assessment should include CHAT on entry, Secure Stairs trauma-informed formulation, and multidisciplinary formulation rather than diagnosis alone. (Evidence maturity: moderate.) (Recommendation 4)

  • Establish national standards for children’s and young people’s crisis

and inpatient care. Every Integrated Care System should provide intensive community and home treatment, therapeutic inpatient care where necessary, structured intervention before discharge, co-produced safety plans and proactive follow-up. Standards should also include 24-hour safe havens, alternatives to admission, discharge funding that follows the young person rather than being disputed between organisations, continuity across the CAMHS–adult transition, and clear accountability where avoidable harm occurs. The Six Core Strategies should guide the reduction of seclusion and restraint[1074]. Safety Planning Intervention with follow-up has been shown to approximately halve suicidal behaviour and double engagement with services[1075]. (Evidence maturity: emerging.) (Recommendation 4)

Case study: Devon Specialist Personality Disorder Service – specialist alternatives to long hospitalisation

The Devon Specialist Personality Disorder Service illustrates how a local specialist therapeutic pathway can reduce reliance on prolonged hospital admissions and out-of-area placements for people with the most complex mental health needs. The Centre for Mental Health evaluation describes this population as people with multiple and interacting needs, including personality disorder or complex emotional needs, eating disorders, persistent physical symptoms, substance use and neurodevelopmental conditions. Their difficulties do not fit easily within single-diagnosis services or traditional service boundaries.

The service provided outreach and assessment for people at risk of, or already placed in, out-of-area services. Where appropriate, it offered intensive psychological treatment, including a year-long psychodynamic therapeutic community programme operating four days a week, followed by outpatient psychodynamic psychotherapy and family therapy where needed. For people unable to participate in intensive psychotherapy, the service instead provided consultation and support to the teams caring for them.

The evaluation found that, following the introduction of the service, both the number and duration of out-of-area placements fell. Specialist personality disorder and High Dependency Inpatient Rehabilitation placements reduced from 27 in 2012 to 13 across the period 2014–2020. Mean length of stay fell from 951 days to 406 days, while average placement costs reduced from £90,487 to £35,947 per case. At the same time, around twice as many people received specialist therapeutic treatment as before the service was established, and use of emergency services also declined.

19.5 Making it happen

19.5.1 A digitally enabled, human-centred system

Digital infrastructure is not a separate pathway. It is one of the things that makes a coordinated pathway possible. A digital front door should not mean a digital-only service. Rather, it should provide the shared infrastructure that allows people to enter through many different routes while building a single, evolving understanding of need rather than multiple records and repeated assessments.

People should have to tell their story only once, while remaining the authors of that story. A clear distinction should be drawn between the information that needs to travel with the person and the personal narrative that should travel only with their knowledge and consent. Information sharing should therefore be co-designed with careful attention to consent, capacity, privacy and the shame that can accompany sensitive mental health information.

Digital tools should support, not replace, professional judgement. Guided digital interventions are more effective than unguided ones, which achieve little sustained engagement[1076, 1077]. National guidance therefore recommends guided digital therapies only when combined with professional assessment and ongoing support[1078]. AI and structured digital tools can assist navigation, information gathering and administration, but responsibility for clinical judgement, diagnosis, safeguarding and assessment of suicide or self-harm risk must remain with appropriately qualified professionals[1079].

Digital innovation can either widen or narrow inequalities. Children and young people experiencing poverty, unstable housing, limited internet access, language barriers, disability or other forms of exclusion may become further disadvantaged if digital routes become the default[1080, 1081]. Experience with single digital points of access in England suggests that uptake may become concentrated among less deprived families unless digital exclusion is actively addressed[738].

  • Deliver national digital transformation through common standards

rather than a single national platform. National data standards, interoperability, information governance, national commissioning of digital services and secure data environments should support real-world evaluation. (Evidence maturity: moderate) (Recommendation 7)

  • Establish a national minimum dataset and treat digital exclusion as a

health inequality. Extend the Mental Health Services Data Set to include functioning, participation and outcomes across all providers, including Right to Choose and independent providers, while preserving non-digital routes into care. (Evidence maturity: moderate) (Recommendations 7 and 8)

  • Apply the same evidence standards to digital tools as to any other

intervention. Digital tools should be endorsed only where they meet the NICE Evidence Standards Framework and DTAC requirements, with MHRA regulation where appropriate. Tools making clinical or diagnostic claims should be evaluated to clinical standards. Bias and equity should be routinely audited and published, and responsibility for assessing suicide risk, self-harm and safeguarding should always remain with appropriately qualified

professionals. AI should never be the sole response to someone at risk. MindSpot provides an example of safe national digital delivery at scale[1082]. (Evidence maturity: emerging) (Recommendation 7)

19.5.2 Capacity, workforce and stewardship

The pathway is not a blueprint to be implemented identically everywhere. Most of its components already exist in different forms: Family Hubs, neighbourhood teams, integrated neurodevelopmental pathways, mental health services, Mental Health Support Teams, community crisis services, digital innovations, social prescribing, employment support and the voluntary sector. The challenge is not to create these services anew, but to connect them into a coherent system.

Four conditions will determine whether a needs-led approach succeeds.

First, capacity. Earlier access without sufficient capacity simply becomes a longer waiting list – or a different form of rationing.

Second, coordination. Integrated pathways work best where collaboration already exists and struggle where it does not[803].

Third, workforce. The challenge is not simply to increase staff numbers, but to organise expertise so that specialist knowledge reaches people where they need it most.

Fourth, learning. Many community-based approaches remain less well evaluated than specialist services and should therefore be introduced with a clear theory of change, agreed outcomes and independent evaluation from the outset.

The Royal Colleges and professional bodies should be partners in developing the standards, competencies and training needed to support this way of working. If the pathway is to shift its emphasis from activity and diagnosis towards outcomes and participation, professional standards, training, supervision and commissioning will need to evolve alongside it. People with lived experience should be involved throughout the design, implementation and evaluation of local pathways.

The Review is realistic about the gap between this vision and the current system. Some contributors, including members of the Lived Experience Panel, questioned whether change on this scale could be delivered and argued instead for more targeted improvements to waiting times and core NHS services. The Review retains the broader ambition because piecemeal solutions are unlikely to resolve a fragmented system. However, implementation should distinguish genuinely new developments from better coordination of what already exists, build earlier-support

capacity before reducing specialist provision, and phase reform realistically rather than assuming it can happen all at once.

Successful implementation also depends on clear accountability. This is not a proposal for the NHS to deliver everything itself. Health, education, social care, employment, welfare and community services all have distinct responsibilities. Each component of the pathway should therefore have a clearly identified lead organisation, an agreed funding mechanism and stable, multi-year commissioning, so that services are not withdrawn just as relationships and effective ways of working become established. Terms such as framework, care, support, specialist and pathway owner should also be used consistently across organisations.

Finally, the evidence for a pathway of this kind necessarily extends beyond randomised trials. Alongside trial evidence, the Review has drawn on service evaluations, qualitative research, grey literature and lived experience because many community and voluntary-sector approaches are widely valued but remain under- evaluated. Clinician-reported outcomes should therefore be interpreted alongside the experiences of children, young people and families themselves.

  • Develop a tiered competency and skill-mix model. Expand the low-

intensity workforce for people with mild-to-moderate need; deploy specialists increasingly through consultation and supervision; reserve direct specialist care for the most complex and highest-risk cases; and ensure experienced clinicians are available at the point of entry. Task-sharing with specialist supervision is endorsed by the World Health Organization, and low-intensity practitioners achieve reliable improvement for around 47% of children receiving treatment (Chapter 13). (Evidence maturity: strong.) (Recommendation 6)

  • Commission the voluntary, community and social enterprise (VCSE)

sector as core partners through stable funding. Multi-year commissioning, shared governance arrangements, common outcome measures and support for organisational sustainability should replace short-term funding cycles. The VCSE sector provides trusted and culturally responsive support but remains organisationally fragile and under-evaluated (Chapter 15). (Evidence maturity: moderate.) (Recommendation 3)

  • Use shared-care and competency frameworks to distribute specialist

expertise safely. Nationally consistent, properly funded shared-care arrangements should enable ADHD and other long-term conditions to be managed closer to home. (Evidence maturity: moderate.) (Recommendation 5)

  • Develop secure cross-sector data linkage. Use de-identified linkage across

health, education, employment and justice within secure environments, while improving fit-note and occupational health data. (Evidence maturity: moderate.) (Recommendation 7)

  • Establish a permanent national surveillance and observatory function.

Rather than creating a new organisation, coordinate the work of existing statistical, public health and academic bodies. Place the Adult Psychiatric Morbidity Survey and Mental Health of Children and Young People surveys on a regular cycle, incorporating measures of functioning, participation and neurodevelopment. (Evidence maturity: moderate.) (Recommendation 8)

  • Build independent evaluation into every major reform. Every programme

should begin with a clear theory of change, agreed outcomes and independent evaluation. Priorities include research on co-occurrence, service models, the drivers of impairment and NEET, digital technologies and AI, and participation in relation to ADHD and autism. Mental health research has historically been underfunded[1083]. (Evidence maturity: moderate.) (Recommendation 8)

19.5.3 Equity and reaching those most likely to be missed

A pathway organised around need will only be more equitable than the system it replaces if it reaches the people current services most often fail to reach. Systems naturally organise themselves around those who already use them. Improvements that make access easier for those already within the system may leave untouched those who never reach it. Around 4% of adolescents and 5% of adults present with multiple interacting difficulties and have the poorest outcomes across every domain measured, including wellbeing, self-harm, peer relationships, physical health, adversity, SEND, educational disengagement, employment, homelessness, suicide attempts and unmet need (Chapter 9, Section 9.2.2). This group should be a clear priority for support.

Designing for equity therefore means designing around those who remain least visible, following the principle of proportionate universalism: support that is universally available but delivered with an intensity proportionate to need. Children living in poverty, care-experienced children, people from minoritised communities, people with intellectual disability and those with complex communication needs should become more visible through integration, not less. Some of the services most likely to reach underserved groups – particularly community and voluntary sector provision – remain less well evaluated than specialist services. They should

therefore expand with evaluation built in, rather than being held back until the evidence base is complete.

Support should be person-centred and recognise that recovery and participation depend on much more than clinical treatment alone. Education, employment, relationships, housing, debt, gambling-related harms and other social determinants all influence mental health and should be considered as part of care. Services should also be culturally responsive and trauma-informed, recognising the influence of people’s backgrounds, experiences and identities, while addressing inequalities in access, experience and outcomes across different communities.

Co-design with people with lived experience, families, carers and communities is essential if services are to build trust, reach underserved groups and remain responsive to changing needs.

  • Commission proportionate universalism explicitly. Local plans should

describe how universally available support will be intensified for those least likely to reach services, and should report access, experience and outcomes separately for these groups rather than only in aggregate. (Evidence maturity: moderate.) (Recommendation 3)

  • Recognise the multiple-difficulties group as a priority population. The

4% of adolescents and 5% of adults with multiple interacting difficulties experience the poorest outcomes across every domain measured and are poorly served by single-condition pathways. They should be recognised explicitly in local planning, with coordinated care and a named professional responsible for continuity. (Evidence maturity: moderate.) (Recommendation 3)

  • Expand community and voluntary sector provision alongside

independent evaluation. Services most likely to reach underserved groups should grow with a clear theory of change and built-in evaluation, rather than waiting for a perfect evidence base before they are commissioned. (Evidence maturity: moderate.) (Recommendations 3 and 8)

19.6 Chapter synthesis

This chapter has described what a pathway would look like if it were organised around need, functioning and participation rather than diagnosis and organisational boundaries.

The pathway has four connected stages: access through many routes into one coordinated system; assessment and formulation that build a shared understanding alongside diagnosis; support matched to changing need and delivered as close as possible to everyday life for individuals and those who care for them; and planned step-down that preserves continuity rather than treating discharge as an endpoint. Three principles run throughout: understanding travels with the person; specialist expertise strengthens earlier and neighbourhood support; and crisis remains part of the same pathway rather than standing outside it.

The Review has also been explicit about the conditions for success. Needs-led care depends on sufficient capacity; integration depends on sustained collaboration; digital infrastructure must support, not replace, professional judgement; community services should grow alongside independent evaluation; and earlier support must strengthen rather than replace specialist and crisis care.

Ultimately, neither diagnosis nor access alone is the goal. The purpose of the pathway is to recognise need early, provide effective support and enable children, young people and adults to participate as fully as possible in family life, education, relationships, work and their communities.

Chapter 20. Recommendations for government and local systems

Three principles run through the recommendations that follow. They apply to mental health, ADHD and autism together, and across the life course.

Tackle the drivers of mental distress and ill health

The evidence reviewed throughout this report shows that mental health outcomes are shaped by a complex interaction of factors, including poverty, insecure housing, school exclusion, unemployment, family adversity, discrimination, social isolation and wider social and environmental influences. Among children and young people, changes in digital environments and educational pressure show the closest correspondence with the timing and pattern of rising distress (Chapter 10). Improving mental health therefore requires action across government and across society, recognising that many of the strongest influences on mental health lie outside healthcare itself. Education, employment, housing, welfare, family support, communities and public health should be understood as part of the mental health system, not separate from it. The aim is to reduce avoidable distress, strengthen participation and resilience, and prevent difficulties from escalating into crisis.

Prevention has a different meaning for neurodevelopmental conditions. ADHD and autism are not conditions to be prevented. Here, the equivalent ambition is preventing traits and symptoms crossing the threshold to impairment, earlier recognition and earlier support, so that people do not accumulate avoidable mental health difficulties, educational disruption and reduced participation while they wait (Chapters 7 and 8).

Move from a diagnosis-dependent system to one led by need alongside diagnosis

Timely, accurate diagnosis remains essential, but it should sit within a broader system organised around need. Too often, people must wait, be assessed and receive a diagnosis before any support begins. By need we mean the impact of a person’s difficulties on their functioning and participation – including distress that

may not yet have reduced outward participation but is being sustained at considerable personal cost – taking account of severity, complexity and the risk of deterioration or harm from delay.

A system led by need alongside diagnosis reverses that sequence. It recognises need, seeks to understand it, begins support early, uses diagnosis to inform treatment and specialist intervention, and adapts as needs change. This is not a way around diagnosis. Understanding what lies behind a person’s difficulties often requires careful clinical assessment, and support should be informed by that understanding rather than substituted for it. Children and young people should be a particular priority within a life-course approach that applies the same principles across development, changing circumstances and major transitions (Chapters 18 and 19).

Support and diagnosis should proceed in parallel, not in strict sequence. The case is strongest in neurodevelopmental pathways, where waits are longest and diagnosis most tightly controls access to support. By the end of 2025, open referrals for ADHD assessment exceeded 560,000, more than seventeen times the number recorded in 2019, while autistic people frequently accumulate avoidable difficulties during prolonged waits (Chapters 7 and 8). Support offered while people wait should be a bridge to timely assessment, not a substitute for it. Reducing waiting times should remain an immediate priority. Waiting lists should be actively reviewed so that each person receives a response proportionate to their needs, and no one should be removed from a pathway without appropriate assessment, communication and discussion.

A shared description of need should open the way to appropriate support; it must never become another hurdle or a mechanism for rationing care. Services should describe need consistently so that people are understood once and that understanding travels with them, rather than requiring repeated assessment at every point of contact. Being understood once does not mean combining every assessment. Where separate assessments for ADHD, autism or mental health are needed, each should be undertaken properly within a coordinated pathway and with shared clinical oversight.

No single organisation can deliver this alone. The drivers of distress, the understanding of need and the outcomes that matter span health, education, social care, employment, welfare, housing, and the voluntary and independent sectors. The NHS remains central to clinical care, but its role increasingly includes convening and coordinating support with partners. The economic case must therefore be considered across government rather than within health alone. Earlier support may require investment upfront, while the benefits – increased participation, improved

employment and fewer avoidable crises – accrue across different public services and over longer timescales. Those benefits should be measured rather than assumed.

Design services with people and measure what matters

Services should be designed with the people they are intended to help, alongside clinicians, professionals and communities. Lived experience should be embedded in the policies, services and communications that follow from this Review, not consulted once and then set aside.

Success should be judged by whether people are better able to participate in relationships, family life, education, work and their communities. For people with ADHD and autism this includes whether reasonable adjustments are available and effective (Chapter 16). Counting assessments, diagnoses or movement through waiting lists measures activity rather than outcomes.

These recommendations should be read as a connected programme of reform, not as separate proposals.

Recommendation 1 Tackle the drivers of mental distress and ill health through an ambitious approach to prevention, early intervention and improving public mental health and wellbeing.

Tackle the key drivers and social determinants of mental health and wellbeing as determined by the evidence, by:

  • Protecting opportunities for rest, recovery and healthy development:

increase access to sport and recreation, play and green spaces, creative arts, promote healthy sleep and target action to reduce screen time during school hours or late at night.

  • Reducing harmful evaluation and comparison: work with regulators and

industry to create safer online environments, including holding platforms accountable for harmful design and content; improve digital literacy; and work with DfE to provide positive experiences in education settings, inclusive environments in universities, deliver evidence-based anti-bullying programmes in schools, and support students to manage academic and exam pressures. Where school-based interventions aim to prevent enduring mental health conditions, prioritise selective and indicated approaches for children and young people at increased risk, and evaluate programmes for potential harms as well as benefits.

  • Strengthening belonging and connection: invest in youth, community and

voluntary organisations and school enrichment activities that promote inclusion and improved mental health and wellbeing.

  • Supporting people to manage uncertainty and increase stability:

appropriate housing, stable employment and financial security, alongside evidence-based population-level approaches that reduce suicide risk, harmful substance use and other preventable causes of mental distress and ill health. Increase protective factors that families and immediate support networks can provide by making evidence-based parenting and family support universally available, with intensity matched to need.

Establish a national public mental health and prevention programme led by DHSC, with clear cross-government accountabilities, funding for delivery of shared outcomes and monitoring impact.

Deliver through local systems and devolved powers, embedding prevention in local planning and shared accountability across Integrated Care Boards (ICBs), mayoral authorities and local partnerships.

Develop with people with lived experience and other partners, including the voluntary sector. Provide longer-term funding for local voluntary and community organisations as key partners to the NHS and local government, and work with communities to build on their strengths and local assets that protect mental health.

Recommendation 2 Move from a ‘diagnosis-dependent’ system to a ‘needs-led plus diagnosis’ system by developing a digitally and AI-enabled shared needs assessment and prioritisation framework for people with mental health and neurodevelopmental needs.

The key components of the needs led plus diagnosis system include:

‘No wrong front door’

  • There should be ‘no wrong front door’ to the needs-led system, meaning

people can enter through any trusted route at a local or national level, including schools, primary care, community services, mental health services, social care, youth provision, employers, family hubs and digital platforms – as well as via self-referral where appropriate.

Shared needs assessment behind the ‘no wrong front door’

  • Behind the ‘no wrong front door’, there should be a shared needs

assessment: an integrated approach to triage and assessment, under senior clinical oversight, that considers functional need, urgency and risk, persistence, clinical complexity and co-occurrence, and context.

  • Adopt a consistent shared way of assessing and describing need across

health, social care, education, employment and other support services, building on the principles of the International Classification of Functioning, Disability and Health (ICF).

  • Building on existing tools that have already been developed and piloted, roll

out validated and clinically overseen digital and AI tools to support information-gathering, navigation, signposting and triage, with clear routes to immediate human escalation (see Recommendation 7). Pilots should begin in 2027, co-designed with people with lived experience, clinicians and other professionals, to build the evidence base for effective implementation, aiming for full rollout across England conditional on evaluation.

  • Ensure data follows the person so they do not need to undergo unnecessary reassessments. (see Recommendation 7)

Prioritisation based on different people’s needs

  • Prioritise individuals with severe functional impairment, multiple difficulties or

escalation triggers such as self-harm, safeguarding concerns, imminent school exclusion, or being at risk of unemployment for rapid, comprehensive specialist assessment.

  • Those with mild to moderate impairment should receive support whilst

awaiting diagnostic assessment including evidence-based psychoeducation; practical, parenting and peer support; monitoring and/or appropriate school or workplace adjustments.

  • Ensure people whose needs exceed what can be managed in primary care,

but who do not meet thresholds for specialist services or face prolonged waits for them, receive timely support matched to need.

  • Where there is active anxiety, depression, eating disorder, trauma related

distress or psychosis, appropriate mental health assessment and treatment should not be delayed until a neurodevelopmental assessment is complete.

  • Retain rapid access to local crisis support for people who need it, in line with specialist and emergency care pathways and standards.
  • Guarantee a post-diagnostic offer tailored to the person’s needs where

diagnosis marks the beginning of continuing care: explanation, psychoeducation, shared planning and a clear onward offer of support, treatment and intervention if needed.

Address the long waits people are currently experiencing for ADHD and autism assessments in particular

  • Address the large numbers of people on long waiting lists for an ADHD or

autism assessment by conducting a clinically-led review of these lists to establish, and identify, for each person, the primary presenting concern, to prioritise the most relevant existing pathway and organise assessments for other potential co-existing conditions around this core assessment and to offer evidence-based support while waiting. People must not just be removed from waiting lists without a proper assessment and need appropriate communication, discussion to ensure they receive a response as early as possible that responds to their needs and the right to a clearly defined and transparent appeals process.

Recommendation 3 Commission and deliver integrated mental health and neurodevelopmental assessment and support that is age-appropriate and responds to the needs of local communities.

Actions include: Easy to access support as close to home as possible

  • Invest in integrated, community-based support in neighbourhood health

services for people with mental health and neurodevelopmental conditions that are easy to access and delivered as close to home as possible.

  • Commission voluntary, community and social enterprise (VCSE)

organisations as core partners in the delivery of integrated mental health and neurodevelopmental support, supported by stable and sustainable funding.

  • Deliver full national coverage of the Early Support Hubs model, focused on

integrated youth participation and support, Best Start Family Hubs, Mental Health Support Teams in schools and colleges, sequencing the remaining expansion so that special schools, alternative provision and post-16 settings are covered first and community-based Mental Health Centres for adults; and strengthen social prescribing and community connection.

Multi-disciplinary teams organise services around people, rather than the other way round

  • Follow the ‘Expert at Hand’ principle and organise services around people

through multidisciplinary teams. Deploy specialist expertise through consultation, supervision and shared care arrangements, enabling timely triage, decision-making and escalation while supporting a wider range of practitioners to deliver care matched to need, including interventions delivered in schools and homes. Progress should also be reviewed over time and stepped up or down according to the individual’s needs (see Recommendation 6).

  • Roll out keyworkers for people with the most complex co-occurring needs,

prioritising children who would benefit from having access to a trusted adult outside the family and recognising the importance of youth work within this. This should be a clearly identified personal contact who can support them with navigating information, appointments and services as required.

Joined up support from childhood through to adulthood to increase participation

  • Make participation in education and work a core outcome of care and support.
  • Deliver joined-up support with clear transfers of care for 0 to 25-year-olds,

including coordinated support from pregnancy through preschool years for children and their families, across primary and community health care, mental health and neurodevelopmental services (including CAMHS), paediatrics, health visiting, perinatal, early years and parenting support, speech and language and occupational therapy, education, social care, employment support and the voluntary sector.

  • Provide structured developmental checks at universal contact points and match neurodevelopmental interventions to functional profile.
  • Make NHS Talking Therapies developmentally appropriate for 16–25s and adapted for ADHD and autism.
  • Scale evidence-based interventions that integrate health and employment

support, such as adapting Individual Placement and Support (IPS) for young people and strengthening workplace reasonable adjustments and access to the Access to Work scheme. Shift from simply certifying absences from work towards actively supporting participation, including fuller use of ‘may be fit for

work’ advice and phased returns, so people are helped to stay in, or return to, employment.

Improve fairness of access and outcomes for all

  • Prioritise ensuring parity of access to integrated mental health and

neurodevelopmental services for groups currently least well served including children and young people who are not in education, employment or training (NEET), care-experienced people, people with learning disabilities, ethnic minority communities, people facing multiple disadvantage, and people in contact with the justice system.

  • Expand access to tailored evidence-based interventions, including

adaptations to ensure therapies and other interventions are effective and accessible for children and young people – and people with neurodevelopmental conditions.

Recommendation 4 Close treatment gaps while protecting, strengthening and investing in specialist services to improve access and outcomes for those with the most severe, complex and enduring needs, including specialist support delivered in the community, crisis and inpatient services.

Actions to include:

  • Improve care for people with the most severe, enduring, complex or lifelong

mental health and neurodevelopmental needs, including people with profound intellectual disability, the highest levels of impairment, eating disorders, other severe mental illnesses, complex co-occurring conditions and lifelong support needs.

  • Close treatment gaps by expanding access to NICE-recommended and

evidence-based interventions and therapies, ensuring that people receiving specialist mental health care can access treatments recommended for their condition.

  • Reduce avoidable mental health-related A&E attendances, paediatric ward

admissions, and the over-representation of people with autism and/or learning disability among specialist inpatients by providing timely local crisis support, meeting national urgent and emergency mental health standards and liaison psychiatry standards, providing appropriate specialist care and evidence- based alternatives to admission and detention, and reducing avoidable escalation and out-of-area placements.

  • Focus specialist capacity where it adds greatest value, including in cases of

diagnostic uncertainty, clinical complexity, complex co-occurring conditions, significant risk and medication initiation, and for population groups with distinct specialist needs, including people receiving care from secondary mental health services and those requiring specialist liaison psychiatry input.

  • Strengthen and standardise timely medication reviews and safe deprescribing practices.
  • Outcomes for people with the highest support needs should include a focus on improving social inclusion and quality of life, not only service metrics.

Recommendation 5 Overhaul the commissioning, funding, oversight and regulation of mental health and neurodevelopmental services, including making all diagnostic assessments for ADHD and autism CQC-regulated activity and banning direct-to- public advertising of NHS-funded ADHD and autism assessments.

Actions to include:

  • Develop a new commissioning framework, funding mechanisms (including

pooled funding), and guidance to support local systems and their partners to move away from commissioning separate ADHD, autism and mental health assessment and support services and towards outcome-based commissioning of integrated support structured around need and functioning.

  • Integration means services working together to provide coordinated support. It

does not mean having a single assessment for ADHD, autism and mental health. If more than one assessment is needed, they should be carried out separately but in a planned sequence, without unnecessary delays and with shared clinical oversight.

  • Ensure consistent quality standards across NHS, independent, voluntary and digital providers, and stronger national oversight and accountability.
  • Bring all ADHD and autism assessment providers within a common regulatory

framework, making these diagnostic assessments CQC-regulated activity with consistent quality standards, data reporting and oversight. Standards should require assessment of co-occurring conditions and functioning, clear guidance on when remote and in-person assessment are appropriate, and consistent training standards for assessors.

  • Shift from episodic specialist care for ADHD to long-term management in primary care where required, supported by funded shared-care arrangements

held to a consistent national standard so people are not left stranded between services or reliant on private prescriptions once diagnosed.

  • Require all providers, including those operating under Right to Choose, to

submit consistent activity, quality and outcomes data as a condition of funding and accreditation, including separate reporting of NHS-funded and privately funded referrals, waiting times and outcomes.

  • Ban direct-to-public advertising of NHS-funded ADHD and autism

assessments, ensuring access is through commissioned pathways based on clinical need rather than provider marketing.

Recommendation 6 Build capacity and capability across the multidisciplinary workforce to deliver more integrated mental health and neurodevelopmental support, backed by investment in training, supervision and career development.

Actions to include:

  • Co-produce national and local multidisciplinary workforce strategies with

Royal Colleges, professional bodies, practitioners and people with lived experience.

  • Workforce planning should extend beyond clinical professions to include

education, youth work, welfare and employment support, social care, family support, peer support and the voluntary sector.

  • Build neighbourhood multidisciplinary teams (MDTs) as a distribution of

capability, not of responsibility, supported by a nationally agreed tiered competency model. Neighbourhood MDTs for children and young people, for example, should span primary care, mental health, paediatric neurodevelopmental services, education, speech and language therapy, occupational therapy, educational psychology, specialist teachers, social care and the voluntary sector, with senior specialists providing consultation, supervision and shared care.

  • Invest in a tiered workforce model that builds capability across the whole

system, ensuring all staff have core competence in mental health and neurodevelopmental needs, with access to increasing levels of specialist expertise as complexity rises. This should be supported by training, supervision and continuing professional development, including use of the nationally agreed shared needs assessment, to enable psychologically informed, evidence-based support that improves prevention, participation, recovery and inclusion.

  • A nationally agreed competency framework should define the knowledge and

skills required at each level of the system, enabling mental health and neurodevelopmental expertise to be distributed across universal, targeted and specialist services.

  • Recommendations 2 and 3 in particular depend on this workforce redesign to

deliver the ‘Expert at Hand’ principle and more integrated support for people with mental health and neurodevelopmental needs in practice.

Recommendation 7: Take an ambitious approach to scaling up safe digital and AI- enabled tools within a digitally enabled and integrated care pathway, while sharing and using data to improve quality of care, efficiency and productivity.

Build an integrated digital infrastructure platform for orchestration of care pathways to allow currently siloed digital tools to be sequenced, selected and matched to individual need. It should use common standards for implementation and collect consistent data and outcomes, extracting an agreed minimum dataset written to a Secure Data Environment, to support continuous learning, improvement and outcomes-based commissioning.

  • Use AI to support delivery of the needs assessment and prioritisation framework (see Recommendation 2).
  • Extend regulated digital therapies within stratified and matched care. DHSC

and NHS England should ensure this is done safely and securely within a robust national assurance and evaluation process that includes strong safeguards and clear standards for safety, effectiveness, equity, consent and privacy.

  • Use digital platforms, including the NHS App, to support people while they are

waiting for assessment or treatment by providing clear, accessible updates on referral status, expected waiting times, available support options and routes to report if their circumstances or needs change.

  • Build interoperable data infrastructure so information can be shared safely

across health, education, employment, social care and justice, with strong privacy and consent protections and particular safeguards for children and young people. Data should follow the person rather than staying locked inside organisations, supported by the single patient record.

  • Develop an ‘ADHD Passport’ as a unified single patient held record within the NHS App, giving people a portable record of their diagnosis, care plans and

medication history. This should improve continuity of care when people move between services, providers or different stages of life.

  • Preserve in-person and primary care routes and extend human care rather

than replacing it, making sure the benefits of digital provision are realised without deepening digital exclusion.

Recommendation 8: Improve data collection and evaluation for mental disorders, ADHD and autism, including through regular population prevalence studies, and update NICE guidelines to strengthen the evidence-base for policy development and implementation.

Actions to include:

  • Build a continuous learning and improvement system that measures what

matters, scales what works, tracks the impact on inequalities, and stops ineffective models and interventions.

  • Invest in more regular national population surveys that measure population

prevalence, including the Mental Health of Children and Young People (MHCYP) survey and the Adult Psychiatric Morbidity Survey (APMS); and improve regular data collection and monitoring at a local and national level for administrative prevalence, need, access, functioning, participation and outcomes.

  • Update NICE guidance to reflect the latest evidence-based interventions,

especially for ADHD and autism, and develop new guidance on responding to co-occurrence and complexity.

  • DHSC and NHS England should improve the national data infrastructure for

improved data collection and sharing, including interoperability standards, a minimum dataset, information governance, assurance and outcome measurement. For example, NHS England, the Department for Education and local government should develop a consistent minimum dataset for people on neurodevelopmental pathways that records intellectual disability, functional communication, and adaptive functioning. This should be incorporated into national health, education and social care data collection, with appropriate consent, governance and safeguards, so that autistic people with lifelong intensive support needs can be identified, counted and planned for more reliably.

  • Report outcomes broken down by ethnicity, deprivation and other characteristics, so inequities are visible and can be acted on. Publish

aggregate outcomes transparently for public accountability while protecting individual privacy.

  • Scale innovation through ‘test, learn and grow’ approaches, with rapid

evaluation and shared learning across local systems to scale innovation at pace.

Recommendation 9 Commission a national public information and education campaign to help people better understand, support and accommodate neurodevelopmental conditions and manage their mental health, supported by a trusted digital platform providing evidence-based information, tools and support without requiring a diagnosis.

Actions to include:

  • Go beyond raising awareness to build genuine understanding and drive

action, carrying a clear message: people with common mental health conditions can recover and, with the right support and adjustments, many people with ADHD and autism can live full lives, whilst recognising that needs vary widely and some people require substantial lifelong support.

  • Use evidence-based communications to improve mental health literacy,

support people to keep well, reduce stigma and misconceptions, and support earlier access to appropriate support and adjustments without the need for a diagnosis.

  • This campaign should address the drivers of poor mental health and

increasing distress, with evidence-based advice and guidance to help people support themselves and others more effectively, recognising the current gap in trusted advice and guidance.

  • Co-produce the campaign with people with lived experience, charities,

campaigning organisations and professional bodies to ensure it is credible, evidence-based, appropriately contextualised and reaches different audiences effectively.

  • Create an open-access digital platform that provides trusted information, tools

and support without requiring a diagnosis, which could build on Every Mind Matters.

  • Maintain the platform through national quality assurance, regular content updates and integration with national and local services and pathways,

ensuring people can access consistent, evidence-based advice, information and support wherever they seek help.

  • Evaluate the campaign against mental health and wellbeing outcomes as well

as awareness and knowledge, including impacts on help-seeking, service demand, waiting times, access to support and inequalities, and adapt delivery in response to findings.

Recommendation 10 Publish a Cross-Government Mental Health Strategy this year setting out how this report’s recommendations will be implemented through immediate actions in the next 12 to 24 months and a 10-year reform plan.

Action to include:

  • Identify lead departments for actions, establish shared national and local

accountability for improving mental health and wellbeing across government and put prevention and improving participation at the centre of government decision-making.

  • Agree a short set of shared cross-government outcomes, while leaving local

systems room to set their own priorities. These could include better school attendance; fewer young people not in education, employment or training; greater participation in education, work and community life; lower premature mortality among people with severe mental illness; and fewer avoidable inequalities for people with neurodevelopmental conditions.

  • Commit to actions that can be implemented within the next 12 to 24 months and those that require longer-term reform and building capacity for change.
  • Connect with related reforms, including the Department for Education’s 2026

Schools White Paper, Every Child Achieving and Thriving (covering SEND reform, pupil engagement, enrichment and home-school partnerships); the DCMS National Youth Strategy, Youth Matters, and its work on loneliness; and Alan Milburn’s review into Young People and Work.

  • Set out the investment and workforce required for addressing the number of

people on long waiting lists for assessment; delivering multidisciplinary assessments and support matched to need across community-based and secondary care; and balancing priorities across prevention, early intervention, diagnosis, treatment and specialist care – including longer-term, sustainable funding for local voluntary and community organisations.

  • Make the financial case across government, not DHSC alone, given

investment in timely, evidence-based support, earlier intervention and prevention has the potential to deliver returns across different budgets and longer timescales, which should be measured through transparent evaluation of social and economic returns.

Annexes

Glossary of terms

Working Definitions

Administrative prevalence – the proportion of people recorded as having a condition within a health, education or other administrative system; reflects recognition, access, diagnosis, recording practice and eligibility as well as underlying need, and is not equivalent to population prevalence.

Assessment – the process of examining domains of health to gain insight on a patient’s condition or functioning. The term is used in two different ways throughout this Review. A diagnostic assessment seeks to determine whether an individual meets criteria for a particular condition. A needs assessment (see Chapter 18) is a structured description of a person’s current functioning, participation, risk, complexity and environmental circumstances, used to understand the nature and urgency of support required, regardless of whether a diagnosis is present.

Clinical pathway / care pathway – standardised steps describing the course of action taken to practically apply guidelines for clinical problems, procedures and treatments.

Common mental health conditions – the grouping of depression and anxiety disorders (e.g. generalised anxiety disorder (GAD), obsessive compulsive disorder (OCD), phobias, panic disorders) that are characterised by patterns of thought, feeling or behaviour that cause significant distress or impairment; they can arise at any stage of life, may come and go, and often respond well to treatment and support.

Demand – the expression of need through requests for help, referrals, assessments or service use; reflects awareness, help-seeking, accessibility and system design as well as underlying need.

Diagnosis – the process of identifying the most likely explanation for a person’s difficulties according to internationally recognised criteria for diagnostic categorisation; a basis for communication, treatment and research.

Disability – under the Equality Act 2010, a physical or mental impairment with a substantial and long-term adverse effect on the ability to carry out normal day-to-day activities, including where a person does not identify as disabled.

Distress – the phenomenological experience of emotional suffering and the extent to which symptoms or circumstances are felt as painful, overwhelming or hard to bear;

a normal part of human life, which may accompany but is not synonymous with disorder.

Functioning – describes how a person carries out everyday activities and roles (learning, communicating, self-care, emotional regulation, relationships, work, parenting) and reflects the interaction between the person and their environment, so that two people with similar symptoms or impairments may function very differently.

Impairment – limitations associated with a health condition or developmental difference that affect functioning; understood in interaction with the environment rather than as a fixed property of the individual, and which may or may not result in a disability (see below).

Incidence – the number of new cases arising in a defined population over a specified period.

Medicalisation / Professionalisation – the process by which experiences once understood in social, educational or personal terms come to be interpreted through medical or other professional clinical frameworks; used descriptively, and recognised as having both beneficial and unintended consequences depending on context.

Mental disorder – a clinically significant, persistent pattern of symptoms that gives rise to significant levels of distress and/or impairment that meets recognised diagnostic criteria.

Misdiagnosis – when someone has a condition, or several conditions, but is given the wrong diagnosis.

Need – the gap between an individual’s status and the support required to improve health, functioning or participation at a particular time; shaped by symptoms and diagnosis but also by functioning, context, risk, co-occurring conditions, goals and available support; dynamic rather than fixed.

Neurodevelopmental conditions – differences in how the brain and nervous system develop that are present from early life, though not always apparent immediately, and shape enduring patterns of cognition, attention, communication, behaviour and learning. They exist through complex genetic, biological and environmental factors. They include ADHD and autism, alongside developmental language disorder, intellectual disability and tic disorders.

Neurodiversity – the concept that variation in brain function, cognition and behaviour (including variation associated with conditions such as autism and ADHD) is part of

natural human diversity rather than pathology and disorder; neurodivergent and neurotypical are the associated terms.

Overdiagnosis – when someone is given a diagnosis for a medical condition they do not have to explain the difficulties and impairment they are experiencing.

Participation – a person’s involvement in education, employment, family life, relationships and community. This is the outcome that functioning serves, and the one that people with lived experience most often said mattered to them, thus what support should seek to improve.

Population (epidemiological) prevalence – the proportion of the population who are estimated to have a condition, or to be above a pre-defined level of symptoms at a given time, whether or not they have sought help or been diagnosed; estimated principally through surveys.

Self-identification / self-diagnosis – the assignment of a medical condition to oneself without a formal diagnosis administered by a professional.

Support – the full range of interventions intended to improve health, functioning, acquisition of skills and participation, extending well beyond healthcare to include educational adjustments, family and community support, workplace accommodations and practical assistance.

Symptoms – subjective experiences which may be accompanied by observable manifestations (signs), such as anxiety, low mood, or inattention. Many vary dimensionally across the population. Their presence alone does not establish a disorder; clinical significance generally depends on their pattern, severity, persistence, context, associated distress or functional impairment, and relevant diagnostic exclusions.

Treatment – clinical interventions intended to improve health outcomes through established therapeutic approaches; one component of support, not a synonym for it.

Under-recognition / underdiagnosis – these terms are closely linked; the former is the precursor to the latter. Under-recognition refers to a condition not being recognised as a possible explanation for a person’s difficulties. Underdiagnosis occurs when someone has a condition but does not receive a diagnosis when they should.

Key conditions

Attention deficit hyperactivity disorder (ADHD) – a neurodevelopmental condition characterised by persistent patterns of inattention, hyperactivity and/or impulsivity that are associated with clinically significant functional impairment.

Autism – a neurodevelopmental condition characterised by differences in social communication and interaction, together with restricted and repetitive behaviours, interests or activities, and sensory differences. Diagnosis requires these characteristics to be present from the early developmental period and to cause clinically significant impairment.

Bipolar disorder – a mental health condition affecting mood, often including manic and depressive episodes.

Conduct disorder – a disorder characterised by severe and frequent behavioural issues that can violate social norms and the rights of others such as through aggression, angry outbursts and violence.

Hyperkinetic disorder – a disorder characterised by difficulty staying focused, controlling behaviour, and managing excessive or impulsive activity that typically has its onset in childhood. Some classifications (e.g. ICD-11) have since merged this term with ADHD (attention deficit hyperactivity disorder) definitions.

Learning disability – a significantly reduced ability to understand new or complex information, to learn new skills (impaired intelligence), with a reduced ability to cope independently (impaired social functioning), which started before adulthood (Department of Health and Social Care, 2001).

Oppositional defiant disorder – a disorder characterised by defiant, hostile and disobedient behaviour that exceeds the expected behaviour of a person of their age or developmental window, to the point of impacting everyday interactions with authority figures.

Personality disorder – a disorder characterised by experiencing thoughts, behaviours and feelings that differ from cultural norms to the point of causing distress and impairing functioning. For example, borderline personality disorder (BPD) a personality disorder that involves symptoms like having intense but unstable relationships, affective dysregulation, impulsivity and patterns of disturbed thinking.

Psychosis – a condition where people lose some contact with reality. May involve hallucinations, delusions and disordered thinking and speaking.

Post-Traumatic Stress Disorder (PTSD) – a disorder arising from experience such as a distressing and traumatic event, that can be characterised by experiencing symptoms like flashbacks, intrusive thoughts or hypervigilance related to the traumatising event.

Data sources and screening tools

Adult ADHD Self-Report Survey (ASRS) – a tool for screening and monitoring severity of ADHD symptoms over time.

APMS – Adult Psychiatric Morbidity Survey, a series of surveys monitoring mental health over time. Measures population prevalence of a range of mental health conditions, including common mental health conditions such as depression and anxiety, posttraumatic stress disorder, psychotic disorder, bipolar disorder, and conditions like autism, ADHD and other health-related behaviours such as alcohol and drug dependence.

Autism Diagnostic Observation Schedule (ADOS) – an assessment tool used by clinicians to help diagnose Autism Spectrum Disorder

Autism Spectrum Quotient (AQ) – a 50-item self-report questionnaire used to measure the presence of autistic traits in adults.

Born in Bradford – a cohort study sampling expectant mothers and their children in Bradford, recruited between 2007 and 2010, to assess their health, wellbeing and educational development over a 20-year period.

Clinical Interview Schedule – Revised (CIS-R) – a structured interview and questionnaire used in population surveys such as APMS to assess common mental health conditions, including depression and anxiety disorders.

Children’s Global Assessment Scale (CGAS) is a clinician-rated measure of a child or young person’s overall psychological, social and educational functioning.

CPRD (Clinical Practice Research Datalink) – a government research service that collects and provides UK wide healthcare data from a network of GP practices and healthcare organisations to support public health research.

Development and Well-Being Assessment (DAWBA) – a package of interviews, questionnaires and rating techniques designed to generate ICD-10 and DSM-IV or DSM-5 psychiatric diagnoses on 5-16 year olds.

General Health Questionnaire (GHQ-12) – a 12-item screening instrument to detect psychiatric disorders in community settings and non-psychiatric clinical settings.

Health Survey for England – monitors trends in the nation’s health and care, providing information about adults aged 16 and over, and children aged 0 to 15, living in private households in England.

MHCYP (Mental Health of Children and Young People) – a cross-sectional survey exploring the mental health of young people in Great Britain. The survey was first run in 1991 by the Office of National Statistics for the Department of Health and Social Care, and most recently conducted between 2017 and 2023.

OxWell – large-scale annual student survey designed to measure the wellbeing of children and young people aged 9-18 years old.

Strengths and Difficulties Questionnaire (SDQ) – a brief behavioural and emotional screening tool used to evaluate the psychological wellbeing of children and adolescents aged 2 to 17.

Understanding Society – a longitudinal study of UK households, started in 2009, that aims to track families across generations to document their thoughts and attitudes regarding their family, work and/or school, wellbeing, health, financial situation and social and political perspectives.

Other

Administrative data – data collected from the public’s interaction with public services like the NHS health systems, school systems, HMRC tax and income data as well as population records.

Adverse childhood experiences (ACE) – events or experiences occurring within childhood that can be a source of stress or result in trauma, for example experiencing or witnessing abuse, neglect or violence.

Alexithymia – a construct that describes struggling with the identification and expression of the emotions of yourself and others, hindering abilities for introspection and communicating emotions.

Antipsychotic Medication – a type of medication used to help treat symptoms of psychosis through the modification of certain neurotransmitters in the brain. It is commonly prescribed for conditions characterised by psychosis, such as schizophrenia, and psychotic symptoms associated with other mental health conditions.

Aware – a support group for autistic children, young people and their families that is led by parents and offers support, advice and activities. This group does not require a formal autism diagnosis to participate.

Blinded trial – a clinical trial technique where aspects of the treatment condition are not shared with the participants and/or researchers, such as their assignment to an experimental or placebo condition.

Causal-inference methods – an analysis method that determines and quantifies the effect of one variable on another.

CBT (Cognitive Behavioural Therapy) – a form of talking therapy that provides intervention by exploring the relationship between thoughts, behaviours and mental health symptoms, with therapists supporting patients to alter dysfunctional thought patterns to improve and cope with symptoms.

Connect to Work – a programme funded by the UK Government that aims to provide employment support to individuals that may experience complex barriers to finding and sustaining work.

Co-occurrence – the occurrence of two or more distinct disorders that can vary in the extent to which they interact by potentially worsening or complicating respective symptoms.

CYPMHS (Children and Young People’s Mental Health Services) – NHS services providing mental health and emotional wellbeing support for children and young people (usually up to age 18 years), sometimes known as CAMHS (Child and Adolescent Mental Health Services).

Diagnostic overshadowing – the action of assigning symptoms to a pre-existing cause or condition consequently overlooking the consideration of alternative explanations for symptoms.

DSM (Diagnostic and Statistical Manual of Mental Disorders) – a handbook created and published by the American Psychiatric Association to provide medical information regarding the classifications of brain-related and mental health conditions. The most recent edition is the DSM-5-TR (Version 5, text revision) published in 2022.

Education and Health Care Plan (EHCP) – a legal document that lists the support needed for a child or young person (up to age 25 years) with special educational needs to inform the provision of support they require from their local authority.

Formulation – clinical formulation is a summary of the origins and nature of a person’s problems, together with opinion on what may go wrong in the future and what steps should be taken to improve matters.

ICD (International Classification of Diseases) – a classification published by the World Health Organisation (WHO) that defines the standardised terminology and classification of diseases and health data. The most recent edition is the ICD-11 (Version 11) published in 2019 and implemented in 2022.

ICF (International Classification of Functioning, Disability and Health) – a classification published by the World Health Organisation (WHO) that defines the classification of health and health-related domains to measure health and disability across population and individual levels.

Integrated Care Boards (ICBs) – a component of the NHS Integrated care system that manages NHS services by supporting local needs, governance arrangements and service delivery within their defined location.

Keep Britain Working – an independent review exploring how employers can address health-related economic inactivity as well as promoting inclusivity and health in workplaces.

Learning system – the routine utilisation of system data and insights drawn from data to inform continual innovation of the system.

Masking – adjusting expression based on the real or anticipated reactions of others, described as a strategy by autistic people in particular.

Matched care – a model of care that encourages the consideration of contextual factors surrounding the patient’s opinions and experiences to promote a personalised and thorough approach to treatment and treatment decisions.

Morbidity – to have an illness, condition or disease.

National Institute for Health and Care Excellence (NICE) – An independent public body that provides evidence-based guidance, advice, and quality standards for health, public health, and social care.

NEET (Not in Education, Employment or Training) – a young person, between the age of 16 and 24 years, who is not in education, employment or undertaking any form of training.

NHS Talking Therapies – a service available through the NHS that provides psychological interventions for UK adults experiencing common mental health problems like depression and anxiety disorders.

Occupational health – a professional field that centres around the promotion and maintenance of the mental and physical health of workers, working conditions and workplace organisation across occupations.

Outcome – the result arising from an event or action taken.

P factor – a score representing an individual’s psychopathology, capturing shared variance across psychiatric symptoms, mental disorders and including indicators across multiple domains like symptom severity, symptom duration, developmental history and life impairment. A high score reflects higher burden and severity of psychopathology.

Population registry studies – studies that utilise data from national or regional registries that track characteristics and features within populations such as demographic characteristics and distributions of diseases.

POTS (Postural Orthostatic Tachycardia Syndrome) – a health condition linked to the autonomic nervous system and symptoms arising from heart rate changes when moving between lying and upright positions. The condition can be characterised by dizziness, fatigue, and light-headedness when moving between lying, sitting and standing positions.

Prevention – action taken to avoid the occurrence or worsening of a risked outcome.

Prognosis – the expected outcome following illness or condition, such as relating to recovery.

Psychiatry – the medical specialty and branch of medicine focusing on mental, personality and behavioural disorders.

Psychoeducation – a therapeutic intervention for patients, families and supporters that provides information and support to better understand and cope with illness.

Psychology – an empirical scientific field focused on the mind and behaviour, with many practitioners involved in therapeutic roles.

Psychosocial – an approach which looks at individuals in the context of the combined influence that psychological factors and the surrounding social environment have on their physical and mental wellness and their ability to function.

Randomised controlled trial (RCT) – a type of experiment involving participants being randomly assigned to groups to test the efficacy of an intervention. Random assignment to a treatment group, control or placebo group aims to reduce bias.

Recovery – the process of returning to a normal or improved state after challenges to mental and/or physical health.

Resilience – the adaptation to adversity through flexibility towards demands, such as by promptly recovering from difficulties or challenges.

Right to Choose – a framework that allows NHS England patients to choose to send their NHS referral to another NHS England provider outside of their local NHS. This may include providers outside a patient’s local area, including independent organisations delivering NHS-funded care. It does not apply in Wales, Scotland or Northern Ireland.

Risk – the chance or probability of an uncertain outcome that induces consequences upon occurrence.

SEND (Special educational needs and disability) – an individual with special educational needs and/or disability that may require additional support around health and/or education.

Severity – the level of intensity associated with symptoms or a condition.

Social prescribing – an approach that connects people to activities, groups, and services in their community to meet the practical, social and emotional needs that affect their health and wellbeing.

SSRIs (Selective Serotonin Reuptake Inhibitors) – a type of antidepressant medication that works to improve mood through preventing the reabsorption of the neurotransmitter serotonin.

Stepped care – a model of care that provides increasingly intensive treatment options in line with the intensity of symptoms, often informed by symptom questionnaires e.g. PHQ-9 (Patient Health Questionnaire) categorising symptom intensity or following the insufficiency of the lower level of intervention.

Stratified care (also known as matched care) – a model of care that tailors treatment decisions to subgroups of patients by acknowledging their characteristics and the potential effects of these characteristics on their treatment needs.

Surveillance – collecting, analysing and disseminating health data to facilitate the improvement of public health decision making and knowledge.

Thrive – an approach in schools to support the social and emotional needs of children and young people.

Trauma-informed service – a service that aims to recognise the effects of exposure to trauma, this knowledge is applied to prevent interaction with the service causing further harm or distress.

Universal support – a government programme that encourages employment amongst people with health conditions, disabilities and/or facing barriers to employment by supporting them to overcome challenges to maintaining sustained employment.

Unmasking – the process of reducing or stopping behaviours used by people with autism to conceal autistic traits and conform to social expectations.

Acknowledgements

Individuals acknowledged in this report have contributed to the work of the Review. The Chair and Vice-Chairs of the Review are grateful for their input, which has been enormously valued. They are also particularly grateful for the support provided by the Secretariat team of dedicated civil servants who have gone above and beyond the call of duty. All the findings and final conclusions of the Review set out in this report are the responsibility of the Chair and Vice-Chairs.

Chair

Professor Peter Fonagy CBE FMedSci FBA FAcSS

Vice-Chairs

Professor Gillian Baird Obe Ma, Mb BChir, FRCPCH, DCH consultant in Children's Neurodisability

Professor Sir Simon Wessely FMedSci FRS GBE

Secretariat

Sarah Dunsdon, Kate Headington, Amelia Horn, Adam Jones, Rhys Jose, Charlotte Kume-Holland, Alessandra Moon, Jo Sampson and Anika Sharma.

Steering group

Dr Michael Absoud, Professor Marios Adamou, Dr Charlotte Caniff, Jane Cannon, Professor Tony Charman, Professor Johnny Downs, Professor Dougal Hargreaves, Professor Chris Hollis, Dr Sarah Hughes, Dr Camilla Kingdon, Professor William Mandy, Professor Sally McManus, Wendy Minhinnett, Dr Anna Moore, Professor Praveetha Patalay, Dr Matthias Pierce, Professor Steve Pilling, Professor Alastair Santhouse, Professor Jenny Shand, Professor Emily Simonoff, Professor Ilina Singh, Professor Argyris Stringaris and Dr Adrian Whittington.

Topic group members

Dr Yushi Bai, Sir Simon Baron-Cohen, Andy Bell, Dr Sunil Bhopal, Professor Kamaldeep Bhui, Dr Ruth Blackburn, Professor Owen Bowden-Jones, Dr Ana-Maria Butura Santos, Professor Richard Byng, Professor Tanya Byron, Paul Calaminus, Dr Alison Cape, Professor Rudolf Cardinal, Matthew Carpenter, Dr Michael Chapman,

Professor Carolyn Chew-Graham, Professor Rhiannon Corcoran, Professor Samuele Cortese, Professor Mike Crawford, Professor Cathy Creswell, Dr Connor Cummings, Professor Prathiba Chitsabesan, Emma-Jane Clark, Professor Jayati Das-Munshi, Professor Anthony David, Professor Jessica Deighton, Dr Jennifer Dixon, Sahil Dodhia, Dr Alex Dregan, Vicky Dutchburn, Thomas Easdale-Cheele, Paul Edwards, Vicki Ensor, Kaisu Fagan, Professor Daisy Fancourt, Professor Mina Fazel, Dr Lucy Foulkes, Professor Tamsin Ford, Dr David Frayman, Dr Matthew Fysh, Professor Catherine Gallop, Professor Ruth Gilbert, Dr Suzi Godson, Professor Karen Guldberg, Gemma Gradwell, Rebecca Gray, Professor Jonathan Green, Dr Nick Grey, Professor Francesca Happé, Dr Val Harpin, Patricia Howlin, Victoria Hornby, Professor Neil Humphrey, Professor Ann John, Ellen Kretschmer, Professor Judi Kidger, Dr Stephanie Lamb, Dame Christine Lenehan, Professor Sonia Livingstone, Rebecca Lucas, Dr Steve Lukito, Dr Sarah Markham, Dr Alex Martin, Roisin McEvoy, Professor Mark Mon-Williams, Professor Craig Morgan, Dr Faraz Mughal, Dr Sharon Neufeld, Associate Professor Tamsin Newlove-Delgado, Professor Courtenay Norbury, Dr Jade Eloise Norris, Professor Dennis Ougrin, Professor Amy Orben, Dr Tom Osborn, Dr Suzanne O'Sullivan, Professor Jeremy Parr, Peter Pratt, Professor Dheeraj Rai, Dr Fiona Rattray, Dr Maria Rodriguez, Mark Rowland, Dr Aws Sadik, Professor Kapil Sayal, Dr Terry Segal, Nidhi Sethi, Professor Philip Shaw, Professor Doug Simkiss, Dr Rachel Sippy, Professor Edmund Sonuga-Barke, Jack Smith, Thea Stein, Associate Professor Maxime Taquet, Dr Emma Tiffin, Dr Samuel Tromans, Dr Catherine Tuffrey, Professor Essi Viding, Varun Warrier, Laiba Waseem, Dr Elizabeth Weir, Dr Simon White, Dr Susannah Whitwell, Alice Wickersham and Professor Jane Wills.

Individuals who have significantly directly contributed

Nicoletta Adamo, Dr Awais Aftab, Mia Akiba, Elizabeth Allison, Professor Carrie Allison, Jessica Armitage, Eleonora Armyra, Arjoon Arunasalam, Professor Louis Appleby, Elena Bagdades, Martin Batty, Dr Zharain Bawa, Professor, Graham Beales, Sir David Behan, Dr Julie Bieles, Dr Lesley Booth, Nicola Buckley, Chloe Campbell, Ayana Cant, Dr Wenzhen Cao, Baroness Hilary Cass, Dr Matteo Catanzano, Jasmine Chavda, Dr Natasha Chilman, Brian CF Ching, Anne Conolly, Dr David Crepaz-Keay, Joseph Crouch, Sir Cyril Chantler, Craig Colling, Nick Cosstick, George Davey Smith, Elizaveta Dimitrova, Corinna Dymond, Dr Sophie Epstein, Robin Evans, Dr Anastasia Fadeeva, Benjamin Fell, Zoe Firth, Dr Mariane Furtado, Professor Sebastian Gaigg, Dr Katerina Gklantzouni, Dr Laura Gimeno, Ladan Hashemi, Dr John Headley Ward, Yihang Huang, Holly Honderich, Dr Amy Rachael Hodgson, Chris Hughes, Dr Jim Hughes, Tassella Isaac, Dr Amber John, Meghann Jones, Dr Dominic P. Kelly, Asa Kerr-Davis, Paul Kirby, Dr Stephanie J Lewis, Dr Su Mon Latt, Emilia Loane, Dr Amanda Ludlow, Professor Daniele

Marcotulli, Tanya Marwaha, Margaret McCartney, George Melrose, Rocío Méndez Pineda, Sasha Menon, Baroness Sally Morgan, Dr Zoe Morgan, Sarah Morris, Anna Morris, Dr Jacqueline Nonweiler, Matilda North, Jessica Norman, Nnenna Omeje, Dr Rashmi Patel, Dr Mirabel Pelton, Jessica Penhallow, Dr William Pettersson-Yeo, Adam Pill, Antonina H. Przyborowska, Sakshi Sangekar, Dr Abigail Runicles, Taj Sallamuddin, Kavitha Saravanakumar, Merle Schlief, A Seker, Perses Sethna, Jayashree Ravi Shetty, Henry Shelford, Evie Shore, Rachel Solomon, Dr Thomas Steare, Joshua Stott, Dr Evangeline Tabor, Dr Vicky P. Taxiarchi, Dr Laurence Telesia, Professor Daksha Trivedi, Alex Truscott, Alex Tsompanidis, Po-Chang Tseng, Professor Russell Viner, Myriam Vriend, Matthew Warburton, Lord Norman Warner, Yuting Wu, Shuwen Yin, Daniel Young, Mirela Zaneva and Dr Shuo Zhang

References

1 Cardoso F and McHayle Z. ‘The economic and social costs of mental ill health’ Centre for Mental Health, 2024. https://www.centreformentalhealth.org.uk/publications/the-economic-and-social- costs-of-mental-ill-health/

2 Liubertiene G and others. ‘Common mental health conditions’. In: Morris S and others, editors. ‘Adult Psychiatric Morbidity Survey: Survey of Mental Health and Wellbeing, England, 2023/4’ NHS England, 2025. https://digital.nhs.uk/data-and- information/publications/statistical/adult-psychiatric-morbidity-survey/survey-of- mental-health-and-wellbeing-england-2023-24/common-mental-health-conditions

3 McManus S and others. ‘Prevalence of non-suicidal self-harm and service contact in England, 2000-14: repeated cross-sectional surveys of the general population’ Lancet Psychiatry 2019: volume 6, pages 573-581. https://doi.org/10.1016/S2215- 0366(19)30188-9

4 Faria C and others. ‘The global prevalence of eating disorders in children and young people: a systematic review and meta-analysis’ European Child & Adolescent Psychiatry 2026: volume 35, pages 1093-1106. https://doi.org/10.1007/s00787-025- 02933-0

5 Brown A and others. ‘Psychotic disorder’. In: Morris S and others, editors. ‘Adult Psychiatric Morbidity Survey: Survey of Mental Health and Wellbeing, England, 2023/4’ NHS England, 2025. https://digital.nhs.uk/data-and- information/publications/statistical/adult-psychiatric-morbidity-survey/survey-of- mental-health-and-wellbeing-england-2023-24/psychotic-disorder

6 Randall E and others. ‘Bipolar disorder’. In: Morris S and others, editors. ‘Adult Psychiatric Morbidity Survey: Survey of Mental Health and Wellbeing, England, 2023/4’ NHS England, 2025. https://digital.nhs.uk/data-and- information/publications/statistical/adult-psychiatric-morbidity-survey/survey-of- mental-health-and-wellbeing-england-2023-24/bipolar-disorder

7 John A and others. ‘Attention-deficit/hyperactivity disorder in children and adults in England, 2000-2025: recorded prevalence and diagnostic trends in a population- based observational study using routinely collected primary care data’ Lancet Regional Health – Europe 2026: volume 67, page 101740. https://doi.org/10.1016/j.lanepe.2026.101740

8 NHS England. ‘ADHD Management Information - February 2026’ 2026. https://digital.nhs.uk/data-and-information/publications/statistical/mi-adhd/february- 2026

9 Frayman D and Mandy W. ‘Growing administrative prevalence of autism spectrum disorder among English school pupils: heterogeneity and implications’ PsyArXiv 2025. https://doi.org/10.31234/osf.io/cr4k5_v2

10 O’Nions E and others. ‘Autism in England: assessing underdiagnosis in a population-based cohort study of prospectively collected primary care data’ Lancet Regional Health – Europe 2023: volume 29, page 100626. https://doi.org/10.1016/j.lanepe.2023.100626

11 Frayman D and Mandy W. ‘Rising autism identification in England, 2016-2024: evidence from England's School Census’ Journal of Child Psychology and Psychiatry 2026. https://doi.org/10.1111/jcpp.70209

12 NHS England. ‘Autism Statistics, January 2025 to December 2025’ 2026. https://digital.nhs.uk/data-and-information/publications/statistical/autism- statistics/january-2025-to-december-2025

13 Mandy W. ‘What is autism, now?’ Zenodo 2025. https://doi.org/10.5281/zenodo.17989621

14 Fazel M and Patel S. ‘OxWell 2025: Summary report’ University of Oxford, 2025. https://ora.ox.ac.uk/objects/uuid:bc01a40c-81b0-4151-b7b4- fbe20b35398f/files/r6108vf03w

15 Morris S and others. ‘Adult Psychiatric Morbidity Survey: Survey of Mental Health and Wellbeing, England, 2023/4.’ NHS England, 2025. https://digital.nhs.uk/data- and-information/publications/statistical/adult-psychiatric-morbidity-survey

16 Vriend M and Atwell S. ‘Young people with mental health conditions are now more likely to be NEET’ The Health Foundation, 2026. https://www.health.org.uk/reports-and-analysis/analysis/young-people-with-mental- health-conditions-are-now-more-likely-to-be-NEET

17 Kelly DP and others. ‘Adolescent health and Not in Education, Employment or Training (NEET) in young adulthood: evidence from a UK prospective longitudinal study’ medRxiv 2026. https://doi.org/10.64898/2026.08.13.26360381

18 Rowland D. ‘Market failure. How the under-regulated market in NHS funded ADHD services impacts patients and the finances of the NHS’ Centre for Health and the Public Interest, 2026. https://www.chpi.org.uk/reports/how-the-under-regulated- market-in-nhs-funded-adhd-services

19 Care Quality Commission. ‘The state of health care and adult social care in England 2024/25’ 2025. https://www.cqc.org.uk/publications/major-report/state- care/2024-2025

20 Meltzer H and others. ‘The mental health of young people looked after by local authorities in England’ The Stationery Office, 2003. https://webarchive.nationalarchives.gov.uk/ukgwa/20160107072810/http://www.ons. gov.uk/ons/rel/psychiatric-morbidity/mental-health-of-young-people-looked-after-by- local-authorities/2002-survey/mental-health-of-young-people-looked-after-by-local- authorities-in-england.pdf

21 Department for Education. ‘Education, children’s social care and offending: local authority level dashboard. Academic year 2019/20’ 2023. https://explore-education- statistics.service.gov.uk/find-statistics/education-children-s-social-care-and- offending-local-authority-level-dashboard/2019-20

22 Mayfield C. ‘Keep Britain working: final report’ 2026. https://assets.publishing.service.gov.uk/media/6909fac488a98da87e292282/keep- britain-working-review-final-report.pdf

23 Department of Health and Social Care. ‘Independent review into mental health conditions, ADHD and autism: terms of reference’ 2025. https://www.gov.uk/government/publications/independent-review-into-mental-health- conditions-adhd-and-autism-terms-of-reference/independent-review-into-mental- health-conditions-adhd-and-autism-terms-of-reference

24 Atwell S and others. ‘Why are a growing number of young people who are NEET reporting work-limiting health conditions?’ The Health Foundation, 2026. https://www.health.org.uk/reports-and-analysis/analysis/why-are-a-growing-number- of-young-people-who-are-neet-reporting-work

25 Akanni L and others. ‘Childhood socio-emotional and cognitive development and adolescents NEET (not in education, employment or training): findings from the UK Millennium Cohort Study’ BMJ Open 2026: volume 16, page e109720. https://doi.org/10.1136/bmjopen-2025-109720

26 Department of Health & Social Care. ‘Independent review into mental health conditions, ADHD and autism: interim report’ 2026. https://www.gov.uk/government/publications/independent-review-into-mental-health- conditions-adhd-and-autism-interim-report

27 Ronaldson A and Henderson C. ‘Investigating changes in mental illness stigma and discrimination after the Time to Change programme in England’ BJPsych Open 2024: volume 10, page e199. https://doi.org/10.1192/bjo.2024.801

28 Mind. ‘The Big Mental Health Report’ 2025. https://www.mind.org.uk/about- us/our-policy-work/the-big-mental-health-report/

29 World Health Organization. ‘International classification of functioning, disability and health: ICF’ World Health Organization, 2001. https://www.who.int/standards/classifications/international-classification-of- functioning-disability-and-health

30 NHS Digital. ‘Adult Psychiatric Morbidity Survey: survey of mental health and wellbeing, England’ 2025. https://digital.nhs.uk/data-and- information/publications/statistical/adult-psychiatric-morbidity-survey

31 NHS Digital. ‘Mental health of children and young people in England, 2017’ 2017. https://digital.nhs.uk/data-and-information/publications/statistical/mental-health-of- children-and-young-people-in-england

32 University of Essex, Institute for Social and Economic Research. ‘Understanding Society: Waves 1-15, 2009-2024 and Harmonised BHPS: Waves 1-18, 1991-2009 [data collection]. 21st edition’ UK Data Service. SN: 6614, 2026. http://doi.org/10.5255/UKDA-SN-6614-22

33 NHS Digital. ‘Health survey for England’ 2024. https://digital.nhs.uk/data-and- information/publications/statistical/health-survey-for-england

34 Rickwood DJ and Coleman-Rose CL. ‘The effect of survey administration mode on youth mental health measures: social desirability bias and sensitive questions’ Heliyon 2023: volume 9, page e20131. https://doi.org/10.1016/j.heliyon.2023.e20131

35 Independent Healthcare Providers Network (IHPN). ‘Building better neurodevelopmental services’ 2026. https://www.ihpn.org.uk/news/new-ihpn-report- building-better-neurodevelopmental-services/

36 Moore A and others. ‘A protocol for a multi-site cohort study to evaluate child and adolescent mental health service transformation in England using the i-THRIVE model’ Plos One 2023: volume 18, page e0265782. https://doi.org/10.1371/journal.pone.0265782

  1. Timely. 2026. https://timely.org.uk/
  2. Global Burden of Disease Collaborative Network. ‘Global Burden of Disease

Study 2023’ 2025. https://www.healthdata.org/research-analysis/library/global- burden-disease-2023-findings-gbd-2023-study

  1. Jorm AF. ‘Mental health literacy. public knowledge and beliefs about mental

disorders’ British Journal of Psychiatry 2000: volume 177, pages 396-401. https://doi.org/10.1192/bjp.177.5.396

  1. Evans-Lacko S and others. ‘Effect of the Time to Change anti-stigma campaign

on trends in mental-illness-related public stigma among the English population in 2003-13: an analysis of survey data’ Lancet Psychiatry 2014: volume 1, pages 121- 128. https://doi.org/10.1016/S2215-0366(14)70243-3

  1. Schnittker J. ‘The diagnostic system: why the classification of psychiatric

disorders is necessary, difficult, and never settled’ Columbia University Press, 2017

  1. World Health Organization. ‘International classification of diseases, 11th revision’

2019. https://icd.who.int/en/

  1. American Psychiatric Association. ‘Diagnostic and statistical manual of mental

disorders, fifth edition, text revision: DSM-5-TR’ American Psychiatric Publishing, Inc, 2022

  1. Constantino JN and Todd RD. ‘Autistic traits in the general population: a twin

study’ Archives of General Psychiatry 2003: volume 60, pages 524-530. https://doi.org/10.1001/archpsyc.60.5.524

  1. Arildskov TW and others. ‘The factor structure of attention-deficit/hyperactivity

disorder in schoolchildren’ Research in Developmental Disabilities 2022: volume 125, page 104220. https://doi.org/10.1016/j.ridd.2022.104220

  1. Botha M and others. ‘The neurodiversity concept was developed collectively: an

overdue correction on the origins of neurodiversity theory’ Autism 2024: volume 28, pages 1591-1594. https://doi.org/10.1177/13623613241237871

  1. Pellicano E and den Houting J. ‘Annual research review: Shifting from ‘normal

science’ to neurodiversity in autism science’ Journal of Child Psychology and Psychiatry 2022: volume 63, pages 381-396. https://doi.org/10.1111/jcpp.13534

  1. Baron-Cohen S. ‘Editorial perspective: Neurodiversity - a revolutionary concept

for autism and psychiatry’ Journal of Child Psychology and Psychiatry 2017: volume 58, pages 744-747. https://doi.org/10.1111/jcpp.12703

  1. Oliver M. ‘The politics of disablement — new social movements’. In: ‘The politics

of disablement’ Red Globe Press, 1990, pages 112-131.

50 ‘Equality Act 2010’ UK Public General Acts, 2010 c. 15. https://www.legislation.gov.uk/ukpga/2010/15

51 Kenny L and others. ‘Which terms should be used to describe autism? Perspectives from the UK autism community’ Autism 2016: volume 20, pages 442- 462. https://doi.org/10.1177/1362361315588200

52 Bottema-Beutel K and others. ‘Avoiding ableist language: suggestions for autism researchers’ Autism in Adulthood 2021: volume 3, pages 18-29. https://doi.org/10.1089/aut.2020.0014

53 French B and others. ‘The power of words: respectful language in ADHD research’ Lancet Psychiatry 2025: volume 12, pages 876-879. https://doi.org/10.1016/S2215-0366(25)00167-1

54 Faraone SV and Miller MJ. ‘Precision language in ADHD research’ Lancet Psychiatry 2026: volume 2026. https://doi.org/10.1016/S2215-0366(26)00117-3

55 Brodersen J and others. ‘Overdiagnosis: what it is and what it isn’t’ BMJ Evidence-Based Medicine 2018: volume 23, page 1. https://doi.org/10.1136/ebmed- 2017-110886

56 Aftab A. ‘The “overdiagnosis” confusion’ 2025. https://www.psychiatrymargins.com/p/the-overdiagnosis-confusion

57 BBC News. ‘'Life being stressful is not an illness' - GPs on mental health over- diagnosis’ 2025. https://www.bbc.co.uk/news/articles/cx2pvxdn9v4o

58 Sadik A and others. ‘Changes in the profile of adults diagnosed as autistic since 2010: population based studies in England and Sweden’ medRxiv 2026. https://doi.org/10.64898/2026.05.20.26353486

59 Cortese S and others. ‘ADHD (over) diagnosis: fiction, fashion and failure’ British Journal of Psychiatry 2026. https://doi.org/10.1192/bjp.2026.10546

60 Sayal K and Hiller R. ‘Debate: Are we overpathologising young people’s mental health? Research shows otherwise – mental health conditions are not being recognised or diagnosed in healthcare settings’ Child and Adolescent Mental Health 2025: volume 30, pages 305-307. https://doi.org/10.1111/camh.70019

61 Centre for Evidence-Based Medicine. ‘The Preventing Overdiagnosis International Conference’ 2026. https://www.cebm.ox.ac.uk/preventing-overdiagnosis

62 David AS and Deeley Q. ‘Dangers of self-diagnosis in neuropsychiatry’ Psychological Medicine 2024: volume 54, pages 1057-1060. https://doi.org/10.1017/S0033291724000308

63 Botha M and others. ‘Does language matter? Identity-first versus person-first language use in autism research: a response to Vivanti’ Journal of Autism and Developmental Disorders 2023: volume 53, pages 870-878. https://doi.org/10.1007/s10803-020-04858-w

64 Botha M and others. ‘“Autism is me”: an investigation of how autistic individuals make sense of autism and stigma’ Disability & Society 2022: volume 37, pages 427- 453. https://doi.org/10.1080/09687599.2020.1822782

65 Davies J and others. ‘Autistic identity: a systematic review of quantitative research’ Autism Research 2024: volume 17, pages 874-897. https://doi.org/10.1002/aur.3105

66 Fletcher-Watson S. ‘What’s in a name? The costs and benefits of a formal autism diagnosis’ Autism 2024: volume 28, pages 257-262. https://doi.org/10.1177/13623613231213300

67 Corzine A and Roy A. ‘Inside the black mirror: current perspectives on the role of social media in mental illness self-diagnosis’ Discover Psychology 2024: volume 4, page 40. https://doi.org/10.1007/s44202-024-00152-3

68 Johnson AL. ‘An exploratory analysis of self-diagnosis on Reddit’ SSM - Mental Health 2026: volume 9, page 100597. https://doi.org/10.1016/j.ssmmh.2026.100597

69 Jutel A. ‘Sociology of diagnosis: a preliminary review’ Sociology of Health & Illness 2009: volume 31, pages 278-299. https://doi.org/10.1111/j.1467- 9566.2008.01152.x

70 Marcotulli D and others. ‘Editorial Perspective: How spreading mental health information can be (un-) helpful – a dynamic systems approach’ Journal of Child Psychology and Psychiatry 2026: volume 67, pages 429-438. https://doi.org/10.1111/jcpp.70055

71 Rivas-Vazquez RA and others. ‘Adult ADHD: underdiagnosis of a treatable condition’ Journal of Health Service Psychology 2023: volume 49, pages 11-19. https://doi.org/10.1007/s42843-023-00077-w

72 Sibley MH and others. ‘Variable patterns of remission from ADHD in the Multimodal Treatment Study of ADHD’ American Journal of Psychiatry 2022: volume 179, pages 142-151. https://doi.org/10.1176/appi.ajp.2021.21010032

73 Keyes KM and others. ‘Cohort effects explain the increase in autism diagnosis among children born from 1992 to 2003 in California’ International Journal of Epidemiology 2012: volume 41, pages 495-503. https://doi.org/10.1093/ije/dyr193

74 Arvidsson O and others. ‘Secular changes in the symptom level of clinically diagnosed autism’ Journal of Child Psychology and Psychiatry 2018: volume 59, pages 744-751. https://doi.org/10.1111/jcpp.12864

75 Pierre JM. ‘Overdiagnosis, underdiagnosis, synthesis: a dialectic for psychiatry and the DSM’. In: Paris J and Phillips, J, editors. ‘Making the DSM-5: concepts and controversies’ Springer New York, 2013, pages 105-124. https://doi.org/10.1007/978-1-4614-6504-1_8

76 Stott J. ‘Unpublished CPRD Aurum dataset analyses’ Personal communication, 2026.

77 Stringaris A and others. ‘How to understand and deal with “normality” and “disorder” in adolescent mental health care’ World Psychiatry in press. https://drive.google.com/file/d/1Xy48To1VjVhi0buttkKt7T9v-BBkUvXF/view

78 Hofstad T and others. ‘Catchment area characteristics do not account for geographical variation in ADHD diagnoses’ European Child & Adolescent Psychiatry 2025: volume 34, pages 3097-3103. https://doi.org/10.1007/s00787-025-02720-x

79 Lai MC and others. ‘Prevalence of co-occurring mental health diagnoses in the autism population: a systematic review and meta-analysis’ Lancet Psychiatry 2019: volume 6, pages 819-829. https://doi.org/10.1016/S2215-0366(19)30289-5

80 Micai M and others. ‘Prevalence of co-occurring conditions in children and adults with autism spectrum disorder: a systematic review and meta-analysis’ Neuroscience and Biobehavioral Reviews 2023: volume 155, page 105436. https://doi.org/10.1016/j.neubiorev.2023.105436

81 Chikaura TA and others. ‘Traumatic experiences, psychological distress and suicide-related behaviors in autistic adults’ Autism Research 2026: volume 19, page e70137. https://doi.org/10.1002/aur.70137

82 Brown CM and others. ‘Hiding in plain sight: eating disorders, autism, and diagnostic overshadowing in women’ Autism in Adulthood 2024: volume 8, pages 410-422. https://doi.org/10.1089/aut.2023.0197

83 Siddiqui U and others. ‘Sex differences in diagnosis and treatment timing of comorbid depression/anxiety and disease subtypes in patients with ADHD: a database study’ Journal of Attention Disorders 2024: volume 28, pages 1347-1356. https://doi.org/10.1177/10870547241251738

84 Zhang R and others. ‘Association of autism diagnosis and polygenic scores with eating disorder severity’ European Eating Disorders Review 2022: volume 30, pages 442-458. https://doi.org/10.1002/erv.2941

85 McCartney M. ‘The patient paradox: why sexed up medicine is bad for your health’ Montag & Martin, 2013

86 Hart JT. ‘The inverse care law’ Lancet 1971: volume 297, pages 405-412. https://doi.org/10.1016/s0140-6736(71)92410-x

87 Hennig-Schmidt H and others. ‘How payment systems affect physicians’ provision behaviour—an experimental investigation’ Journal of Health Economics 2011: volume 30, pages 637-646. https://doi.org/10.1016/j.jhealeco.2011.05.001

88 Silverman E and Skinner J. ‘Medicare upcoding and hospital ownership’ Journal of Health Economics 2004: volume 23, pages 369-389. https://doi.org/10.1016/j.jhealeco.2003.09.007

89 Porter R. ‘The greatest benefit to mankind: a medical history of humanity’ HarperCollins, 1997

90 O’Connor C and others. ‘How does psychiatric diagnosis affect young people's self-concept and social identity? A systematic review and synthesis of the qualitative

literature’ Social Science & Medicine 2018: volume 212, pages 94-119. https://doi.org/10.1016/j.socscimed.2018.07.011

91 Haslam N. ‘Concept creep: psychology’s expanding concepts of harm and pathology’ Psychological Inquiry 2016: volume 27, pages 1-17. https://doi.org/10.1080/1047840X.2016.1082418

92 Singh I. ‘Not robots: children’s perspectives on authenticity, moral agency and stimulant drug treatments’ Journal of Medical Ethics 2013: volume 39, pages 359- 366. https://doi.org/10.1136/medethics-2011-100224

93 Beeker T and others. ‘Editorial: Psychiatrization of society’ Frontiers in Sociology 2023: volume 8, page 1258264. https://doi.org/10.3389/fsoc.2023.1258264

94 Brinkmann S. ‘Languages of suffering’ Theory & Psychology 2014: volume 24, pages 630-648. https://doi.org/10.1177/0959354314531523

95 Nayyar JM and others. ‘Exploring lived experiences of receiving a diagnosis of autism in adulthood: a systematic review’ Autism in Adulthood 2025: volume 7, pages 1-12. https://doi.org/10.1089/aut.2023.0152

96 Hansson Halleröd SL and others. ‘Experienced consequences of being diagnosed with ADHD as an adult – a qualitative study’ BMC Psychiatry 2015: volume 15, page 31. https://doi.org/10.1186/s12888-015-0410-4

97 Cortese S. ‘Incidence of ADHD diagnoses on the rise—good or bad news?’ JAMA Network Open 2026: volume 9, page e265784. https://doi.org/10.1001/jamanetworkopen.2026.5784

98 Ahuvia IL and others. ‘Depression self-labeling in U.S. college students: associations with perceived control and coping strategies’ Journal of Affective Disorders 2024: volume 351, pages 202-210. https://doi.org/10.1016/j.jad.2024.01.229

99 Ahuvia IL and Link BG. ‘The mental illness self-labeling model: a conceptual model for studying the effects of mental-illness self-labeling on clinical outcomes’ Clinical Psychological Science 2025: volume 13, pages 1031-1050. https://doi.org/10.1177/21677026251338829

100 Harari L and others. ‘The power of self-labels: examining self-esteem consequences for youth with mental health problems’ Journal of Health and Social

Behavior 2023: volume 64, pages 578-592. https://doi.org/10.1177/00221465231175936

101 Smith JL. ‘Sir Arbuthnot Lane, chronic intestinal stasis, and autointoxication’ Annals of Internal Medicine 1982: volume 96, pages 365-369. https://doi.org/10.7326/0003-4819-96-3-365

102 Duffy A and others. ‘Pre-pubertal bipolar disorder: origins and current status of the controversy’ International Journal of Bipolar Disorders 2020: volume 8, page 18. https://doi.org/10.1186/s40345-020-00185-2

  1. O’Sullivan S. ‘The age of diagnosis’ Penguin Random House, 2025
  2. Lilley R and others. ‘’A way to be me‘: autobiographical reflections of autistic

adults diagnosed in mid-to-late adulthood’ Autism 2022: volume 26, pages 1395- 1408. https://doi.org/10.1177/13623613211050694

  1. O’Nions E and others. ‘Life expectancy and years of life lost for adults with

diagnosed ADHD in the UK: matched cohort study’ British Journal of Psychiatry 2025: volume 226, pages 261-268. https://doi.org/10.1192/bjp.2024.199

  1. Zhang L and others. ‘ADHD drug treatment and risk of suicidal behaviours,

substance misuse, accidental injuries, transport accidents, and criminality: emulation of target trials’ BMJ 2025: volume 390, page e083658. https://doi.org/10.1136/bmj- 2024-083658

  1. French B and others. ‘Risks associated with undiagnosed ADHD and/or autism:

a mixed-method systematic review’ Journal of Attention Disorders 2023: volume 27, pages 1393-1410. https://doi.org/10.1177/10870547231176862

  1. Bottema-Beutel K and others. ‘A systematic review and meta-regression

analysis of social functioning correlates in autism and typical development’ Autism Research 2019: volume 12, pages 152-175. https://doi.org/10.1002/aur.2055

  1. Kolbas V and others. ‘A typology of mental health difficulties among children and

young people in England: a latent class analysis of data from the 2017 Mental Health of Children and Young People Survey’ MIND, 2021.

110 Mykletun A and others. ‘Causal modelling of variation in clinical practice and long-term outcomes of ADHD using Norwegian registry data: the ADHD controversy project’ BMJ Open 2021: volume 11, page e041698. https://doi.org/10.1136/bmjopen-2020-041698

111 Lyhmann I and others. ‘Variation in attitudes toward diagnosis and medication of ADHD: a survey among clinicians in the Norwegian child and adolescent mental health services’ European Child & Adolescent Psychiatry 2023: volume 32, pages 2557-2567. https://doi.org/10.1007/s00787-022-02110-7

112 Widding-Havneraas T and others. ‘Geographical variation in ADHD: do diagnoses reflect symptom levels?’ European Child & Adolescent Psychiatry 2023: volume 32, pages 1795-1803. https://doi.org/10.1007/s00787-022-01996-7

113 Chaulagain A and others. ‘Effect of pharmacological treatment of attention– deficit hyperactivity disorder on contacts with child welfare services’ The British Journal of Psychiatry 2026: pages 1-8. https://doi.org/10.1192/bjp.2026.10554

114 Widding-Havneraas T and others. ‘Effect of pharmacological treatment of attention-deficit/hyperactivity disorder on criminality’ Journal of the American Academy of Child & Adolescent Psychiatry 2024: volume 63, pages 433-442. https://doi.org/10.1016/j.jaac.2023.05.025

115 Arnstein M and others. ‘Effect of pharmacological treatment of ADHD on academic performance: a registry-based population study’ British Journal of Psychiatry in press.

116 Lyhmann I and others. ‘Effect of pharmacological treatment of attention- deficit/hyperactivity disorder on later psychiatric comorbidity: a population-based prospective long-term study’ BMJ Mental Health 2024: volume 27, page e301003. https://doi.org/10.1136/bmjment-2024-301003

117 Widding-Havneraas T and others. ‘Effect of ADHD medication on risk of injuries: a preference-based instrumental variable analysis’ European Child & Adolescent Psychiatry 2024: volume 33, pages 1987-1996. https://doi.org/10.1007/s00787-023- 02294-6

118 Mykletun A and others. ‘The ADHD controversy project: Long-term effects of ADHD diagnosis (ISRCTN11573246)’ ISRCTN Registry 2020. https://doi.org/10.1186/ISRCTN11573246

119 Kazda L and others. ‘Association of attention-deficit/hyperactivity disorder diagnosis with adolescent quality of life’ JAMA Network Open 2022: volume 5, page e2236364. https://doi.org/10.1001/jamanetworkopen.2022.36364

120 Owens J. ‘Relationships between an ADHD diagnosis and future school behaviors among children with mild behavioral problems’ Sociology of Education 2020: volume 93, pages 191-214. https://doi.org/10.1177/0038040720909296

121 Chaulagain A and others. ‘A systematic meta-review of systematic reviews on attention deficit hyperactivity disorder’ European Psychiatry 2023: volume 66, page e90. https://doi.org/10.1192/j.eurpsy.2023.2451

122 Davies W. ‘Young people are different’ 2026. https://www.lrb.co.uk/the- paper/v48/n14/william-davies/young-people-are-different

123 Haase CB and others. ‘New hypertension guidance risks overdiagnosis and overtreatment’ BMJ 2019: volume 365, page l1657. https://doi.org/10.1136/bmj.l1657

124 Regan M. ‘The interface between dementia and mental health: an evidence review’ Mental Health Foundation, 2016. https://www.mentalhealth.org.uk/sites/default/files/2022-09/MHF-dementia-mental- health.pdf

125 Matthews FE and others. ‘A two-decade comparison of prevalence of dementia in individuals aged 65 years and older from three geographical areas of England: results of the Cognitive Function and Ageing Study I and II’ Lancet 2013: volume 382, pages 1405-1412. https://doi.org/10.1016/S0140-6736(13)61570-6

126 Huang Y and others. ‘“I’ve spent my whole life striving to be normal”: internalized stigma and perceived impact of diagnosis in autistic adults’ Autism in Adulthood 2023: volume 5, pages 423-436. https://doi.org/10.1089/aut.2022.0066

127 Perkins A and others. ‘Experiencing mental health diagnosis: a systematic review of service user, clinician, and carer perspectives across clinical settings’ Lancet Psychiatry 2018: volume 5, pages 747-764. https://doi.org/10.1016/S2215- 0366(18)30095-6

128 Collins B and others. ‘Investigating the impact of undiagnosed anxiety and depression on health and social care costs and quality of life: cross-sectional study using household health survey data’ BJPsych Open 2023: volume 9, page e201. https://doi.org/10.1192/bjo.2023.596

129 Duong P and others. ‘Health-related quality of life and economic burden associated with likely and diagnosed generalized anxiety disorder among the US general adult population’ Journal of Affective Disorders 2026: volume 400, page 120972. https://doi.org/10.1016/j.jad.2025.120972

130 Baek CH and others. ‘Influence of biogenetic explanations of mental disorders on stigma and help-seeking behavior: a systematic review and meta-analysis’ Journal of Korean Medical Science 2023: volume 38, page e25. https://doi.org/10.3346/jkms.2023.38.e25

131 Loughman A and Haslam N. ‘Neuroscientific explanations and the stigma of mental disorder: a meta-analytic study’ Cognitive Research: Principles and Implications 2018: volume 3, page 43. https://doi.org/10.1186/s41235-018-0136-1

132 Goodman R and others. ‘The Development and Well-Being Assessment: description and initial validation of an integrated assessment of child and adolescent psychopathology’ Journal of Child Psychology and Psychiatry 2000: volume 41, pages 645-655. https://doi.org/10.1111/j.1469-7610.2000.tb02345.x

133 Lewis G and others. ‘Measuring psychiatric disorder in the community: a standardized assessment for use by lay interviewers’ Psychological Medicine 1992: volume 22, pages 465-486. https://doi.org/10.1017/s0033291700030415

134 Vizard T and others. ‘Mental Health of Children and Young People in England, 2017: Emotional disorders’ NHS Digital, 2018. https://files.digital.nhs.uk/14/0E2282/MHCYP%202017%20Emotional%20Disorders. pdf

135 Taxiarchi V and others. ‘Trends in mental health and behavioural outcomes among young adolescents in the UK: an updated analysis using the UK Household Longitudinal Study (2009–2023)’ PsyArXiv 2025. https://doi.org/10.31234/osf.io/e2u7j_v1

136 Fink E and others. ‘Mental health difficulties in early adolescence: a comparison of two cross-sectional studies in England from 2009 to 2014’ Journal of Adolescent Health 2015: volume 56, pages 502-507. https://doi.org/10.1016/j.jadohealth.2015.01.023

137 Patalay P and Gage SH. ‘Changes in millennial adolescent mental health and health-related behaviours over 10 years: a population cohort comparison study’ International Journal of Epidemiology 2019: volume 48, pages 1650-1664. https://doi.org/10.1093/ije/dyz006

138 McElroy E and others. ‘Socioeconomic and sex inequalities in parent-reported adolescent mental ill-health: time trends in four British birth cohorts’ Journal of Child Psychology and Psychiatry 2023: volume 64, pages 758-767. https://doi.org/10.1111/jcpp.13730

139 Institute for Social and Economic Research. ‘Insights 2026: Barriers to opportunity for children and young people’ University of Essex, 2026. https://www.understandingsociety.ac.uk/wp-content/uploads/2026/02/Insights-2026- Opportunity.pdf

140 Holt-White E and others. ‘Mental health and wellbeing: Wave 1 initial findings – Briefing No. 4’ 2022. https://cosmostudy.uk/publication_pdfs/mental-health-and- wellbeing.pdf

141 Gondek D and others. ‘Psychological distress from early adulthood to early old age: evidence from the 1946, 1958 and 1970 British birth cohorts’ Psychological Medicine 2022: volume 52, pages 1471-1480. https://doi.org/10.1017/S003329172000327X

142 Forbes N and others. ‘Mental Health of Children and Young People in England, 2017: Hyperactivity disorders’ NHS Digital, 2018. https://files.digital.nhs.uk/C7/874005/MHCYP%202017%20Hyperactivity%20Disorde rs.pdf

143 Sadler K and others. ‘Mental health of children and young people in England, 2017: Trends and characteristics’ NHS Digital, 2018. https://files.digital.nhs.uk/A0/273EE3/MHCYP%202017%20Trends%20Characteristic s.pdf

144 Taxiarchi VP and others. ‘Emotional disorders among children and young people in England from 2004 to 2017: analysis of a probability sample survey series’ British Journal of Psychiatry 2026: volume 229, pages 36-42. https://doi.org/10.1192/bjp.2025.124

145 Lutz NM and others. ‘Why is non-suicidal self-injury more common in women? Mediation and moderation analyses of psychological distress, emotion dysregulation, and impulsivity’ Archives of Suicide Research 2023: volume 27, pages 905-921. https://doi.org/10.1080/13811118.2022.2084004

146 Armitage JM and others. ‘Validation of the Strengths and Difficulties Questionnaire (SDQ) emotional subscale in assessing depression and anxiety across development’ Plos One 2023: volume 18, page e0288882. https://doi.org/10.1371/journal.pone.0288882

147 Amos R and others. ‘Mental health, social adversity, and health-related outcomes in sexual minority adolescents: a contemporary national cohort study’

Lancet Child & Adolescent Health 2020: volume 4, pages 36-45. https://doi.org/10.1016/S2352-4642(19)30339-6

148 Irish M and others. ‘Depression and self-harm from adolescence to young adulthood in sexual minorities compared with heterosexuals in the UK: a population- based cohort study’ Lancet Child & Adolescent Health 2019: volume 3, pages 91-98. https://doi.org/10.1016/S2352-4642(18)30343-2

149 Patalay P and Fitzsimons E. ‘Psychological distress, self-harm and attempted suicide in UK 17-year olds: prevalence and sociodemographic inequalities’ British Journal of Psychiatry 2021: volume 219, pages 437-439. https://doi.org/10.1192/bjp.2020.258

150 Sellers R and others. ‘Cross-cohort change in adolescent outcomes for children with mental health problems’ Journal of Child Psychology and Psychiatry 2019: volume 60, pages 813-821. https://doi.org/10.1111/jcpp.13029

151 Thompson EJ and others. ‘Changes in the adult consequences of adolescent mental ill-health: findings from the 1958 and 1970 British birth cohorts’ Psychological Medicine 2023: volume 53, pages 1074-1083. https://doi.org/10.1017/S0033291721002506

152 Goodman A and others. ‘The long shadow cast by childhood physical and mental problems on adult life’ Proceedings of the National Academy of Sciences of the United States of America 2011: volume 108, pages 6032-6037. https://doi.org/10.1073/pnas.1016970108

153 Copeland We and others. ‘Adult functional outcomes of common childhood psychiatric problems: a prospective, longitudinal study’ JAMA Psychiatry 2015: volume 72, pages 892-899. https://doi.org/10.1001/jamapsychiatry.2015.0730

154 Dykxhoorn J and others. ‘Common mental disorders in young adults: temporal trends in primary care episodes and self-reported symptoms’ BMJ Mental Health 2025: volume 28, page e301457. https://doi.org/10.1136/bmjment-2024-301457

155 Mathews F and others. ‘Children and young people's reported contact with professional services for mental health concerns: a secondary data analysis’ European Child & Adolescent Psychiatry 2024: volume 33, pages 2647-2655. https://doi.org/10.1007/s00787-023-02328-z

156 Trethewey SP and others. ‘Socio-demographic and clinical characteristics associated with mental health-related support and service contact in children and

young people aged 5-16 in England’ European Child & Adolescent Psychiatry 2025: volume 34, pages 2599-2608. https://doi.org/10.1007/s00787-025-02666-0

157 Ridout K and others. ‘Eating disorders’. In: Morris S and others, editors. ‘Adult Psychiatric Morbidity Survey: Survey of Mental Health and Wellbeing, England, 2023/4’  NHS England, 2025. https://digital.nhs.uk/data-and- information/publications/statistical/adult-psychiatric-morbidity-survey/survey-of- mental-health-and-wellbeing-england-2023-24/eating-disorders

158 Treasure J and others. ‘Eating disorders’ Lancet 2020: volume 395, pages 899- 911. https://doi.org/10.1016/S0140-6736(20)30059-3

159 Fear NT and others. ‘Posttraumatic stress disorder’. In: McManus S and others, editors. ‘Mental Health and Wellbeing in England: Adult Psychiatric Morbidity Survey 2014’ NHS Digital, 2016. https://files.digital.nhs.uk/pdf/q/3/mental_health_and_wellbeing_in_england_full_repo rt.pdf

160 Wilson H and others. ‘Posttraumatic stress disorder’. In: Morris S and others, editors. ‘Adult Psychiatric Morbidity Survey: Survey of Mental Health and Wellbeing, England, 2023/4’ NHS England, 2025. https://digital.nhs.uk/data-and- information/publications/statistical/adult-psychiatric-morbidity-survey/survey-of- mental-health-and-wellbeing-england-2023-24/posttraumatic-stress-disorder

161 Young E and others. ‘Alcohol dependence’. In: Morris S and others, editors. ‘Adult Psychiatric Morbidity Survey: Survey of Mental Health and Wellbeing, England, 2023/4’ NHS England, 2025. https://digital.nhs.uk/data-and- information/publications/statistical/adult-psychiatric-morbidity-survey/survey-of- mental-health-and-wellbeing-england-2023-24/alcohol-dependence

162 Sadler K and Bebbington P. ‘Psychosis’. In: McManus S and others, editors. ‘Adult psychiatric morbidity in England, 2007: Results of a household survey’ NHS Information Centre for Health and Social Care, 2009. https://doc.ukdataservice.ac.uk/doc/6379/mrdoc/pdf/6379_apms_2007_research_rep ort.pdf

163 Bebbington P and others. ‘Psychotic disorder’. In: McManus S and others, editors. ‘Mental Health And Wellbeing in England: Adult Psychiatric Morbidity Survey 2014’ NHS Digital, 2016. https://files.digital.nhs.uk/pdf/q/3/mental_health_and_wellbeing_in_england_full_repo rt.pdf

164 Moran P and others. ‘Personality disorder’. In: McManus S and others, editors. ‘Mental Health and Wellbeing in England: Adult Psychiatric Morbidity Survey 2014’ NHS Digital, 2016. https://files.digital.nhs.uk/pdf/q/3/mental_health_and_wellbeing_in_england_full_repo rt.pdf

165 Clery E and others. ‘Personality disorder’. In: Morris S and others, editors. ‘Adult Psychiatric Morbidity Survey: Survey of Mental Health and Wellbeing, England, 2023/4’ NHS England, 2025. https://digital.nhs.uk/data-and- information/publications/statistical/adult-psychiatric-morbidity-survey/survey-of- mental-health-and-wellbeing-england-2023-24/personality-disorder

166 Charlson FJ and others. ‘Global epidemiology and burden of schizophrenia: findings from the Global Burden of Disease Study 2016’ Schizophrenia Bulletin 2018: volume 44, pages 1195-1203. https://doi.org/10.1093/schbul/sby058

167 Jin H and Mosweu I. ‘The societal cost of schizophrenia: a systematic review’ Pharmacoeconomics 2017: volume 35, pages 25-42. https://doi.org/10.1007/s40273- 016-0444-6

168 Knapp M. ‘Hidden costs of mental illness’ British Journal of Psychiatry 2003: volume 183, pages 477-478. https://doi.org/10.1192/bjp.183.6.477

169 Hoppe s and others. ‘Reducing the delay in the diagnosis of bipolar disorder: a qualitative study’ Health Expectations 2025: volume 28, page e70398. https://doi.org/10.1111/hex.70398

170 Solmi M and others. ‘Age at onset of mental disorders worldwide: large-scale meta-analysis of 192 epidemiological studies’ Molecular Psychiatry 2022: volume 27, pages 281-295. https://doi.org/10.1038/s41380-021-01161-7

171 Ford T and others. ‘The predictors of persistent DSM-IV disorders in 3-year follow-ups of the British Child and Adolescent Mental Health Surveys 1999 and 2004’ Psychological Medicine 2017: volume 47, pages 1126-1137. https://doi.org/10.1017/S0033291716003214

172 Schlechter P and others. ‘How do parent-teacher reporting patterns predict adolescent depression and school problems? a response surface analysis’ PsyArXiv 2026. https://doi.org/10.31234/osf.io/2xy5g_v3

173 Izett E and others. ‘Prevention of mental health difficulties for children aged 0-3 years: a review’ Frontiers in Psychology 2020: volume 11, page 500361. https://doi.org/10.3389/fpsyg.2020.500361

174 Barger BD and others. ‘Prevalence and onset of regression within autism spectrum disorders: a meta-analytic review’ Journal of Autism and Developmental Disorders 2013: volume 43, pages 817-828. https://doi.org/10.1007/s10803-012- 1621-x

175 Ploubidis GB and others. ‘Association of early-life mental health with biomarkers in midlife and premature mortality: evidence from the 1958 British Birth Cohort’ JAMA Psychiatry 2021: volume 78, pages 38-46. https://doi.org/10.1001/jamapsychiatry.2020.2893

176 Ahmad G and others. ‘Prevalence of common mental disorders and treatment receipt for people from ethnic minority backgrounds in England: repeated cross- sectional surveys of the general population in 2007 and 2014’ British Journal of Psychiatry 2022: volume 221, pages 520-527. https://doi.org/10.1192/bjp.2021.179

177 Eke H and others. ‘Transition between child and adult services for young people with attention-deficit hyperactivity disorder (ADHD): findings from a British national surveillance study’ British Journal of Psychiatry 2020: volume 217, pages 616-622. https://doi.org/10.1192/bjp.2019.131

178 Chakraborty A and others. ‘Mental health of the non-heterosexual population of England’ British Journal of Psychiatry 2011: volume 198, pages 143-148. https://doi.org/10.1192/bjp.bp.110.082271

179 Semlyen J and others. ‘Sexual orientation and symptoms of common mental disorder or low wellbeing: combined meta-analysis of 12 UK population health surveys’ BMC Psychiatry 2016: volume 16, page 67. https://doi.org/10.1186/s12888- 016-0767-z

180 Pitman A and others. ‘The mental health of lesbian, gay, and bisexual adults compared with heterosexual adults: results of two nationally representative English household probability samples’ Psychological Medicine 2022: volume 52, pages 3402-3411. https://doi.org/10.1017/S0033291721000052

181 Bai Y and others. ‘Long-term trends in mental health disparities across sexual orientations in the UK: a longitudinal analysis (2010-2021)’ Social Psychiatry and Psychiatric Epidemiology 2025: volume 60, pages 1187-1196. https://doi.org/10.1007/s00127-024-02751-w

182 Tromans SJ and others. ‘Trends in self-reported autism among adults in England: analysis of a repeated cross-sectional patient survey series of 5,999,433 adults’ Journal of Autism and Developmental Disorders 2026. https://doi.org/10.1007/s10803-026-07421-1

183 Warrier V and others. ‘Elevated rates of autism, other neurodevelopmental and psychiatric diagnoses, and autistic traits in transgender and gender-diverse individuals’ Nature Communications 2020: volume 11, page 3959. https://doi.org/10.1038/s41467-020-17794-1

184 Pierce M and others. ‘Understanding drivers of recent trends in young people’s mental health’ Youth Futures Foundation, 2025. https://youthfuturesfoundation.org/wp-content/uploads/2025/07/Understanding- drivers-of-recent-trends-in-young-peoples-mental-health-July-2025-final.pdf

185 MBRRACE-UK. ‘Saving lives, improving mothers’ care: state of the nation report’ National Perinatal Epidemiology Unit, University of Oxford, 2025. https://www.hqip.org.uk/wp-content/uploads/2025/09/Ref.-578-MNI-CORP-Maternal- SON-Report-2025_FINAL.pdf

186 Armitage JM and others. ‘Characteristics of children with a psychiatric disorder in 1999, 2004 and 2017: an analysis of the national child mental health surveys of England’ Journal of Child Psychology and Psychiatry 2025: volume 66, pages 167- 177. https://doi.org/10.1111/jcpp.14040

187 Collishaw S. ‘Annual research review: Secular trends in child and adolescent mental health’ Journal of Child Psychology and Psychiatry 2015: volume 56, pages 370-393. https://doi.org/10.1111/jcpp.12372

188 Tabor E and others. ‘Mental health in higher education students and non- students: evidence from a nationally representative panel study’ Social Psychiatry and Psychiatric Epidemiology 2021: volume 56, pages 879-882. https://doi.org/10.1007/s00127-021-02032-w

189 McCloud T and others. ‘The association between higher education attendance and common mental health problems among young people in England: evidence from two population-based cohorts’ Lancet Public Health 2023: volume 8, pages e811-e819. https://doi.org/10.1016/S2468-2667(23)00188-3

190 Parker C and others. ‘Are children with unrecognised psychiatric disorders being excluded from school? A secondary analysis of the British Child and Adolescent

Mental Health Surveys 2004 and 2007’ Psychological Medicine 2019: volume 49, pages 2561-2572. https://doi.org/10.1017/S0033291718003513

191 Mathews F and others. ‘Teachers' concerns about pupils' mental health in a cross-sectional survey of a population sample of British schoolchildren’ Child and Adolescent Mental Health 2021: volume 26, pages 99-105. https://doi.org/10.1111/camh.12390

192 Committee of Public Accounts. ‘Progress in improving NHS mental health services’ House of Commons, 2023. https://publications.parliament.uk/pa/cm5803/cmselect/cmpubacc/1000/report.html

193 Bebbington PE and others. ‘The mental health of ex-prisoners: analysis of the 2014 English National Survey of Psychiatric Morbidity’ Social Psychiatry and Psychiatric Epidemiology 2021: volume 56, pages 2083-2093. https://doi.org/10.1007/s00127-021-02066-0

194 Chitsabesan P and others. ‘Mental health needs of young offenders in custody and in the community’ British Journal of Psychiatry 2006: volume 188, pages 534- 540. https://doi.org/10.1192/bjp.bp.105.010116

195 Ford T and others. ‘Psychiatric disorder among British children looked after by local authorities: comparison with children living in private households’ British Journal of Psychiatry 2007: volume 190, pages 319-325. https://doi.org/10.1192/bjp.bp.106.025023

196 Chilman N and others. ‘The public health significance of prior homelessness: findings on multimorbidity and mental health from a nationally representative survey’ Epidemiology and Psychiatric Sciences 2024: volume 33, page e63. https://doi.org/10.1017/S2045796024000659

197 McManus S and others. ‘Intimate partner violence, suicidality, and self-harm: a probability sample survey of the general population in England’ Lancet Psychiatry 2022: volume 9, pages 574-583. https://doi.org/10.1016/S2215-0366(22)00151-1

198 Harris J and others. ‘Health, mental health and housing conditions in England’ National Centre for Social Research, 2010. https://www.fuelpovertylibrary.info/sites/default/files/EAGA14%20%282010%29%20 %20REPORT%20Health%20mental%20health%20and%20housing%20conditions% 20in%20England.pdf

199 McManus S and others. ‘Joining the dots: the combined burden of violence, abuse and poverty in the lives of women’ Agenda, 2016. https://www.agendaalliance.org/documents/121/Joining_The_Dots_Report.pdf

200 Cooper SA and others. ‘Mental ill-health in adults with intellectual disabilities: prevalence and associated factors’ British Journal of Psychiatry 2007: volume 190, pages 27-35. https://doi.org/10.1192/bjp.bp.106.022483

201 Tromans SJ and others. ‘Primary care experiences of adults reporting learning disability: a probability sample survey’ British Journal of Psychiatry 2024: volume 74, pages e845-e853. https://doi.org/10.3399/BJGP.2024.0056

202 Mason J and Scior K. ‘‘Diagnostic overshadowing’ amongst clinicians working with people with intellectual disabilities in the UK’ Journal of Applied Research in Intellectual Disabilities 2004: volume 17, pages 85-90. https://doi.org/10.1111/j.1360- 2322.2004.00184.x

203 NHS England. ‘Report of the independent ADHD Taskforce: Part 1’ 2025. https://www.england.nhs.uk/wp-content/uploads/2025/06/PRN02031-interim-report- of-the-independent-adhd-taskforce-part-1.pdf

204 Healthwatch England. ‘Recognising ADHD: how to improve support for people who need it’ 2025. https://nds.healthwatch.co.uk/sites/default/files/reports_library/20250523%20ADHD %20report.pdf

205 Ridout K and others. ‘Attention deficit hyperactivity disorder’. In: Morris S and others, editors. ‘Adult Psychiatric Morbidity Survey: Survey of Mental Health and Wellbeing, England, 2023/4’ NHS England, 2025. https://digital.nhs.uk/data-and- information/publications/statistical/adult-psychiatric-morbidity-survey/survey-of- mental-health-and-wellbeing-england-2023-24/attention-deficit-hyperactivity-disorder

206 Ayano G and others. ‘The global prevalence of attention deficit hyperactivity disorder in children and adolescents: an umbrella review of meta-analyses’ Journal of Affective Disorders 2023: volume 339, pages 860-866. https://doi.org/10.1016/j.jad.2023.07.071

207 Popit S and others. ‘Prevalence of attention-deficit hyperactivity disorder (ADHD): systematic review and meta-analysis’ European Psychiatry 2024: volume 67, page e68. https://doi.org/10.1192/j.eurpsy.2024.1786

208 Salari N and others. ‘The global prevalence of ADHD in children and adolescents: a systematic review and meta-analysis’ Italian Journal of Pediatrics 2023: volume 49, page 48. https://doi.org/10.1186/s13052-023-01456-1

209 National Institute for Health and Care Excellence. ‘Attention deficit hyperactivity disorder: how common is it?’ NICE Clinical Knowledge Summary, 2025. https://cks.nice.org.uk/topics/attention-deficit-hyperactivity-disorder/background- information/prevalence/

210 Martin AF and others. ‘The changing prevalence of ADHD? A systematic review’ Journal of Affective Disorders 2025: volume 388, page 119427. https://doi.org/10.1016/j.jad.2025.119427

211 Toffalini E and others. ‘Categories or dimensions? A critical systematic review of taxometric evidence in autism, ADHD, and language disorders’ PsyArXiv 2026. https://doi.org/10.31234/osf.io/eszg6_v1

212 Arildskov TW and others. ‘How much impairment is required for ADHD? No evidence of a discrete threshold’ Journal of Child Psychology and Psychiatry 2022: volume 63, pages 229-237. https://doi.org/10.1111/jcpp.13440

213 Epstein JN and Loren RE. ‘Changes in the definition of ADHD in DSM-5: Subtle but important’ Neuropsychiatry 2013: volume 3, pages 455-458. https://doi.org/10.2217/npy.13.59

214 Swanson JM and others. ‘Attention-deficit hyperactivity disorder and hyperkinetic disorder’ Lancet 1998: volume 351, pages 429-433. https://doi.org/10.1016/s0140-6736(97)11450-7

215 Fan Y and others. ‘Global burden of mental disorders in 204 countries and territories, 1990-2021: results from the Global Burden of Disease Study 2021’ BMC Psychiatry 2025: volume 25, page 486. https://doi.org/10.1186/s12888-025-06932-y

216 Song P and others. ‘The prevalence of adult attention-deficit hyperactivity disorder: A global systematic review and meta-analysis’ Journal of Global Health 2021: volume 11, page 04009. https://doi.org/10.7189/jogh.11.04009

217 Stott J. ‘Personal communication; unpublished CPRD Aurum dataset analyses’ 2026.

218 Wickersham A and others. ‘Mental health and educational attainment: rneplicating diminishing associations in an England cohort’ medRxiv 2026: page 2026.2003.2020.26348881. https://doi.org/10.64898/2026.03.20.26348881

219 Nordmo M and others. ‘The diminishing association between adolescent mental disorders and educational performance from 2006-2019’ JCPP Advances 2024: volume 4, page e12239. https://doi.org/10.1002/jcv2.12239

220 McKechnie DGJ and others. ‘Attention-deficit hyperactivity disorder diagnoses and prescriptions in UK primary care, 2000-2018: population-based cohort study’ BJPsych Open 2023: volume 9, page e121. https://doi.org/10.1192/bjo.2023.512

221 Eccles J and others. ‘Autism and ADHD: does the global rise in diagnosed neurodivergence reflect increased awareness of undiagnosed cases?’ BMJ 2026: volume 392, page r2648. https://doi.org/10.1136/bmj.r2648

222 Bowden-Jones O. ‘ADHD medication and harm with a focus on risk dependence in England’ Report commissioned by and submitted to the Independent Review, 2026.

223 Moffitt TE and others. ‘Is adult ADHD a childhood-onset neurodevelopmental disorder? Evidence from a four-decade longitudinal cohort study’ American Journal of Psychiatry 2015: volume 172, pages 967-977. https://doi.org/10.1176/appi.ajp.2015.14101266

224 Soneson E and others. ‘The OxWell Student Survey, 2025 Wave’ University of Oxford, Department of Psychiatry, 2025. https://oxwell.org/

225 Tarp ME and others. ‘Changes in characteristics associated with ADHD and ASD diagnoses over time’ JAMA Psychiatry 2026. https://doi.org/10.1001/jamapsychiatry.2026.2722

226 Babinski DE and others. ‘Late adolescent and young adult outcomes of girls diagnosed with ADHD in childhood: an exploratory investigation’ Journal of Attention Disorders 2011: volume 15, pages 204-214. https://doi.org/10.1177/1087054710361586

227 De Rossi P and others. ‘Age-related clinical characteristics of children and adolescents with ADHD’ Frontiers in Psychiatry 2023: volume 14, page 1069934. https://doi.org/10.3389/fpsyt.2023.1069934

228 Owens EB and Hinshaw SP. ‘Childhood conduct problems and young adult outcomes among women with childhood attention-deficit/hyperactivity disorder (ADHD)’ Journal of Abnormal Psychology 2016: volume 125, pages 220-232. https://doi.org/10.1037/abn0000084

229 von Wirth E and others. ‘Prediction of educational attainment and occupational functioning in young adults with a childhood diagnosis of ADHD: results from the Cologne Adaptive Multimodal Treatment (CAMT) study’ Journal of Attention Disorders 2022: volume 26, pages 1018-1032. https://doi.org/10.1177/10870547211045740

230 Biederman J and others. ‘Predictors of persistent ADHD: an 11-year follow-up study’ Journal of Psychiatric Research 2011: volume 45, pages 150-155. https://doi.org/10.1016/j.jpsychires.2010.06.009

231 Adler LA and others. ‘Self-reported quality of life in adults with attention- deficit/hyperactivity disorder and executive function impairment treated with lisdexamfetamine dimesylate: a randomized, double-blind, multicenter, placebo- controlled, parallel-group study’ BMC Psychiatry 2013: volume 13, page 253. https://doi.org/10.1186/1471-244X-13-253

232 Mannuzza S and others. ‘Impact of the impairment criterion in the diagnosis of adult ADHD: 33-year follow-up study of boys with ADHD’ Journal of Attention Disorders 2011: volume 15, pages 122-129. https://doi.org/10.1177/1087054709359907

233 Kuriyan AB and others. ‘Young adult educational and vocational outcomes of children diagnosed with ADHD’ Journal of Abnormal Child Psychology 2013: volume 41, pages 27-41. https://doi.org/10.1007/s10802-012-9658-z

234 Cadman T and others. ‘Six-year follow-up study of combined type ADHD from childhood to young adulthood: Predictors of functional impairment and comorbid symptoms’ European Psychiatry 2016: volume 35, pages 47-54. https://doi.org/10.1016/j.eurpsy.2015.08.007

235 Choi WS and others. ‘The prevalence of psychiatric comorbidities in adult ADHD compared with non-ADHD populations: A systematic literature review’ Plos One 2022: volume 17, page e0277175. https://doi.org/10.1371/journal.pone.0277175

236 Njardvik U and others. ‘Psychiatric comorbidity in children and adolescents with ADHD: a systematic review and meta-analysis’ Clinical Psychology Review 2025: volume 118, page 102571. https://doi.org/10.1016/j.cpr.2025.102571

237 Larson K and others. ‘Patterns of comorbidity, functioning, and service use for US children with ADHD, 2007’ Pediatrics 2011: volume 127, pages 462-470. https://doi.org/10.1542/peds.2010-0165

238 Reale L and others. ‘Comorbidity prevalence and treatment outcome in children and adolescents with ADHD’ European Child & Adolescent Psychiatry 2017: volume 26, pages 1443-1457. https://doi.org/10.1007/s00787-017-1005-z

239 Kent KM and others. ‘The academic experience of male high school students with ADHD’ Journal of Abnormal Child Psychology 2011: volume 39, pages 451-462. https://doi.org/10.1007/s10802-010-9472-4

240 Caye A and others. ‘Predictors of persistence of ADHD into adulthood: a systematic review of the literature and meta-analysis’ European Child & Adolescent Psychiatry 2016: volume 25, pages 1151-1159. https://doi.org/10.1007/s00787-016- 0831-8

241 Agnew-Blais JC and others. ‘Evaluation of the persistence, remission, and emergence of attention-deficit/hyperactivity disorder in young adulthood’ JAMA Psychiatry 2016: volume 73, pages 713-720. https://doi.org/10.1001/jamapsychiatry.2016.0465

242 Shakeel HM and others. ‘Towards clinically actionable AI in psychiatry for risk stratification of adult ADHD using explainable clustering’ Manuscript submitted for publication 2026.

243 Young S and others. ‘A meta-analysis of the prevalence of attention deficit hyperactivity disorder in incarcerated populations’ Psychological Medicine 2015: volume 45, pages 247-258. https://doi.org/10.1017/S0033291714000762

244 Abdelnour E and others. ‘ADHD diagnostic trends: Increased recognition or overdiagnosis?’ Missouri Medicine 2022: volume 119, pages 467-473. https://pmc.ncbi.nlm.nih.gov/articles/PMC9616454/

245 Branigan K. ‘What is the quality of ADHD content on TikTok, and how might this affect self-diagnosis? A systematic literature review’ Educational Psychology in Practice 2026: volume 42, pages 1-17. https://doi.org/10.1080/02667363.2026.2644889

246 Timimi S. ‘Debate: Are we over-pathologising young people’s mental health? You’d be in denial if you think we aren't’ Child and Adolescent Mental Health 2025: volume 30, pages 296-298. https://doi.org/10.1111/camh.70015

247 Underhill R and Foulkes L. ‘Self-diagnosis of mental disorders: a qualitative study of attitudes on Reddit’ Qualitative Health Research 2025: volume 35, pages 779-792. https://doi.org/10.1177/10497323241288785

248 Layton TJ and others. ‘Attention deficit-hyperactivity disorder and month of school enrollment’ New England Journal of Medicine 2018: volume 379, pages 2122- 2130. https://doi.org/10.1056/NEJMoa1806828

249 Johnson J and others. ‘Misdiagnosis and missed diagnosis of adult attention- deficit hyperactivity disorder’ BJPsych Advances 2021: volume 27, pages 60-61. https://doi.org/10.1192/bja.2020.34

250 Kazda L and others. ‘Overdiagnosis of attention-deficit/hyperactivity disorder in children and adolescents: a systematic scoping review’ JAMA Network Open 2021: volume 4, page e215335. https://doi.org/10.1001/jamanetworkopen.2021.5335

251 Chief Medical Officer for England. ‘The health of people in prison, on probation and in the secure NHS estate in England’ Department of Health and Social Care and Ministry of Justice, 2025. https://www.gov.uk/government/publications/the-health-of- people-in-prison-on-probation-and-in-the-secure-nhs-estate-in-england

252 Chitsabesan P. ‘Children and young people in the secure estate’ Report commissioned by and submitted to the Independent Review, 2026.

253 Khan L and others. ‘Out of sight: girls in the children and young people’s secure estate’ 2021. https://www.centreformentalhealth.org.uk/wp- content/uploads/2021/10/CentreforMentalHealth_OutOfSight_PDF_4.pdf

254 Hancock S. ‘Delivering the best for girls in custody’ Ministry of Justice, 2025. https://assets.publishing.service.gov.uk/media/67c6acc516dc9038974dbdf3/Girls_Pl acement_Review_Report_FINAL_002.pdf

255 Hechtman L and others. ‘Functional adult outcomes 16 years after childhood diagnosis of attention-deficit/hyperactivity disorder: MTA results’ Journal of the American Academy of Child and Adolescent Psychiatry 2016: volume 55, pages 945-952 e942. https://doi.org/10.1016/j.jaac.2016.07.774

256 Groenman AP and others. ‘An individual participant data meta-analysis: behavioral treatments for children and adolescents with attention-deficit/hyperactivity disorder’ Journal of the American Academy of Child and Adolescent Psychiatry 2022: volume 61, pages 144-158. https://doi.org/10.1016/j.jaac.2021.02.024

257 Gosling CJ and others. ‘Benefits and harms of ADHD interventions: umbrella review and platform for shared decision making’ BMJ 2025: volume 391, page e085875. https://doi.org/10.1136/bmj-2025-085875

258 Rimestad ML and others. ‘Short- and long-term effects of parent training for preschool children with or at risk of ADHD: a systematic review and meta-analysis’ Journal of Attention Disorders 2019: volume 23, pages 423-434. https://doi.org/10.1177/1087054716648775

259 Lange AM and others. ‘Parent training for preschool ADHD in routine, specialist care: a randomized controlled trial’ Journal of the American Academy of Child and Adolescent Psychiatry 2018: volume 57, pages 593-602. https://doi.org/10.1016/j.jaac.2018.04.014

260 National Institute for Health and Care Excellence. ‘Attention deficit hyperactivity disorder: diagnosis and management (NG87)’ 2018. https://www.nice.org.uk/guidance/ng87

261 Faraone SV and others. ‘The World Federation of ADHD International Consensus Statement: 208 evidence-based conclusions about the disorder’ Neuroscience and Biobehavioral Reviews 2021: volume 128, pages 789-818. https://doi.org/10.1016/j.neubiorev.2021.01.022

262 Li L and others. ‘ADHD pharmacotherapy and mortality in individuals with ADHD’ JAMA 2024: volume 331, pages 850-860. https://doi.org/10.1001/jama.2024.0851

263 Hollis C and others. ‘Methylphenidate and the risk of psychosis in adolescents and young adults: a population-based cohort study’ Lancet Psychiatry 2019: volume 6, pages 651-658. https://doi.org/10.1016/S2215-0366(19)30189-0

264 National Institute for Health and Care Excellence. ‘Methylphenidate’ NICE Clinical Knowledge Summaries, 2025. https://cks.nice.org.uk/topics/attention-deficit- hyperactivity-disorder/prescribing-information/methylphenidate/

265 Clemow DB and Walker DJ. ‘The potential for misuse and abuse of medications in ADHD: a review’ Postgraduate Medicine 2014: volume 126, pages 64-81. https://doi.org/10.3810/pgm.2014.09.2801

266 Konstenius M and others. ‘Methylphenidate for attention deficit hyperactivity disorder and drug relapse in criminal offenders with substance dependence: a 24-

week randomized placebo-controlled trial’ Addiction 2014: volume 109, pages 440- 449. https://doi.org/10.1111/add.12369

267 Matthijssen AM and others. ‘Continued benefits of methylphenidate in ADHD after 2 years in clinical practice: a randomized placebo-controlled discontinuation study’ American Journal of Psychiatry 2019: volume 176, pages 754-762. https://doi.org/10.1176/appi.ajp.2019.18111296

268 Trafford AM and others. ‘Temporal trends in primary care-recorded psychiatric diagnoses and psychotropic medication prescribing among children and young people in the UK: a population-based study’ British Journal of General Practice 2025: volume 75, pages e566-e576. https://doi.org/10.3399/bjgp.2024.0804

269 Hussey L and others. ‘Attention-deficit hyperactivity disorder in children and young persons during the COVID-19 pandemic. A temporal trends analysis of electronic heath records in Greater Manchester, England’ JCPP Advances 2026: volume 6, page e70042. https://doi.org/10.1002/jcv2.70042

270 American Psychiatric Association. ‘Diagnostic and statistical manual of mental disorders, fifth edition (DSM-5)’ American Psychiatric Publishing, Inc, 2013

271 Lord C and others. ‘The Lancet Commission on the future of care and clinical research in autism’ Lancet 2022: volume 399, pages 271-334. https://doi.org/10.1016/S0140-6736(21)01541-5

272 Fombonne E. ‘Epidemiological surveys of autism and other pervasive developmental disorders: an update’ Journal of Autism and Developmental Disorders 2003: volume 33, pages 365-382. https://doi.org/10.1023/a:1025054610557

273 Kentrou V and others. ‘Delayed autism spectrum disorder recognition in children and adolescents previously diagnosed with attention-deficit/hyperactivity disorder’ Autism 2019: volume 23, pages 1065-1072. https://doi.org/10.1177/1362361318785171

274 Leedham A and others. ‘‘I was exhausted trying to figure it out’: the experiences of females receiving an autism diagnosis in middle to late adulthood’ Autism 2020: volume 24, pages 135-146. https://doi.org/10.1177/1362361319853442

275 Maciver D and others. ‘Prevalence of neurodevelopmental differences and autism in Scottish primary schools 2018-2022’ Autism Research 2023: volume 16, pages 2403-2414. https://doi.org/10.1002/aur.3063

276 Furley K and others. ‘Developmental regression in children: current and future directions’ Cortex 2023: volume 169, pages 5-17. https://doi.org/10.1016/j.cortex.2023.09.001

277 Croen LA and others. ‘The health status of adults on the autism spectrum’ Autism 2015: volume 19, pages 814-823. https://doi.org/10.1177/1362361315577517

278 Nimmo-Smith V and others. ‘Anxiety disorders in adults with autism spectrum disorder: A population-based study’ Journal of Autism and Developmental Disorders 2020: volume 50, pages 308-318. https://doi.org/10.1007/s10803-019-04234-3

279 Liu S and others. ‘Age-related physical health of older autistic adults in Sweden: a longitudinal, retrospective, population-based cohort study’ Lancet Healthy Longevity 2023: volume 4, pages e307-e315. https://doi.org/10.1016/S2666- 7568(23)00067-3

280 Hand BN and others. ‘Prevalence of physical and mental health conditions in Medicare-enrolled, autistic older adults’ Autism 2020: volume 24, pages 755-764. https://doi.org/10.1177/1362361319890793

281 Toft G and others. ‘Assessment of educational attainment and employment among individuals with autism spectrum disorder in Denmark’ JAMA Pediatrics 2021: volume 175, pages 601-608. https://doi.org/10.1001/jamapediatrics.2021.0124

282 Stark I and others. ‘Qualification for upper secondary education in individuals with autism without intellectual disability: total population study, Stockholm, Sweden’ Autism 2021: volume 25, pages 1036-1046. https://doi.org/10.1177/1362361320975929

283 Smart R and others. ‘The who, how and what of educational outcome research for autistic students published in the last decade: A systematic quantitative literature review’ British Journal of Educational Psychology 2026: volume 96, pages 286-305. https://doi.org/10.1111/bjep.70015

284 Office for National Statistics. ‘Outcomes for disabled people in the UK: 2021’ 2022. https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/disabilit y/articles/outcomesfordisabledpeopleintheuk/2021

285 Baird G and others. ‘Prevalence of disorders of the autism spectrum in a population cohort of children in South Thames: the Special Needs and Autism

Project (SNAP)’ Lancet 2006: volume 368, pages 210-215. https://doi.org/10.1016/S0140-6736(06)69041-7

286 Loomes R and others. ‘What is the male-to-female ratio in autism spectrum disorder? A systematic review and meta-analysis’ Journal of the American Academy of Child and Adolescent Psychiatry 2017: volume 56, pages 466-474. https://doi.org/10.1016/j.jaac.2017.03.013

287 Furnier SM and others. ‘Trends over time in the prevalence of autism by adaptive and intellectual functioning levels’ Autism Research 2026: volume 19, page e70167. https://doi.org/10.1002/aur.70167

288 Shenouda J and others. ‘Prevalence and disparities in the detection of autism without intellectual disability’ Pediatrics 2023: volume 151, page e2022056594. https://doi.org/10.1542/peds.2022-056594

289 Brugha TS and others. ‘Epidemiology of autism spectrum disorders in adults in the community in England’ Archives of General Psychiatry 2011: volume 68, pages 459-465. https://doi.org/10.1001/archgenpsychiatry.2011.38

290 Tick B and others. ‘Heritability of autism spectrum disorders: a meta-analysis of twin studies’ Journal of Child Psychology and Psychiatry 2016: volume 57, pages 585-595. https://doi.org/10.1111/jcpp.12499

291 Lundström S and others. ‘Autism phenotype versus registered diagnosis in Swedish children: prevalence trends over 10 years in general population samples’ BMJ 2015: volume 350, page h1961. https://doi.org/10.1136/bmj.h1961

292 Alhadeff A and others. ‘Adults who suspect they may be autistic or have ADHD show corresponding neurodevelopmental polygenic effects ’ medRxiv 2026. https://doi.org/10.64898/2026.08.03.26359559

293 Fysh M and others. ‘Autistic traits in the UK Biobank’ Research Square 2026. https://doi.org/10.21203/rs.3.rs-10021244/v1

294 Schalbroeck R and others. ‘A large-scale comparison of formally diagnosed vs. self-identified autistic and non-autistic adults’ PsyArXiv 2026. https://doi.org/10.31234/osf.io/9kx6q_v1

295 Lai M-C. ‘Mental health challenges faced by autistic people’ Nature Human Behaviour 2023: volume 7, pages 1620-1637. https://doi.org/10.1038/s41562-023- 01718-2

296 Morris J. ‘The rapidly growing waiting lists for autism and ADHD assessments’ Nuffield Trust, 2024. https://www.nuffieldtrust.org.uk/news-item/the-rapidly-growing- waiting-lists-for-autism-and-adhd-assessments

297 Abrahamson V and others. ‘Realist evaluation of Autism ServiCe Delivery (RE- ASCeD): which diagnostic pathways work best, for whom and in what context? Findings from a rapid realist review’ BMJ Open 2021: volume 11, page e051241. https://doi.org/10.1136/bmjopen-2021-051241

298 Male I and others. ‘How can we improve the timeliness and quality of diagnostic assessment for children with possible autism? Qualitative findings and recommendations from a Realist Evaluation of Autism Service delivery in the United Kingdom’ Autism 2026: volume 30, pages 1562-1575. https://doi.org/10.1177/13623613261430914

299 Norris JE and others. ‘Post-diagnostic support for adults diagnosed with autism in adulthood in the UK: a systematic review with narrative synthesis’ Autism 2025: volume 29, pages 284-309. https://doi.org/10.1177/13623613241273073

300 Mason D and others. ‘A systematic review of what barriers and facilitators prevent and enable physical healthcare services access for autistic adults’ Journal of Autism and Developmental Disorders 2019: volume 49, pages 3387-3400. https://doi.org/10.1007/s10803-019-04049-2

301 Gellini H and Marczak M. ‘“I always knew i was different”: experiences of receiving a diagnosis of autistic spectrum disorder in adulthood—a meta- ethnographic systematic review’ Review Journal of Autism and Developmental Disorders 2024: volume 11, pages 620-639. https://doi.org/10.1007/s40489-023- 00356-8

302 Cooper K and others. ‘The impact of a positive autism identity and autistic community solidarity on social anxiety and mental health in autistic young people’ Autism 2023: volume 27, pages 848-857. https://doi.org/10.1177/13623613221118351

303 Corden K and others. ‘Personal identity after an autism diagnosis: relationships with self-esteem, mental wellbeing, and diagnostic timing’ Frontiers in Psychology 2021: volume 12, page 699335. https://doi.org/10.3389/fpsyg.2021.699335

304 French B and Cassidy S. ‘"Going through life on hard mode"—the experience of late diagnosis of autism and/or ADHD: a qualitative study’ Autism in Adulthood 2026: volume 8, pages 127-136. https://doi.org/10.1089/aut.2024.0085

305 Crane L and others. ‘Autism diagnosis in the United Kingdom: perspectives of autistic adults, parents and professionals’ Journal of Autism and Developmental Disorders 2018: volume 48, pages 3761-3772. https://doi.org/10.1007/s10803-018- 3639-1

306 Roman-Urrestarazu A and others. ‘Association of race/ethnicity and social disadvantage with autism prevalence in 7 million school children in England’ JAMA Pediatrics 2021: volume 175, page e210054. https://doi.org/10.1001/jamapediatrics.2021.0054

307 Absoud M. ‘Social determinants, inequality, and autism’ Lancet Child & Adolescent Health 2022: volume 6, pages 832-833. https://doi.org/10.1016/S2352- 4642(22)00273-5

308 Weir E and others. ‘Risk of becoming 'NEET' (not in education, employment, or training) among students with special educational needs and disabilities in England’ PsyArXiv 2026. https://doi.org/10.31234/osf.io/h4vfk_v1

309 Davies J and others. ‘Career progression for autistic people: a scoping review’ Autism 2024: volume 28, pages 2690-2706. https://doi.org/10.1177/13623613241236110

310 House of Commons Library. ‘Supporting neurodivergent people into employment’ 2025. https://commonslibrary.parliament.uk/research-briefings/cdp- 2025-0179/

311 Department for Work and Pensions. ‘The Buckland Review of Autism Employment: report and recommendations’ 2024. https://www.gov.uk/government/publications/the-buckland-review-of-autism- employment-report-and-recommendations/the-buckland-review-of-autism- employment-report-and-recommendations

312 Hughes N and others. ‘Nobody made the connection: the prevalence of neuro- disability in young people who offend’ Office of the Children’s Commissioner, 2012. https://www.childrenscommissioner.gov.uk/resource/nobody-made-the-connection/

313 Borschmann R and others. ‘The health of adolescents in detention: a global scoping review’ Lancet Public Health 2020: volume 5, pages e114-e126. https://doi.org/10.1016/S2468-2667(19)30217-8

314 Holland L and others. ‘Neurodevelopmental disorders in youth justice: a systematic review of screening, assessment and interventions’ Journal of

Experimental Criminology 2023: volume 19, pages 31-70. https://doi.org/10.1007/s11292-021-09475-w

315 Moseley RL and others. ‘Community priorities for preventing suicide in autistic people: an approach to guide policy and practice’ Autism in Adulthood 2026. https://doi.org/10.1177/25739581261415764

316 Ward JH and others. ‘Increased rates of chronic physical health conditions across all organ systems in autistic adolescents and adults’ Molecular Autism 2023: volume 14, page 35. https://doi.org/10.1186/s13229-023-00565-2

317 Baird G and others. ‘Recognising and diagnosing autism in children and young people: summary of NICE guidance’ BMJ 2011: volume 343, pages d6360-d6360. https://doi.org/10.1136/bmj.d6360

318 Charman T and others. ‘Non-ASD outcomes at 36 months in siblings at familial risk for autism spectrum disorder (ASD): a baby siblings research consortium (BSRC) study’ Autism Research 2017: volume 10, pages 169-178. https://doi.org/10.1002/aur.1669

319 Bölte S and others. ‘The Gestalt of functioning in autism spectrum disorder: Results of the international conference to develop final consensus International Classification of Functioning, Disability and Health core sets’ Autism 2019: volume 23, pages 449-467. https://doi.org/10.1177/1362361318755522

320 Arnold WM and others. ‘Camouflaging and autism: conceptualisation and methodological issues’ Autism 2026: volume 30, pages 1-16. https://doi.org/10.1177/13623613261420085

321 Bradley L and others. ‘Autistic adults’ experiences of camouflaging and its perceived impact on mental health’ Autism in Adulthood 2021: volume 3, pages 320- 329. https://doi.org/10.1089/aut.2020.0071

322 Raymaker DM and others. ‘“Having all of your internal resources exhausted beyond measure and being left with no clean-up crew”: defining autistic burnout’ Autism in Adulthood 2020: volume 2, pages 132-143. https://doi.org/10.1089/aut.2019.0079

323 Hannon BR and others. ‘The use and misuse of the Camouflaging Autistic Traits Questionnaire in autism research and clinical practice: issues, considerations, and suggestions’ Autism in Adulthood 2026. https://doi.org/10.1177/25739581261435868

324 Hughes MM and others. ‘The prevalence and characteristics of children with profound autism, 15 sites, United States, 2000-2016’ Public Health Reports 2023: volume 138, pages 971-980. https://doi.org/10.1177/00333549231163551

325 Clarke EB and others. ‘Understanding profound autism: implications for stigma and supports’ Frontiers in Psychiatry 2024: volume 15, page 1287096. https://doi.org/10.3389/fpsyt.2024.1287096

326 Zeidan J and others. ‘Global prevalence of autism: a systematic review update’ Autism Research 2022: volume 15, pages 778-790. https://doi.org/10.1002/aur.2696

327 Shattuck PT. ‘The contribution of diagnostic substitution to the growing administrative prevalence of autism in US special education’ Pediatrics 2006: volume 117, pages 1028-1037. https://doi.org/10.1542/peds.2005-1516

328 Howlin P and Magiati I. ‘Autism spectrum disorder: outcomes in adulthood’ Current Opinion in Psychiatry 2017: volume 30, pages 69-76. https://doi.org/10.1097/YCO.0000000000000308

329 Ghirardi L and others. ‘Familial and genetic associations between autism spectrum disorder and other neurodevelopmental and psychiatric disorders’ Journal of Child Psychology and Psychiatry 2021: volume 62, pages 1274-1284. https://doi.org/10.1111/jcpp.13508

330 Rutter M. ‘Environmentally mediated risks for psychopathology: research strategies and findings’ Journal of the American Academy of Child and Adolescent Psychiatry 2005: volume 44, pages 3-18. https://doi.org/10.1097/01.chi.0000145374.45992.c9

331 Caspi A and others. ‘Why psychopathology research should avoid studying one mental disorder at a time: an intergenerational and developmental evidence base for understanding “p”’ Journal of Psychopathology and Clinical Science 2026: volume 135, pages 461-494. https://doi.org/10.1037/abn0001042

332 Martin RC and others. ‘Overlapping symptoms of attention deficit hyperactivity disorder (ADHD), autism, and personality disorder in a community sample’ British Journal of Clinical Psychology 2026: volume 65, pages 508-529. https://doi.org/10.1111/bjc.70037

333 Michelini G and others. ‘Where do neurodevelopmental conditions fit in transdiagnostic psychiatric frameworks? Incorporating a new neurodevelopmental

spectrum’ World Psychiatry 2024: volume 23, pages 333-357. https://doi.org/10.1002/wps.21225

334 Simonoff E and others. ‘Psychiatric disorders in children with autism spectrum disorders: prevalence, comorbidity, and associated factors in a population-derived sample’ Journal of the American Academy of Child and Adolescent Psychiatry 2008: volume 47, pages 921-929. https://doi.org/10.1097/CHI.0b013e318179964f

335 Salazar F and others. ‘Co-occurring psychiatric disorders in preschool and elementary school-aged children with autism spectrum disorder’ Journal of Autism and Developmental Disorders 2015: volume 45, pages 2283-2294. https://doi.org/10.1007/s10803-015-2361-5

336 Crowley J and others. ‘Patterning of mental health needs and support in England: a latent class analysis of the Adult Psychiatric Morbidity Survey.’ Mind, 2019.

337 Fleming M and others. ‘Neurodevelopmental multimorbidity and educational outcomes of Scottish schoolchildren: a population-based record linkage cohort study’ PLoS Medicine 2020: volume 17, page e1003290. https://doi.org/10.1371/journal.pmed.1003290

338 John A and others. ‘Association of school absence and exclusion with recorded neurodevelopmental disorders, mental disorders, or self-harm: a nationwide, retrospective, electronic cohort study of children and young people in Wales, UK’ Lancet Psychiatry 2022: volume 9, pages 23-34. https://doi.org/10.1016/S2215- 0366(21)00367-9

339 Stark I and others. ‘Completion of upper secondary mainstream school in autistic students in Sweden’ Journal of Autism and Developmental Disorders 2025: volume 55, pages 3622-3630. https://doi.org/10.1007/s10803-024-06470-8

340 Helgesson M and others. ‘Sustainable labour market participation among working young adults with diagnosed attention deficit/hyperactivity disorder (ADHD)’ SSM Population Health 2023: volume 23, page 101444. https://doi.org/10.1016/j.ssmph.2023.101444

341 Virtanen M and others. ‘Neurodevelopmental disorders among young adults and the risk of sickness absence and disability pension: a nationwide register linkage study’ Scandinavian Journal of Work, Environment & Health 2020: volume 46, pages 410-416. https://doi.org/10.5271/sjweh.3888

342 Jangmo A and others. ‘Attention-deficit/hyperactivity disorder and occupational outcomes: the role of educational attainment, comorbid developmental disorders, and intellectual disability’ Plos One 2021: volume 16, page e0247724. https://doi.org/10.1371/journal.pone.0247724

343 Bury SM and others. ‘Employment profiles of autistic people: an 8-year longitudinal study’ Autism 2024: volume 28, pages 2322-2333. https://doi.org/10.1177/13623613231225798

344 Chandler S and others. ‘Pathways to adaptive functioning in autism from early childhood to adolescence’ Autism Research 2022: volume 15, pages 1883-1893. https://doi.org/10.1002/aur.2785

345 Yerys BE and others. ‘Co-occurring ADHD symptoms in autistic adults are associated with less independence in daily living activities and lower subjective quality of life’ Autism 2022: volume 26, pages 2188-2195. https://doi.org/10.1177/13623613221112198

346 Sikora DM and others. ‘Attention-deficit/hyperactivity disorder symptoms, adaptive functioning, and quality of life in children with autism spectrum disorder’ Pediatrics 2012: volume 130 (Suppl 2), pages S91-S97. https://doi.org/10.1542/peds.2012-0900G

347 Ahnemark E and others. ‘Health-related quality of life and burden of illness in adults with newly diagnosed attention-deficit/hyperactivity disorder in Sweden’ BMC Psychiatry 2018: volume 18, page 223. https://doi.org/10.1186/s12888-018-1803-y

348 Sáez-Suanes GP and others. ‘Emotion regulation as a mediator between depressive symptoms and autism spectrum disorder (ASD) in adults with ASD and intellectual disabilities’ Research in Autism Spectrum Disorders 2020: volume 78, page 101654. https://doi.org/10.1016/j.rasd.2020.101654

349 Oakley BF and others. ‘How do core autism traits and associated symptoms relate to quality of life? Findings from the Longitudinal European Autism Project’ Autism 2021: volume 25, pages 389-404. https://doi.org/10.1177/1362361320959959

350 Arim RG and others. ‘Children with neurodevelopmental disorders and disabilities: a population-based study of healthcare service utilization using administrative data’ Developmental Medicine & Child Neurology 2017: volume 59, pages 1284-1290. https://doi.org/10.1111/dmcn.13557

351 Sun S and others. ‘Association of psychiatric comorbidity with the risk of premature death among children and adults with attention-deficit/hyperactivity disorder’ JAMA Psychiatry 2019: volume 76, pages 1141-1149. https://doi.org/10.1001/jamapsychiatry.2019.1944

352 Hwang YIJ and others. ‘Mortality and cause of death of Australians on the autism spectrum’ Autism Research 2019: volume 12, pages 806-815. https://doi.org/10.1002/aur.2086

353 Reiss S, Levitan, G. W., & Szyszko, J. . ‘Emotional disturbance and mental retardation: diagnostic overshadowing’ American Journal of Mental Deficiency 1982: volume 86, pages 567-574.

354 Dell’Armo K and Tasse MJ. ‘Diagnostic overshadowing of psychological disorders in people with intellectual disability: a systematic review’ American Journal on Intellectual and Developmental Disabilities 2024: volume 129, pages 116-134. https://doi.org/10.1352/1944-7558-129.2.116

355 Russell G and others. ‘Time trends in autism diagnosis over 20 years: a UK population-based cohort study’ Journal of Child Psychology and Psychiatry 2022: volume 63, pages 674-682. https://doi.org/10.1111/jcpp.13505

356 McKechnie DGJ and others. ‘Attention-deficit hyperactivity disorder diagnoses and prescriptions in UK primary care, 2000–2018: population-based cohort study’ BJPsych Open 2023: volume 9, page e121. https://doi.org/10.1192/bjo.2023.512

357 Mutluer T and others. ‘Population-based psychiatric comorbidity in children and adolescents with autism spectrum disorder: a meta-analysis’ Frontiers in Psychiatry 2022: volume 13, page 856208. https://doi.org/10.3389/fpsyt.2022.856208

358 Hollingdale J and others. ‘Autistic spectrum disorder symptoms in children and adolescents with attention-deficit/hyperactivity disorder: a meta-analytical review’ Psychological Medicine 2020: volume 50, pages 2240-2253. https://doi.org/10.1017/S0033291719002368

359 Claussen AH and others. ‘Impact of Tourette syndrome on school measures in a nationally representative sample’ Journal of Developmental and Behavioral Pediatrics 2018: volume 39, pages 335-342. https://doi.org/10.1097/DBP.0000000000000550

360 Hirschtritt ME and others. ‘Lifetime prevalence, age of risk, and genetic relationships of comorbid psychiatric disorders in Tourette syndrome’ JAMA

Psychiatry 2015: volume 72, page 325. https://doi.org/10.1001/jamapsychiatry.2014.2650

361 Shaw KA and others. ‘Prevalence and early identification of autism spectrum disorder among children aged 4 and 8 years - Autism and Developmental Disabilities Monitoring Network, 16 sites, United States, 2022’ Morbidity and Mortality Weekly Report. Surveillance Summaries 2025: volume 74, pages 1-22. https://doi.org/10.15585/mmwr.ss7402a1

362 Norbury CF and others. ‘The impact of nonverbal ability on prevalence and clinical presentation of language disorder: evidence from a population study’ Journal of Child Psychology and Psychiatry 2016: volume 57, pages 1247-1257. https://doi.org/10.1111/jcpp.12573

363 Norbury C and others. ‘Socioeconomic disadvantage is associated with prevalence of developmental language disorders, but not rate of language or literacy growth in children from 4 to 11 years: Evidence from the Surrey Communication and Language in Education Study (Scales)’ PsyArXiv 2021. https://doi.org/10.31234/osf.io/nyv7z

364 Clegg J and others. ‘Developmental language disorder (DLD) in young people leaving care in England: a study profiling the language, literacy and communication abilities of young people transitioning from care to independence’ International Journal of Environmental Research and Public Health 2021: volume 18, page 4107. https://doi.org/10.3390/ijerph18084107

365 Sciberras E and others. ‘Language problems in children with ADHD: a community-based study’ Pediatrics 2014: volume 133, pages 793-800. https://doi.org/10.1542/peds.2013-3355

366 NHS Digital. ‘Mental Health of Children and Young People in England, 2017: Summary of key findings’ 2018. https://files.digital.nhs.uk/A6/EA7D58/MHCYP%202017%20Summary.pdf

367 Van Bergen E and others. ‘Co-occurrence and causality among ADHD, dyslexia, and dyscalculia’ Psychological Science 2025: volume 36, pages 204-217. https://doi.org/10.1177/09567976241293999

368 Piek JP and others. ‘Depressive symptomatology in child and adolescent twins with attention-deficit hyperactivity disorder and/or developmental coordination disorder’ Twin Research and Human Genetics 2007: volume 10, pages 587-596. https://doi.org/10.1375/twin.10.4.587

369 Harrowell I and others. ‘Mental health outcomes of developmental coordination disorder in late adolescence’ Developmental Medicine & Child Neurology 2017: volume 59, pages 973-979. https://doi.org/10.1111/dmcn.13469

370 Khandaker GM and others. ‘A population-based longitudinal study of childhood neurodevelopmental disorders, IQ and subsequent risk of psychotic experiences in adolescence’ Psychological Medicine 2014: volume 44, pages 3229-3238. https://doi.org/10.1017/S0033291714000750

371 Donolato E and others. ‘Research Review: Language and specific learning disorders in children and their co‐occurrence with internalizing and externalizing problems: a systematic review and meta‐analysis’ Journal of Child Psychology and Psychiatry 2022: volume 63, pages 507-518. https://doi.org/10.1111/jcpp.13536

372 Griffiths SL and Norbury C. ‘The role of low language proficiency in school absence and subsequent academic attainment over the course of primary education’ PsyArXiv 2026. https://doi.org/10.31234/osf.io/n7qe3_v1

373 Kessler RC and others. ‘The World Health Organization Adult ADHD Self-Report Scale (ASRS): a short screening scale for use in the general population’ Psychological Medicine 2005: volume 35, pages 245-256. https://doi.org/10.1017/s0033291704002892

374 Winsper C and others. ‘Associations between infant and toddler regulatory problems, childhood co-developing internalising and externalising trajectories, and adolescent depression, psychotic and borderline personality disorder symptoms’ Journal of Child Psychology and Psychiatry 2020: volume 61, pages 182-194. https://doi.org/10.1111/jcpp.13125

375 Hirvikoski T and others. ‘Individual risk and familial liability for suicide attempt and suicide in autism: a population-based study’ Psychological Medicine 2020: volume 50, pages 1463-1474. https://doi.org/10.1017/S0033291719001405

376 Lye V and others. ‘Behaviours that challenge in children with intellectual disability: systematic review and meta-analysis of pharmacological and non- pharmacological interventions’ BJPsych Open 2025: volume 11, page e256. https://doi.org/10.1192/bjo.2025.10871

377 Sheehan R and others. ‘Mental illness, challenging behaviour, and psychotropic drug prescribing in people with intellectual disability: UK population based cohort study’ BMJ 2015: volume 351, page h4326. https://doi.org/10.1136/bmj.h4326

378 National Institute for Health and Care Excellence. ‘Challenging behaviour and learning disabilities: prevention and interventions for people with learning disabilities whose behaviour challenges’ 2015. https://www.nice.org.uk/guidance/ng11

379 Green J and others. ‘Pathological demand avoidance: symptoms but not a syndrome’ Lancet Child & Adolescent Health 2018: volume 2, pages 455-464. https://doi.org/10.1016/S2352-4642(18)30044-0

380 Cederlöf M and others. ‘Nationwide population-based cohort study of psychiatric disorders in individuals with Ehlers-Danlos syndrome or hypermobility syndrome and their siblings’ BMC Psychiatry 2016: volume 16, page 207. https://doi.org/10.1186/s12888-016-0922-6

381 Quadt L and others. ‘Childhood neurodivergent traits, inflammation and chronic disabling fatigue in adolescence: a longitudinal case-control study’ BMJ Open 2024: volume 14, page e084203. https://doi.org/10.1136/bmjopen-2024-084203

382 Smythe L and others. ‘Co-occurring functional neurological disorder and autism: an exploratory study of comorbidities in a retrospective cohort study using TriNetX’ Journal of Neurology 2025: volume 272, page 653. https://doi.org/10.1007/s00415- 025-13385-6

383 Tamilson B and others. ‘The co-occurrence of functional neurological disorder and autism spectrum disorder: a systematic literature review and meta-analysis’ Cognitive Neuropsychiatry 2024: volume 29, pages 358-385. https://doi.org/10.1080/13546805.2025.2452259

384 Instanes JT and others. ‘Adult ADHD and comorbid somatic disease: a systematic literature review’ Journal of Attention Disorders 2018: volume 22, pages 203-228. https://doi.org/10.1177/1087054716669589

385 Nigg JT and others. ‘Attention-deficit/hyperactivity disorder (ADHD) and being overweight/obesity: new data and meta-analysis’ Clinical Psychology Review 2016: volume 43, pages 67-79. https://doi.org/10.1016/j.cpr.2015.11.005

386 Doherty M and others. ‘Barriers to healthcare and self-reported adverse outcomes for autistic adults: a cross-sectional study’ BMJ Open 2022: volume 12, page e056904. https://doi.org/10.1136/bmjopen-2021-056904

387 Tromans SJ and others. ‘Characteristics and primary care experiences of people who self-report as autistic: a probability sample survey of adults registered with

primary care services in England’ BMJ Open 2024: volume 14, page e081388. https://doi.org/10.1136/bmjopen-2023-081388

388 Stewart CS and others. ‘Impact of ASD traits on treatment outcomes of eating disorders in girls’ European Eating Disorders Review 2017: volume 25, pages 123- 128. https://doi.org/10.1002/erv.2497

389 Saure E and others. ‘Characteristics of autism spectrum disorders are associated with longer duration of anorexia nervosa: a systematic review and meta- analysis’ International Journal of Eating Disorders 2020: volume 53, pages 1056- 1079. https://doi.org/10.1002/eat.23259

390 Faraone SV and Wilens TE. ‘Effect of stimulant medications for attention- deficit/hyperactivity disorder on later substance use and the potential for stimulant misuse, abuse, and diversion’ Journal of Clinical Psychiatry 2007: volume 68 (Suppl 11), pages 15-22. https://www.ncbi.nlm.nih.gov/pubmed/18307377

391 Chang Z and others. ‘Stimulant ADHD medication and risk for substance abuse’ Journal of Child Psychology and Psychiatry 2014: volume 55, pages 878-885. https://doi.org/10.1111/jcpp.12164

392 Volotinen L and others. ‘Age at first attention-deficit/hyperactivity disorder diagnosis and educational outcomes’ JAMA Psychiatry 2026: volume 83, pages 620- 629. https://doi.org/10.1001/jamapsychiatry.2026.0181

393 Hughes K and others. ‘The effect of multiple adverse childhood experiences on health: a systematic review and meta-analysis’ Lancet Public Health 2017: volume 2, pages e356-e366. https://doi.org/10.1016/S2468-2667(17)30118-4

394 Ching BCF and others. ‘Research Review: The impact of the COVID-19 pandemic on the mental health of children and young people with pre-existing mental health and neurodevelopmental conditions - a systematic review and meta-analysis of longitudinal studies’ Journal of Child Psychology and Psychiatry 2025: volume 66, pages 1274-1298. https://doi.org/10.1111/jcpp.14117

395 McLaughlin KA and Sheridan MA. ‘Beyond cumulative risk: a dimensional approach to childhood adversity’ Current Directions in Psychological Science 2016: volume 25, pages 239-245. https://doi.org/10.1177/0963721416655883

396 Mazurek MO. ‘Loneliness, friendship, and well-being in adults with autism spectrum disorders’ Autism 2014: volume 18, pages 223-232. https://doi.org/10.1177/1362361312474121

397 Haruvi-Lamdan N and others. ‘Autism spectrum disorder and post-traumatic stress disorder: an unexplored co-occurrence of conditions’ Autism 2020: volume 24, pages 884-898. https://doi.org/10.1177/1362361320912143

398 Brown-Lavoie SM and others. ‘Sexual knowledge and victimization in adults with autism spectrum disorders’ Journal of Autism and Developmental Disorders 2014: volume 44, pages 2185-2196. https://doi.org/10.1007/s10803-014-2093-y

399 Beaudry G and others. ‘An updated systematic review and meta-regression analysis: mental disorders among adolescents in juvenile detention and correctional facilities’ Journal of the American Academy of Child and Adolescent Psychiatry 2021: volume 60, pages 46-60. https://doi.org/10.1016/j.jaac.2020.01.015

400 Murie M and others. ‘The mental health of young people entering custody: a systematic review and meta-analysis’ Child & Youth Care Forum 2025: volume 55, pages 21-49. https://doi.org/10.1007/s10566-025-09878-0

401 Young S and others. ‘Neurodevelopmental disorders in prison inmates: comorbidity and combined associations with psychiatric symptoms and behavioural disturbance’ Psychiatry Research 2018: volume 261, pages 109-115. https://doi.org/10.1016/j.psychres.2017.12.036

402 Bond-Taylor S and Ellis A. ‘Rural marginality and violence affecting young people: experiences from an English county’ Journal of Youth Studies 2025. https://doi.org/10.1080/13676261.2025.2571510

403 Martin A and others. ‘Exploring the prevalence of adverse childhood experiences in secure children’s home admissions’ Child & Youth Care Forum 2021: volume 51, pages 921-935. https://doi.org/10.1007/s10566-021-09660-y

404 Khan L. ‘Missed opportunities: a review of recent evidence into children and young people’s mental health’ Centre for Mental Health, 2016. https://www.centreformentalhealth.org.uk/publications/missed-opportunities/

405 Youth Endowment Fund. ‘Beyond the headlines: trends in violence affecting children’ 2024. https://youthendowmentfund.org.uk/reports/beyond-the-headlines- 2024/

406 Chitsabesan P and others. ‘The development of the comprehensive health assessment tool for young offenders within the secure estate’ Journal of Forensic Psychiatry & Psychology 2014: volume 25, pages 1-25. https://doi.org/10.1080/14789949.2014.882387

407 Hughes N and others. ‘Ensuring the rights of children with neurodevelopmental disabilities within child justice systems’ Lancet Child & Adolescent Health 2020: volume 4, pages 163-166. https://doi.org/10.1016/S2352-4642(19)30401-8

408 Gajwani R and Minnis H. ‘Double jeopardy: implications of neurodevelopmental conditions and adverse childhood experiences for child health’ European Child & Adolescent Psychiatry 2023: volume 32, pages 1-4. https://doi.org/10.1007/s00787- 022-02081-9

409 Woolgar M and Baldock E. ‘Attachment disorders versus more common problems in looked after and adopted children: comparing community and expert assessments’ Child and Adolescent Mental Health 2015: volume 20, pages 34-40. https://doi.org/10.1111/camh.12052

410 Research Units on Pediatric Psychopharmacology Autism Network. ‘Randomized, controlled, crossover trial of methylphenidate in pervasive developmental disorders with hyperactivity’ Archives of General Psychiatry 2005: volume 62, pages 1266-1274. https://doi.org/10.1001/archpsyc.62.11.1266

411 Walkup JT and others. ‘Cognitive behavioral therapy, sertraline, or a combination in childhood anxiety’ New England Journal of Medicine 2008: volume 359, pages 2753-2766. https://doi.org/10.1056/NEJMoa0804633

412 Simonoff E and others. ‘Citalopram did not significantly improve anxiety in children with autism spectrum disorder undergoing treatment for core symptoms: secondary analysis of a trial to reduce repetitive behaviors’ Journal of Child and Adolescent Psychopharmacology 2022: volume 32, pages 233-241. https://doi.org/10.1089/cap.2021.0137

413 Reddihough DS and others. ‘Effect of fluoxetine on obsessive-compulsive behaviors in children and adolescents with autism spectrum disorders: a randomized clinical trial’ JAMA 2019: volume 322, pages 1561-1569. https://doi.org/10.1001/jama.2019.14685

414 Herscu P and others. ‘The Sofia study: negative multi-center study of low dose fluoxetine on repetitive behaviors in children and adolescents with autistic disorder’ Journal of Autism and Developmental Disorders 2020: volume 50, pages 3233-3244. https://doi.org/10.1007/s10803-019-04120-y

415 Poku AC and others. ‘Adapting psychological therapies for individuals with intellectual disabilities: a systematic review’ Clinical Psychology & Psychotherapy 2026: volume 33, page e70202. https://doi.org/10.1002/cpp.70202

416 Paynter J and others. ‘How can we make therapy better for autistic adults? Autistic adults’ ratings of helpfulness of adaptations to therapy’ Autism 2025: volume 29, pages 1540-1553. https://doi.org/10.1177/13623613251313569

417 National Institute for Health and Care Excellence. ‘Autism spectrum disorder in under 19s: support and management’ 2013. https://www.nice.org.uk/guidance/cg170

418 Walters S and others. ‘A systematic review of effective modifications to cognitive behavioural therapy for young people with autism spectrum disorders’ Review Journal of Autism and Developmental Disorders 2016: volume 3, pages 137-153. https://doi.org/10.1007/s40489-016-0072-2

419 Roberts K and Rankin PM. ‘A cognitive help or hindrance? A systematic review of cognitive behavioural therapy to treat anxiety in young people with autism spectrum disorder’ Clinical Child Psychology and Psychiatry 2025: volume 30, pages 419-435. https://doi.org/10.1177/13591045251314906

420 Linden A and others. ‘Benefits and harms of interventions to improve anxiety, depression, and other mental health outcomes for autistic people: a systematic review and network meta-analysis of randomised controlled trials’ Autism 2023: volume 27, pages 7-30. https://doi.org/10.1177/13623613221117931

421 Wood JJ and others. ‘Cognitive behavioral treatments for anxiety in children with autism spectrum disorder: a randomized clinical trial’ JAMA Psychiatry 2020: volume 77, pages 474-483. https://doi.org/10.1001/jamapsychiatry.2019.4160

422 Gallego-Jiménez MG and others. ‘Comorbidity between ASD and ED: a systematic review of therapeutic and educational interventions’ Research in Autism 2026: volume 133, page 202877. https://doi.org/10.1016/j.reia.2026.202877

423 Loomes R and Bryant-Waugh R. ‘Widening the reach of family-based interventions for anorexia nervosa: autism-adaptations for children and adolescents’ Journal of Eating Disorders 2021: volume 9, page 157. https://doi.org/10.1186/s40337-021-00511-8

424 Tchanturia K and others. ‘A novel approach for autism spectrum condition patients with eating disorders: analysis of treatment cost-savings’ European Eating Disorders Review 2021: volume 29, pages 514-518. https://doi.org/10.1002/erv.2760

425 Abel KM and others. ‘Prevalence of maternal mental illness among children and adolescents in the UK between 2005 and 2017: a national retrospective cohort

analysis’ Lancet Public Health 2019: volume 4, pages e291-e300. https://doi.org/10.1016/S2468-2667(19)30059-3

426 Sweeney S and Wilson C. ‘Parental anxiety and offspring development: a systematic review’ Journal of Affective Disorders 2023: volume 327, pages 64-78. https://doi.org/10.1016/j.jad.2023.01.128

427 Sandler IN and others. ‘Long-term impact of prevention programs to promote effective parenting: lasting effects but uncertain processes’ Annual Review of Psychology 2011: volume 62, pages 299-329. https://doi.org/10.1146/annurev.psych.121208.131619

428 Oppenheim C and others. ‘Time for Parents’ Nuffield Foundation, 2022. https://www.nuffieldfoundation.org/publications/time-for-parents

429 Office for National Statistics. ‘Families and the labour market, UK: 2021’ 2021. https://www.ons.gov.uk/employmentandlabourmarket/peopleinwork/employmentand employeetypes/articles/familiesandthelabourmarketengland/2021

430 Fagan C and Norman H. ‘Which fathers are involved in caring for pre-school age children in the United Kingdom? A longitudinal analysis of the influence of work hours in employment on shared childcare arrangements in couple households’. In: Ruspini E and Crespi, I, editors. ‘Balancing work and family in a changing society: The fathers’ perspective’ Palgrave MacMillan, 2016, pages 83-98. https://doi.org/10.1057/978-1-137-53354-8_6

431 Rusu PP and others. ‘Parental stress and well-being: a meta-analysis’ Clinical Child and Family Psychology Review 2025: volume 28, pages 255-274. https://doi.org/10.1007/s10567-025-00515-9

432 Gray P and others. ‘Decline in independent activity as a cause of decline in children’s mental well-being: summary of the evidence’ Journal of Pediatrics 2023: volume 260, page 113352. https://doi.org/10.1016/j.jpeds.2023.02.004

433 Oliver CW and others. ‘There is too much traffic for Alex to walk to school, so we drive: a call to action based on a 42-year trend’ British Journal of Sports Medicine 2019: volume 53, pages 323-324. https://doi.org/10.1136/bjsports-2017-098933

434 Hu N and others. ‘Associations between overparenting and offspring's mental health: a meta-analysis of multiple moderators’ Behaviour Science 2025: volume 15, page 1235. https://doi.org/10.3390/bs15091235

435 McCoy SS and others. ‘Parenting in overdrive: a meta-analysis of helicopter parenting across multiple indices of emerging adult functioning’ Journal of Adult Development 2025: volume 32, pages 222-245. https://doi.org/10.1007/s10804-024- 09496-5

436 Ferreira IA and others. ‘Autonomy as key to healthy psychological well-being: a systematic literature review on children’s independent mobility, cognitive and socio- emotional development’ Journal of Transport & Health 2024: volume 38, page 101837. https://doi.org/10.1016/j.jth.2024.101837

437 Office for National Statistics. ‘Families and households in the UK: 2024’ 2025. https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/f amilies/bulletins/familiesandhouseholds/2024

438 Kiernan K and others. ‘Families and Inequalities’ Institute for Fiscal Studies, 2022. https://ifs.org.uk/publications/families-and-inequalities

439 Abate BB and others. ‘Adverse childhood experiences are associated with mental health problems later in life: an umbrella review of systematic review and meta-analysis’ Neuropsychobiology 2025: volume 84, pages 48-64. https://doi.org/10.1159/000542392

440 Sands A and others. ‘Long-term influences of parental divorce on offspring affective disorders: a systematic review and meta-analysis’ Journal of Affective Disorders 2017: volume 218, pages 105-114. https://doi.org/10.1016/j.jad.2017.04.015

441 Garriga A and Pennoni F. ‘The causal effects of parental divorce and parental temporary separation on childrenְ’s cognitive abilities and psychological well-being according to parental relationship quality’ Social Indicators Research 2020: volume 161, pages 963-987. https://doi.org/10.1007/s11205-020-02428-2

442 Lee D and McLanahan S. ‘Family structure transitions and child development: instability, selection, and population heterogeneity’ American Sociological Review 2015: volume 80, pages 738-763. https://doi.org/10.1177/0003122415592129

443 Wootton RE and others. ‘Intergenerational transmission of parental mental health and wellbeing traits on offspring emotional and behavioural difficulties’ medRxiv 2025: page 2025.2006. 2017.25329751. https://doi.org/10.1101/2025.06.17.25329751

444 Opoku R and others. ‘Mapping the evidence on outcomes of childhood out-of- home care: a scoping review of reviews’ Plos One 2026: volume 21, page e0325409. https://doi.org/10.1371/journal.pone.0325409

445 Herbell K and others. ‘Identifying family engagement priorities in youth residential treatment settings: a group concept mapping study’ Child and Youth Services Review 2025: volume 168, page 108037. https://doi.org/10.1016/j.childyouth.2024.108037

446 Chollet D and others. ‘The Good Childhood Report 2025’ The Children’s Society, 2025. https://www.childrenssociety.org.uk/sites/default/files/2025- 10/GCR2025%20Main%20Report.pdf

447 McPherson C and others. ‘Schools for All? Young people's experiences of alienation in the English secondary school system’ Edge Foundation, 2023. https://www.edge.co.uk/documents/349/DD0940_- _Young_Futures_Young_Lives_FINAL_JVbavJ7.pdf

448 Ofqual. ‘GCSE Reform Consultation’ 2013. https://assets.publishing.service.gov.uk/media/5a7f5f3ced915d74e622a0f8/2013-06- 11-gcse-reform-consultation-june-2013.pdf

449 Department for Education. ‘The national curriculum in England: framework document’ 2014. https://assets.publishing.service.gov.uk/media/5a7db9e9e5274a5eaea65f58/Master_ final_national_curriculum_28_Nov.pdf

450 Department for Education. ‘16 to 19 accountability measures: technical guidance’ 2016. https://assets.publishing.service.gov.uk/media/698db6ca7da91680ad7f4301/16- 18_technical_guidance_-_February_2026.pdf

451 Abdinasir K. ‘Making the grade: how education shapes young people’s mental health’ Centre for Mental Health, 2019. https://www.centreformentalhealth.org.uk/publications/making-grade/

452 HBSC. ‘Schoolwork pressure’ 2026. https://data- browser.hbsc.org/measure/schoolwork-pressure/

453 Armitage JM and others. ‘Change over time in perceived schoolwork pressure and associations with emotional problems among 11-16-year-olds: a repeat cross-

sectional study in Wales, UK’ JCPP Advances 2025: volume 5, page e70005. https://doi.org/10.1002/jcv2.70005

454 Lofstedt P and others. ‘School satisfaction and school pressure in the WHO European region and North America: an analysis of time trends (2002-2018) and patterns of co-occurrence in 32 countries’ Journal of Adolescent Health 2020: volume 66, pages S59-S69. https://doi.org/10.1016/j.jadohealth.2020.03.007

455 Baines E and Blatchford P. ‘School break times and young people’s social lives’ UCL Institute of Education, 2019. https://www.breaktime.org.uk/Publications/BaSiS- StudyResearchBriefingFinal.pdf

456 Youth Sport Trust. ‘PE and school sport: the annual report 2025’ 2025. https://www.youthsporttrust.org/media/qw5i5s4h/yst_pe_school_sport_report_2025_f inal.pdf

457 Cullinane C and Montacute R. ‘Tutoring: the new landscape’ The Sutton Trust, 2023. https://www.suttontrust.com/wp-content/uploads/2023/03/Tutoring-The-New- Landscape.pdf

458 Lu S and others. ‘Instrumented measures of sedentary behavior and physical activity are associated with depression among children and adolescents: a systematic review and dose-response meta-analysis of observational studies’ Frontiers in Psychology 2024: volume 15, page 1465974. https://doi.org/10.3389/fpsyg.2024.1465974

459 Kandola A and others. ‘Depressive symptoms and objectively measured physical activity and sedentary behaviour throughout adolescence: a prospective cohort study’ Lancet Psychiatry 2020: volume 7, pages 262-271. https://doi.org/10.1016/S2215-0366(20)30034-1

460 Department for Education. ‘Suspensions and permanent exclusions in England: 2022 to 2023’ 2024. https://www.gov.uk/government/statistics/suspensions-and- permanent-exclusions-in-england-2022-to-2023

461 Department for Education. ‘Pupil absence in schools in England: 2022 to 2023’ 2024. https://www.gov.uk/government/statistics/pupil-absence-in-schools-in-england- 2022-to-2023

462 Department for Education. ‘Timpson review of school exclusion (CP 92)’ 2019. https://data.parliament.uk/DepositedPapers/Files/DEP2019- 0527/Timpson_Review_of_School_Exclusion_May_2019.pdf

463 Department for Education. ‘Suspensions and permanent exclusions in England: 2023 to 2024’ 2025. https://www.gov.uk/government/statistics/suspensions-and- permanent-exclusions-in-england-2023-to-2024

464 Whitehouse M. ‘Illegal school exclusion in English education policy’ Emotional and Behavioural Difficulties 2022: volume 27, pages 220-230. https://doi.org/10.1080/13632752.2022.2092274

465 The Health Foundation. ‘Inequalities in absences by mental health status’ 2025. https://www.health.org.uk/evidence-hub/education/inequalities-in-absences-by- mental-health-status

466 Parker C and others. ‘School exclusion in children with psychiatric disorder or impairing psychopathology: a systematic review’ Emotional and Behavioural Difficulties 2014: volume 20, pages 229-251. https://doi.org/10.1080/13632752.2014.945741

467 Viner RM and others. ‘The impact of school absence on mental health in children and young people: analysis of an English national birth cohort’ Plos One 2026: volume 21, page e0336137. https://doi.org/10.1371/journal.pone.0336137

468 Villadsen A and others. ‘Longitudinal association of conduct and emotional problems with school exclusion and truancy: a fixed effect analysis of the UK Millennium Cohort Study’ Child and Adolescent Mental Health 2024: volume 29, pages 299-307. https://doi.org/10.1111/camh.12681

469 Children & Young People’s Mental Health Coalition. ‘Behaviour and mental health in schools’ 2023. https://cypmhc.org.uk/publications/behaviour-and-mental- health-in-schools-report/

470 Department for Education. ‘Mental health and behaviour in schools’ 2018. https://assets.publishing.service.gov.uk/media/625ee6148fa8f54a8bb65ba9/Mental_ health_and_behaviour_in_schools.pdf

471 Allen K-A and others. ‘Adolescent school belonging and mental health outcomes in young adulthood: findings from a multi-wave prospective cohort study’ School Mental Health 2024: volume 16, pages 149-160. https://doi.org/10.1007/s12310-023- 09626-6

472 Butler N and others. ‘The contributing role of family, school, and peer supportive relationships in protecting the mental wellbeing of children and adolescents’ School

Mental Health 2022: volume 14, pages 776-788. https://doi.org/10.1007/s12310-022- 09502-9

473 Donaldson C and others. ‘Mental health, bullying and school connectedness: a comparative analysis of school transition at age 11 from within the Welsh education system’ British Educational Research Journal 2024: volume 50, pages 1533-1555. https://doi.org/10.1002/berj.3985

474 Donaldson C and others. ‘Individual and school environment predictors of mental health and wellbeing across the primary-to-secondary school transition’ School Mental Health 2025: volume 17, pages 890-902. https://doi.org/10.1007/s12310-025-09776-9

475 Symonds J and others. ‘Mental health improves after transition from comprehensive school to vocational education or employment in England: a national cohort study’ Developmental Psychology 2016: volume 52, pages 652-665. https://doi.org/10.1037/a0040118

476 Shaw RJ and others. ‘Mental health of UK youth following the removal of the Education Maintenance Allowance in England: a natural experiment study using Understanding Society data’ BMJ Public Health 2025: volume 3, page e001677. https://doi.org/10.1136/bmjph-2024-001677

477 Connolly SE and others. ‘School distress and the school attendance crisis: a story dominated by neurodivergence and unmet need’ Frontiers in Psychiatry 2023: volume 14, page 1237052. https://doi.org/10.3389/fpsyt.2023.1237052

478 YMCA England & Wales. ‘YMCA calls for urgent action to reverse decline in youth services funding’ 2025. https://ymca.org.uk/ymca-calls-for-urgent-action-to- reverse-decline-in-youth-services-funding/

479 Larkham J and Ren A. ‘A long road to recovery: local authority spending on early intervention children’s services 2010/11 to 2023/24’ 2025. https://pbe.co.uk/wp- content/uploads/2025/05/acch01_report_final.pdf

480 Woodhouse J and Zayed Y. ‘Public libraries briefing paper CBP-5875’ House of Commons Library, 2021. https://researchbriefings.files.parliament.uk/documents/SN05875/SN05875.pdf

481 Chartered Institute of Public Finance and Accountancy. ‘CIPFA Public Library Statistics 2023/24’ 2024.

482 Smith G and others. ‘Stop start. Survival, decline or closure? Children’s centres in England, 2018’ The Sutton Trust, 2018. https://www.suttontrust.com/wp- content/uploads/2018/04/StopStart-FINAL.pdf

483 Abreu L and Harker R. ‘Youth services in the UK’ House of Commons Library, 2026. https://researchbriefings.files.parliament.uk/documents/CBP-10132/CBP- 10132.pdf

484 Raising the Nation Play Commission. ‘Everything to play for: a plan to ensure every child in England can play’ Raising the Nation Play Commission & Centre for Young Lives, 2025. https://playwork.foundation/wp-content/uploads/2025/06/2025- 06-11-everything-to-play-for-raising-the-nation-play-commission-report-final.pdf

485 Department for Culture, Media and Sport,. ‘Community Life Survey 2023/24 annual publication’ 2024. https://www.gov.uk/government/statistics/community-life- survey-202324-annual-publication

486 Cattan S and others. ‘The health effects of universal early childhood interventions: evidence from Sure Start’ Institute for Fiscal Studies, 2022. https://ifs.org.uk/sites/default/files/2022-10/WP202243-The-health-effects-of- universal-early-childhood-interventions-evidence-from-Sure-Start.pdf

487 Ciocanel O and others. ‘Effectiveness of positive youth development interventions: a meta-analysis of randomized controlled trials’ Journal of Youth and Adolescence 2017: volume 46, pages 483-504. https://doi.org/10.1007/s10964-016- 0555-6

488 Fahy K and others. ‘Mental health impact of cuts to local government spending on cultural, environmental and planning services in England: a longitudinal ecological study’ BMC Public Health 2023: volume 23, page 1441. https://doi.org/10.1186/s12889-023-16340-0

489 Kingsbury M and others. ‘The protective effect of neighbourhood social cohesion on adolescent mental health following stressful life events’ Psychological Medicine 2020: volume 50, pages 1292-1299. https://doi.org/10.1017/S0033291719001235

490 Holt-Lunstad J and others. ‘Loneliness and social isolation as risk factors for mortality: a meta-analytic review’ Perspectives on Psychological Science 2015: volume 10, pages 227-237. https://doi.org/10.1177/1745691614568352

491 Bedük S and Yong A. ‘Long-term childhood poverty in Britain: trends and drivers across the 1991–2017 birth cohorts’ Journal of Social Policy 2026. https://doi.org/10.1017/s0047279426101366

492 Department for Communities and Local Government. ‘English Housing Survey: private rented sector, 2015-16’ 2017. https://assets.publishing.service.gov.uk/media/5a8226c2ed915d74e340201f/Private_ rented_sector_report_2015-16.pdf

493 Office for National Statistics. ‘More adults living with their parents’ 2023. https://www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/pop ulationestimates/articles/moreadultslivingwiththeirparents/2023-05-10

494 Department for Levelling Up, Housing & Communities,. ‘English Housing Survey 2021 to 2022: private rented sector’ 2025. https://www.gov.uk/government/statistics/english-housing-survey-2021-to-2022- private-rented-sector/english-housing-survey-2021-to-2022-private-rented-sector

495 Mason KE and others. ‘The impact of housing insecurity on mental health, sleep and hypertension: analysis of the UK Household Longitudinal Study and linked data, 2009-2019’ Social Science and Medicine 2024: volume 351, page 116939. https://doi.org/10.1016/j.socscimed.2024.116939

496 Office for National Statistics. ‘EMP17: People in employment on zero hours contracts’ 2026. https://www.ons.gov.uk/employmentandlabourmarket/peopleinwork/employmentand employeetypes/datasets/emp17peopleinemploymentonzerohourscontracts

497 Broome M and others. ‘An intergenerational audit for the UK: 2023’ Resolution Foundation, 2023. https://www.resolutionfoundation.org/app/uploads/2023/11/An- intergenerational-audit-for-the-UK-2023.pdf

498 Bolton P. ‘Student loan statistics’ House of Commons Library, 2025. https://researchbriefings.files.parliament.uk/documents/SN01079/SN01079.pdf

499 Davies J and others. ‘Access to employment: a comparison of autistic, neurodivergent and neurotypical adults' experiences of hiring processes in the United Kingdom’ Autism 2023: volume 27, pages 1746-1763. https://doi.org/10.1177/13623613221145377

500 Vargas-Salas O and others. ‘Neurodivergence and the workplace: a systematic review of the literature’ Journal of Vocational Rehabilitation 2025: volume 63, pages 83-94. https://doi.org/10.1177/10522263251337564

501 Organisation for Economic Co‑operation and Development. ‘The State of Global Teenage Career Preparation’ 2025. https://www.oecd.org/en/publications/the-state- of-global-teenage-career-preparation_d5f8e3f2-en.html

502 Pulford A and others. ‘Does persistent precarious employment affect health outcomes among working age adults? A systematic review and meta-analysis’ Journal of Epidemiology and Community Health 2022: volume 76, pages 909-917. https://doi.org/10.1136/jech-2022-219292

503 Svane-Petersen AC and others. ‘Psychosocial working conditions and depressive disorder: disentangling effects of job control from socioeconomic status using a life-course approach’ Social Psychiatry and Psychiatric Epidemiology 2020: volume 55, pages 217-228. https://doi.org/10.1007/s00127-019-01769-9

504 Thern E and others. ‘Precarious employment in early adulthood and later mental health problems: a register-linked cohort study’ Journal of Epidemiology and Community Health 2023: volume 77, pages 755-761. https://doi.org/10.1136/jech- 2023-220817

505 Cribb J and others. ‘Living standards, poverty and inequality in the UK: 2024’ Institute for Fiscal Studies, 2024. https://ifs.org.uk/publications/living-standards- poverty-and-inequality-uk-2024

506 Wickham S and others. ‘The effect of a transition into poverty on child and maternal mental health: a longitudinal analysis of the UK Millennium Cohort Study’ Lancet Public Health 2017: volume 2, pages e141-e148. https://doi.org/10.1016/S2468-2667(17)30011-7

507 Cooper K and Stewart K. ‘Does household income affect children’s outcomes? A systematic review of the evidence’ Child Indicators Research 2021: volume 14, pages 981-1005. https://doi.org/10.1007/s12187-020-09782-0

508 Office for National Statistics. ‘Consumer price inflation, UK: October 2022’ 2022. https://www.ons.gov.uk/economy/inflationandpriceindices

509 Financial Conduct Authority. ‘Financial Lives 2024 survey: forbearance and debt advice’ 2024. https://www.fca.org.uk/financial-lives/financial-lives-2024

510 NSPCC. ‘The impact of financial pressures on young peopleod’ 2024. https://learning.nspcc.org.uk/research-resources/2024/financial-pressures-young- people

511 Ridge T. ‘Childhood poverty: a barrier to social participation and inclusion’. In: Tisdall K and others, editors. ‘Children, young people and social inclusion: participation for what?’ Policy Press, 2006. https://doi.org/10.1332/policypress/9781861346629.003.0002

512 Skinner GC and Kennedy E. ‘The cost of poverty for child development: the adverse impact on maltreatment, education and mental health outcomes cannot be ignored’ Clinical Child Psychology and Psychiatry 2025: volume 30, pages 309-313. https://doi.org/10.1177/13591045241302121

513 Adjei NK and others. ‘Impact of poverty and family adversity on adolescent health: a multi-trajectory analysis using the UK Millennium Cohort Study’ Lancet Regional Health – Europe 2022: volume 13, page 100279. https://doi.org/10.1016/j.lanepe.2021.100279

514 Children & Young People’s Mental Health Coalition. ‘A dual crisis: the hidden link between poverty and children's mental health’ 2024. https://cypmhc.org.uk/wp- content/uploads/2024/07/CentreforMH_ADualCrisis.final_.pdf

515 Elliot Major L and others. ‘Learning loss since lockdown: variation across the home nations’

2021. https://www.understandingsociety.ac.uk/research/publications/publication- 536878/

516 Roberts N and Danechi S. ‘Coronavirus and schools’ 2022. https://researchbriefings.files.parliament.uk/documents/CBP-8915/CBP-8915.pdf

517 Andrew A and others. ‘Learning during the lockdown: real‑time data on children’s experiences during home learning’ Institute for Fiscal Studies, 2020. https://dera.ioe.ac.uk/id/eprint/35632/1/BN288-Learning-during-the-lockdown-1.pdf

518 McOwat K and others. ‘The CLoCk study: a retrospective exploration of loneliness in children and young people during the COVID-19 pandemic, in England’ Plos One 2023: volume 18, page e0294165. https://doi.org/10.1371/journal.pone.0294165

519 Natural England. ‘The People and Nature Survey for England: Children’s Survey’ 2020. https://www.gov.uk/government/collections/the-people-and-nature- surveys-for-england

520 Sport England. ‘Active Lives Children and Young People Survey: Academic Year 2021–22’ 2022. https://sportengland-production-files.s3.eu-west- 2.amazonaws.com/s3fs-public/2022- 12/Active%20Lives%20Children%20and%20Young%20People%20Survey%20Acad emic%20Year%202021- 22%20Report.pdf?VersionId=R5_hmJHw5M4yKFsewm2vGDMRGHWW7q3E

521 Yarde J and others. ‘Covid Social Mobility & Opportunities (COSMO) Study. Wave 1 Initial Findings Briefing No. 3 – Future plans and aspirations’ UCL & Sutton Trust, 2022. https://cosmostudy.uk/publications/future-plans-and-aspirations

522 Holt‑White E and others. ‘Covid Social Mobility & Opportunities (COSMO) Study. Wave 2 Initial Findings – Briefing No. 1: Mental and physical health’ UCL & Sutton Trust, 2023. https://cls.ucl.ac.uk/wp-content/uploads/2017/02/Mental-and- physical-health-FINAL.pdf

523 Office for National Statistics. ‘Young people from disadvantaged backgrounds feel less in control of their futures’ 2023. https://www.ons.gov.uk/peoplepopulationandcommunity/educationandchildcare/articl es/youngpeoplefromdisadvantagedbackgroundsfeellessincontroloftheirfutures/2023- 11-06

524 Booth C and Fitzsimons E. ‘Early family transitions: Initial findings from the Millennium Cohort Study at Age 23’ 2026. https://cls.ucl.ac.uk/wp- content/uploads/2026/02/Early-family-transitions-Initial-findings-from-the-Millennium- Cohort-Study-at-Age-23.pdf

525 Tong M and others. ‘The longitudinal impact of the covid-19 pandemic on children’s and parents’ mental health outcomes: a systematic review’ European Journal of Developmental Psychology 2025: volume 22, pages 693-728. https://doi.org/10.1080/17405629.2025.2533160

526 Kasaven LS and others. ‘The impact of the COVID-19 pandemic on perinatal services and maternal mental health in the UK’ BJPsych Open 2023: volume 9, page e13. https://doi.org/10.1192/bjo.2022.632

527 McLeish J and others. ‘Learning from a crisis: a qualitative study of the impact on mothers’ emotional wellbeing of changes to maternity care during the COVID-19

pandemic in England, using the National Maternity Survey 2020’ BMC Pregnancy and Childbirth 2022: volume 22, page 868. https://doi.org/10.1186/s12884-022- 05208-7

528 Botting N and others. ‘Developmental outcomes in 4-year-old children born in the first COVID-19 lockdown in England: preliminary findings from the Born In Covid Year - Core Lockdown Effects (Bicycle) study’ Archives of Disease in Childhood 2026. https://doi.org/10.1136/archdischild-2025-329529

529 Bevilacqua L and others. ‘Impact of COVID-19 on the mental health of children and young people: an umbrella review’ Journal of Epidemiology and Community Health 2023: volume 77, pages 704-709. https://doi.org/10.1136/jech-2022-220259

530 Bower M and others. ‘A hidden pandemic? An umbrella review of global evidence on mental health in the time of COVID-19’ Frontiers in Psychiatry 2023: volume 14, page 1107560. https://doi.org/10.3389/fpsyt.2023.1107560

531 Duan C and others. ‘Evidence linking COVID-19 and the health/well-being of children and adolescents: an umbrella review’ BMC Medicine 2024: volume 22, page 116. https://doi.org/10.1186/s12916-024-03334-x

532 Harrison L and others. ‘Indirect effects of COVID-19 on child and adolescent mental health: an overview of systematic reviews’ BMJ Global Health 2022: volume

  • https://doi.org/10.1136/bmjgh-2022-010713

533 Witteveen AB and others. ‘COVID-19 and common mental health symptoms in the early phase of the pandemic: an umbrella review of the evidence’ PLoS Medicine 2023: volume 20, page e1004206. https://doi.org/10.1371/journal.pmed.1004206

534 NHS England. ‘Mental Health of Children and Young People in England, 2022 - wave 3 follow-up to the 2017 survey’ 2022. https://digital.nhs.uk/data-and- information/publications/statistical/mental-health-of-children-and-young-people-in- england/2022-follow-up-to-the-2017-survey

535 Newlove-Delgado T and others. ‘Mental Health of Children and Young People in England, 2023 - wave 4 follow-up to the 2017 survey’ NHS England, 2023. https://digital.nhs.uk/data-and-information/publications/statistical/mental-health-of- children-and-young-people-in-england/2023-wave-4-follow-up

536 van Oers HA and others. ‘Changes in child and adolescent mental health across the COVID-19 pandemic (2018-2023): Insights from general population and clinical

samples in the Netherlands’ JCPP Advances 2024: volume 4, page e12213. https://doi.org/10.1002/jcv2.12213

537 Kiviruusu O and others. ‘Mental health after the COVID-19 pandemic among Finnish youth: a repeated, cross-sectional, population-based study’ Lancet Psychiatry 2024: volume 11, pages 451-460. https://doi.org/10.1016/S2215- 0366(24)00108-1

538 Sourander A and others. ‘Self-reported changes in adolescent mental health, deliberate self-harm, substance use, and help-seeking behavior before and after the COVID-19 pandemic - A Finnish time-trend study’ Child and Adolescent Mental Health 2026: volume 31, pages 13-22. https://doi.org/10.1111/camh.70040

539 Newlove-Delgado T and others. ‘Annual Research Review: The impact of Covid- 19 on psychopathology in children and young people worldwide: systematic review of studies with pre- and within-pandemic data’ Journal of Child Psychology and Psychiatry 2023: volume 64, pages 611-640. https://doi.org/10.1111/jcpp.13716

540 Pellicano L and others. ‘“I want to see my friends”: The everyday experiences of autistic people and their families during COVID-19’ Sydney Policy Lab, University of Sydney, 2020. https://www.sydney.edu.au/sydney-policy-lab/news-and- analysis/latest-news/2020/11/11/the-everyday-experiences-of-autistic-people-and- their-families-d.html

541 Bobo E and others. ‘How do children and adolescents with attention deficit hyperactivity disorder (ADHD) experience lockdown during the COVID-19 outbreak?]’ Encephale 2020: volume 46, pages S85-S92. https://doi.org/10.1016/j.encep.2020.05.011

542 Sciberras E and others. ‘Physical health, media use, and mental health in children and adolescents with ADHD during the COVID-19 pandemic in Australia’ Journal of Attention Disorders 2022: volume 26, pages 549-562. https://doi.org/10.1177/1087054720978549

543 Hamilton LG and others. ‘“I’m able to function better when I know there’s a beginning and an end time”: autistic adolescents’ experiences of lockdowns during the COVID-19 pandemic’ Autism & Developmental Language Impairments 2023: volume 8, page 239694152311595. https://doi.org/10.1177/23969415231159552

544 Vincent C and others. ‘Lessons from lockdown: autistic students, parents and mainstream schools’ British Educational Research Journal 2023: volume 49, pages 522-537. https://doi.org/10.1002/berj.3854

545 Toseeb U and Asbury K. ‘A longitudinal study of the mental health of autistic children and adolescents and their parents during COVID-19: Part 1, quantitative findings’ Autism 2023: volume 27, pages 105-116. https://doi.org/10.1177/13623613221082715

546 Castle VE and others. ‘The impact of COVID-19 on the mental health and wellbeing of children with special education needs and disabilities: a systematic review’ Review Journal of Autism and Developmental Disorders 2024. https://doi.org/10.1007/s40489-024-00453-2

547 Nonweiler J and others. ‘Prevalence and associated factors of emotional and behavioural difficulties during COVID-19 pandemic in children with neurodevelopmental disorders’ Children 2020: volume 7, page 128. https://doi.org/10.3390/children7090128

548 Mullen L and others. ‘What are the effects of the COVID-19 pandemic on the development of children with special educational needs and disabilities from parents’ experiences? An integrative review’ European Journal of Special Needs Education 2025: volume 40, pages 424-439. https://doi.org/10.1080/08856257.2024.2372966

549 Garagiola ER and others. ‘Adolescent resilience during the COVID-19 pandemic: a review of the impact of the pandemic on developmental milestones’ Behavioral Sciences 2022: volume 12, page 220. https://doi.org/10.3390/bs12070220

550 Lipkin M and Crepeau-Hobson F. ‘The impact of the COVID-19 school closures on families with children with disabilities: a qualitative analysis’ Psychology in the Schools 2022: volume 60, pages 1544-1559. https://doi.org/10.1002/pits.22706

551 Paterson J and others. ‘The impact of the coronavirus pandemic on the lives of children and young people who have special educational needs and/or disabilities in the UK: a scoping review’ Journal of Research in Special Educational Needs 2024: volume 24, pages 12-24. https://doi.org/10.1111/1471-3802.12608

552 All-Party Parliamentary Group for Special Educational Needs and Disabilities. ‘Forgotten. Left behind. Overlooked. The experiences of young people with SEND and their educational transitions during the Covid-19 pandemic in 2020’ 2021. https://www.dioceseofbrentwood.net/wp-content/uploads/2021/06/APPG-for-SEND- Report-Spring-2021-1.pdf

553 Ofcom. ‘Children and parents: media use and attitudes report’ 2025. https://www.ofcom.org.uk/siteassets/resources/documents/research-and-data/media-

literacy-research/children/childrens-media-use-and-attitudes-report-2025/childrens- media-literacy-report-2025.pdf?v=396621

554 Ofcom. ‘Children’s online experiences’ 2026. https://www.ofcom.org.uk/siteassets/resources/documents/online-safety/research- statistics-and-data/protecting-children/childrens-online-experiences-research- report.pdf?v=418198

555 Cardy R and others. ‘Patterns and impact of technology use in autistic children’ Research in Autism Spectrum Disorders 2023: volume 108, page 102253. https://doi.org/10.1016/j.rasd.2023.102253

556 Valencia K and others. ‘The impact of technology on people with autism spectrum disorder: a systematic literature review’ Sensors 2019: volume 19, page 4485. https://doi.org/10.3390/s19204485

557 Winter H and O'Neill J. ‘The impact of screen time on ADHD symptoms in children and adolescents: a narrative review of treatment approaches’ Clinical Child Psychology and Psychiatry 2026: volume 31, pages 764-776. https://doi.org/10.1177/13591045251399697

558 Caton S and others. ‘Internet use for family carers of people with intellectual disabilities: a literature review and thematic synthesis’ Journal of Intellectual Disabilities 2019: volume 23, pages 446-468. https://doi.org/10.1177/1744629519874214

559 Hall CM and others. ‘Online dissemination of resources and services for parents of children with autism spectrum disorders (ASDs): a systematic review of evidence’ Review Journal of Autism and Developmental Disorders 2016: volume 3, pages 273- 285. https://doi.org/10.1007/s40489-016-0083-z

560 Kostyrka-Allchorne K and others. ‘Understanding youth online experiences and mental health: development and validation of the Digital Activity and Feelings Inventory (DAFI)’ International Journal of Methods in Psychiatric Research 2025: volume 34, page e70028. https://doi.org/10.1002/mpr.70028

561 Kostyrka-Allchorne K and others. ‘Differentiating specific digital activity as risks for adolescent mental health problems’ Unpublished work 2026.

562 Fardouly J and others. ‘Social comparisons on social media: the impact of Facebook on young women’s body image concerns and mood’ Body Image 2015: volume 13, pages 38-45. https://doi.org/10.1016/j.bodyim.2014.12.002

563 Fardouly J and Vartanian LR. ‘Social media and body image concerns: current research and future directions’ Current Opinion in Psychology 2016: volume 9, pages 1-5. https://doi.org/10.1016/j.copsyc.2015.09.005

564 Tiggemann M and Slater A. ‘NetGirls: the Internet, Facebook, and body image concern in adolescent girls’ International Journal of Eating Disorders 2013: volume 46, pages 630-633. https://doi.org/10.1002/eat.22141

565 Tiggemann M and Slater A. ‘NetTweens: the internet and body image concerns in preteenage girls’ Journal of Early Adolescence 2014: volume 34, pages 606-620. https://doi.org/10.1177/0272431613501083

566 Kelly Y and others. ‘Social media use and adolescent mental health: findings from the UK Millennium Cohort Study’ EClinicalMedicine 2019: volume 6, pages 59-

  • https://doi.org/10.1016/j.eclinm.2018.12.005

567 Yoon S and others. ‘Is social network site usage related to depression? A meta- analysis of Facebook-depression relations’ Journal of Affective Disorders 2019: volume 248, pages 65-72. https://doi.org/10.1016/j.jad.2019.01.026

568 Irmer A and Schmiedek F. ‘Associations between youth’s daily social media use and well-being are mediated by upward comparisons’ Communications Psychology 2023: volume 1, page 12. https://doi.org/10.1038/s44271-023-00013-0

569 Stothart C and others. ‘The attentional cost of receiving a cell phone notification’ Journal of Experimental Psychology: Human Perception and Performance 2015: volume 41, pages 893-897. https://doi.org/10.1037/xhp0000100

570 Upshaw JD and others. ‘The hidden cost of a smartphone: the effects of smartphone notifications on cognitive control from a behavioral and electrophysiological perspective’ Plos One 2022: volume 17, page e0277220. https://doi.org/10.1371/journal.pone.0277220

571 Skowronek J and others. ‘The mere presence of a smartphone reduces basal attentional performance’ Scientific Reports 2023: volume 13, page 9363. https://doi.org/10.1038/s41598-023-36256-4

572 Böttger T and others. ‘Does the brain drain effect really exist? A meta-analysis’ Behavioral Sciences 2023: volume 13, page 751. https://doi.org/10.3390/bs13090751

573 Ruiz Pardo AC and Minda JP. ‘Reexamining the “brain drain” effect: a replication of Ward et al. (2017)’ Acta Psychologica 2022: volume 230, page 103717. https://doi.org/10.1016/j.actpsy.2022.103717

574 Sohn SY and others. ‘Prevalence of problematic smartphone usage and associated mental health outcomes amongst children and young people: a systematic review, meta-analysis and Grade of the evidence’ BMC Psychiatry 2019: volume 19, page 356. https://doi.org/10.1186/s12888-019-2350-x

575 Ding J and others. ‘The association between problematic social media use and attention deficit/hyperactivity disorder symptomatology: a systematic review and meta-analysis’ Journal of Psychiatric Research 2025: volume 189, pages 544-553. https://doi.org/10.1016/j.jpsychires.2025.07.009

576 Nikkelen SW and others. ‘Media use and ADHD-related behaviors in children and adolescents: a meta-analysis’ Developmental Psychology 2014: volume 50, pages 2228-2241. https://doi.org/10.1037/a0037318

577 Jourdren M and others. ‘The impact of screen exposure on attention abilities in young children: a systematic review’ Pediatric Neurology 2023: volume 142, pages 76-88. https://doi.org/10.1016/j.pediatrneurol.2023.01.005

578 Thorell LB and others. ‘Longitudinal associations between digital media use and ADHD symptoms in children and adolescents: a systematic literature review’ European Child & Adolescent Psychiatry 2024: volume 33, pages 2503-2526. https://doi.org/10.1007/s00787-022-02130-3

579 Nguyen L and others. ‘Feeds, feelings, and focus: a systematic review and meta-analysis examining the cognitive and mental health correlates of short-form video use’ Psychological Bulletin 2025: volume 151, pages 1125-1146. https://doi.org/10.1037/bul0000498

580 Shou Q and others. ‘Association of screen time with attention- deficit/hyperactivity disorder symptoms and their development: the mediating role of brain structure’ Translational Psychiatry 2025: volume 15, page 447. https://doi.org/10.1038/s41398-025-03672-1

581 Zhao Y and others. ‘Brain structural co-development is associated with internalizing symptoms two years later in the ABCD cohort’ Journal of Behavioral Addictions 2023: volume 12, pages 80-93. https://doi.org/10.1556/2006.2023.00006

582 Fiore M and others. ‘Digital devices use and sleep in adolescents: an umbrella review’ International Journal of Environmental Research and Public Health 2025: volume 22, page 1517. https://doi.org/10.3390/ijerph22101517

583 LeBourgeois MK and others. ‘Digital media and sleep in childhood and adolescence’ Pediatrics 2017: volume 140, pages S92-S96. https://doi.org/10.1542/peds.2016-1758J

584 Bauducco S and others. ‘A bidirectional model of sleep and technology use: a theoretical review of how much, for whom, and which mechanisms’ Sleep Medicine Reviews 2024: volume 76, page 101933. https://doi.org/10.1016/j.smrv.2024.101933

  1. Hall JA. ‘Relating through technology’ Cambridge University Press, 2020
  2. Hall JA and Liu D. ‘Social media use, social displacement, and well-being’

Current Opinion in Psychology 2022: volume 46, page 101339. https://doi.org/10.1016/j.copsyc.2022.101339

  1. Przybylski AK and Bowes L. ‘Cyberbullying and adolescent well-being in

England: a population-based cross-sectional study’ Lancet Child & Adolescent Health 2017: volume 1, pages 19-26. https://doi.org/10.1016/S2352-4642(17)30011- 1

  1. Kowalski RM and others. ‘Bullying in the digital age: a critical review and meta-

analysis of cyberbullying research among youth’ Psychological Bulletin 2014: volume 140, pages 1073-1137. https://doi.org/10.1037/a0035618

  1. John A and others. ‘Self-harm, suicidal behaviours, and cyberbullying in children

and young people: systematic review’ Journal of Medical Internet Research 2018: volume 20, page e129. https://doi.org/10.2196/jmir.9044

  1. Hill NTM and others. ‘Association of suicidal behavior with exposure to suicide

and suicide attempt: a systematic review and multilevel meta-analysis’ PLoS Medicine 2020: volume 17, page e1003074. https://doi.org/10.1371/journal.pmed.1003074

  1. Walker A. ‘Molly Russell: Prevention of future deaths report’ Courts and

Tribunals Judiciary, 2022. https://www.judiciary.uk/prevention-of-future-death- reports/molly-russell-prevention-of-future-deaths-report/

  1. Center for Countering Digital Hate. ‘Deadly by design: how social media

companies enable online abuse’ 2022. https://counterhate.com/research/deadly-by- design/

  1. Amnesty International. ‘Driven into darkness: How TikTok’s “For You” feed

encourages self‑harm and suicidal ideation’ 2023. https://www.amnesty.org/en/documents/pol40/7350/2023/en/

  1. Fairplay. ‘Designing for disorder: Instagram’s pro‑eating disorder bubble’ 2023.

https://fairplayforkids.org/wp-content/uploads/2022/04/designing_for_disorder.pdf

  1. Craddock N and Schneider J. ‘Body image in the age of GLP-1s: emerging

questions for research and practice’ Body Image 2026: volume 57, page 102095. https://doi.org/10.1016/j.bodyim.2026.102095

  1. Suhag K and Rauniyar S. ‘Social media effects regarding eating disorders and

body image in young adolescents’ Cureus 2024: volume 16, page e58674. https://doi.org/10.7759/cureus.58674

  1. Karasavva V and others. ‘A double-edged hashtag: evaluation of #ADHD-related

TikTok content and its associations with perceptions of ADHD’ Plos One 2025: volume 20, page e0319335. https://doi.org/10.1371/journal.pone.0319335

  1. Orben A and others. ‘Mechanisms linking social media use to adolescent mental

health vulnerability’ Nature Reviews Psychology 2024: volume 3, pages 407-423. https://doi.org/10.1038/s44159-024-00307-y

  1. Bölte S and others. ‘Reconciling the seemingly irreconcilable: the Who’s ICF

system integrates biological and psychosocial environmental determinants of autism and ADHD’ Bioessays 2021: volume 43, page e2000254. https://doi.org/10.1002/bies.202000254

  1. Tam MT and others. ‘A systematic review of the impacts of media mental health

awareness campaigns on young people’ Health Promotion Practice 2024: volume 25, pages 907-920. https://doi.org/10.1177/15248399241232646

  1. Evans-Lacko S and others. ‘The double disadvantage faced by adolescents

from low socioeconomic backgrounds with mental health problems affects earnings up to mid-life’ Social Science & Medicine 2024: volume 362, page 117385. https://doi.org/10.1016/j.socscimed.2024.117385

  1. Hahn JS and others. ‘The Every Mind Matters campaign: changes in mental

health literacy and its associations with campaign awareness’ European Journal of Public Health 2023: volume 33, pages 1008-1013. https://doi.org/10.1093/eurpub/ckad145

  1. Kim SY and others. ‘Time to level up: a systematic review of interventions

aiming to reduce stigma toward autistic people’ Autism 2024: volume 28, pages 798- 815. https://doi.org/10.1177/13623613231205915

  1. Bisset M and others. ‘Recent attitudes toward ADHD in the broader community:

a systematic review’ Journal of Attention Disorders 2022: volume 26, pages 537-548. https://doi.org/10.1177/10870547211003671

  1. Henderson C and others. ‘Relationships between anti-stigma programme

awareness, disclosure comfort and intended help-seeking regarding a mental health problem’ British Journal of Psychiatry 2017: volume 211, pages 316-322. https://doi.org/10.1192/bjp.bp.116.195867

  1. Booth RG and others. ‘Youth mental health services utilization rates after a

large-scale social media campaign: population-based interrupted time-series analysis’ JMIR Mental Health 2018: volume 5, page e27. https://doi.org/10.2196/mental.8808

  1. Ross AM and Bassilios B. ‘Australian R U OK?Day campaign: improving helping

beliefs, intentions and behaviours’ International Journal of Mental Health Systems 2019: volume 13, page 61. https://doi.org/10.1186/s13033-019-0317-4

  1. Hacking I. ‘Kinds of people: moving targets: British Academy Lecture’

Proceedings of the British Academy 2007: volume 151, pages 285-318. https://doi.org/10.5871/bacad/9780197264249.003.0010

  1. Hacking I. ‘The looping effects of human kinds’. In: Sperber D and others,

editors. ‘Causal cognition: a multidisciplinary debate’ Clarendon Press, 1995, pages 351-383.

610 Baes N and others. ‘The semantic inflation of “trauma” in psychology’ Psychology of Language and Communication 2023: pages 23-45. https://doi.org/10.58734/plc-2023-0002

611 O’Connor C and others. ‘How trauma is represented on social media: analysis of #trauma content on TikTok’ Psychological Trauma: Theory, Research, Practice, and Policy 2025: volume 17, pages S132-S138. https://doi.org/10.1037/tra0001792

612 Ju R and others. ‘#MentalHealthDiagnosis and self-diagnosis: harmful or helpful on TikTok’ Health Communication 2026. https://doi.org/10.1080/10410236.2026.2652419

613 Alper M. ‘Kids across the spectrums: growing up autistic in the digital age’ MIT Press, 2023. https://doi.org/10.7551/mitpress/14152.001.0001

614 Kiehl I and others. ‘The adult experience of being diagnosed with autism spectrum disorder: a qualitative meta-synthesis’ Autism 2024: volume 28, pages 1060-1074. https://doi.org/10.1177/13623613231220419

615 Turpin G and others. ‘Effectiveness of providing self-help information following acute traumatic injury: randomised controlled trial’ British Journal of Psychiatry 2005: volume 187, pages 76-82. https://doi.org/10.1192/bjp.187.1.76

616 Scholes C and others. ‘A randomised controlled trial to assess the effectiveness of providing self-help information to people with symptoms of acute stress disorder following a traumatic injury’ Behaviour Research and Therapy 2007: volume 45, pages 2527-2536. https://doi.org/10.1016/j.brat.2007.06.009

617 Foulkes L and others. ‘The psychological consequences of mental health awareness efforts’ Nature Reviews Psychology 2026: volume 5, pages 173-184. https://doi.org/10.1038/s44159-026-00532-7

618 Sandra DA and others. ‘Inform and do no harm: nocebo education reduces false self-diagnosis caused by mental health awareness’ Psychological Medicine 2025: volume 55, page e330. https://doi.org/10.1017/s0033291725101979

619 Hasan F and others. ‘Normalizing anxiety on social media increases self- diagnosis of anxiety: the mediating effect of identification (but not stigma)’ Journal of Health Communication 2023: volume 28, pages 563-572. https://doi.org/10.1080/10810730.2023.2235563

620 Deighton J and others. ‘Effectiveness of school mental health awareness interventions: universal approaches in English secondary schools’ Department for Education, 2025. https://assets.publishing.service.gov.uk/media/680f52ef5072e9b7db83cce9/Effective ness_of_school_mental_health_awareness_interventions.pdf

621 Guzman-Holst C and others. ‘Scoping review: potential harm from school-based group mental health interventions’ Child and Adolescent Mental Health 2025: volume 30, pages 208-222. https://doi.org/10.1111/camh.12760

622 Tsujimoto K and others. ‘Meta-analysis: the effectiveness of universal school- based mental health programs for children and adolescents’ JAACP Open 2026. https://doi.org/10.1016/j.jaacop.2026.07.001

623 Foulkes L and others. ‘Research recommendations for assessing potential harm from universal school-based mental health interventions’ Nature Mental Health 2024: volume 2, pages 270-277. https://doi.org/10.1038/s44220-024-00208-2

624 Neumann M and others. ‘Increasing self- and desired psychiatric diagnoses among emerging adults: mixed-methods insights from clinical psychologists’ International Journal of Clinical and Health Psychology 2026: volume 26, page 100661. https://doi.org/10.1016/j.ijchp.2025.100661

625 Royal College of Psychiatrists. ‘Survey commissioned by and submitted to the Independent Review’ 2026.

626 Armstrong S and others. ‘Self-diagnosis in the age of social media: A pilot study of youth entering mental health treatment for mood and anxiety disorders’ Acta Psychologica 2025: volume 256, page 105015. https://doi.org/10.1016/j.actpsy.2025.105015

627 Cunningham-Rowe K and Foulkes L. ‘Adolescents’ self-diagnosis of mental disorders: a qualitative study of mental health clinicians’ perspectives’ Child and Adolescent Mental Health 2026. https://doi.org/10.1111/camh.70102

628 Walker N. ‘Neuroqueer heresies: notes on the neurodiversity paradigm, autistic empowerment, and postnormal possibilities’ Autonomous Press, 2021

629 Chapman R. ‘Empire of normality: neurodiversity and capitalism’ Pluto Press, 2023

630 Kapp SK and others. ‘Deficit, difference, or both? Autism and neurodiversity’ Developmental Psychology 2013: volume 49, pages 59-71. https://doi.org/10.1037/a0028353

631 Croce L and Fusaro I. ‘Diagnostic inflation in autism spectrum disorder: an epistemological and methodological reappraisal’ Frontiers in Psychology 2026: volume 17, page 1847929. https://doi.org/10.3389/fpsyt.2026.1847929

632 Adamou M and Wharton L. ‘Comparative analysis of diagnostic expectations between adults referred for ADHD and autism assessment: a thematic analysis’

Frontiers in Psychology 2026: volume 17, page 1780570. https://doi.org/10.3389/fpsyt.2026.1780570

633 Constantino JN and others. ‘Deficits in reciprocal social behavior in male twins: evidence for a genetically independent domain of psychopathology’ Journal of the American Academy of Child and Adolescent Psychiatry 2003: volume 42, pages 458-467. https://doi.org/10.1097/01.CHI.0000046811.95464.21

634 Levy F and others. ‘Attention-deficit hyperactivity disorder: a category or a continuum? Genetic analysis of a large-scale twin study’ Journal of the American Academy of Child and Adolescent Psychiatry 1997: volume 36, pages 737-744. https://doi.org/10.1097/00004583-199706000-00009

635 World Health Organization. ‘Autism’ 2025. https://www.who.int/news-room/fact- sheets/detail/autism-spectrum-disorders

636 Nelson RH. ‘A critique of the neurodiversity view’ Journal of Applied Philosophy 2020: volume 38, pages 335-347. https://doi.org/10.1111/japp.12470

637 Jaarsma P and Welin S. ‘Autism as a natural human variation: reflections on the claims of the neurodiversity movement’ Health Care Analysis 2012: volume 20, pages 20-30. https://doi.org/10.1007/s10728-011-0169-9

638 Hughes JA. ‘Does the heterogeneity of autism undermine the neurodiversity paradigm?’ Bioethics 2021: volume 35, pages 47-60. https://doi.org/10.1111/bioe.12780

639 Green J. ‘Debate: Neurodiversity, autism and healthcare’ Child and Adolescent Mental Health 2023: volume 28, pages 438-442. https://doi.org/10.1111/camh.12663

640 Dwyer P and others. ‘Neurodiversity movement identification and perceived appropriateness of terms used to describe autism’ Autism in Adulthood 2025. https://doi.org/10.1089/aut.2024.0297

641 Student Minds. ‘Student Minds Research Briefing – February ’23’ 2023. https://www.studentminds.org.uk/wp- content/uploads/2025/12/student_minds_insight_briefing_feb23.pdf

642 Young S and others. ‘Females with ADHD: an expert consensus statement taking a lifespan approach providing guidance for the identification and treatment of attention-deficit/hyperactivity disorder in girls and women’ BMC Psychiatry 2020: volume 20, page 404. https://doi.org/10.1186/s12888-020-02707-9

643 Asherson P and others. ‘SASC Guidance on the assessment and identification of the characteristics of an Attention Deficit Hyperactivity Disorder (ADHD)’ SpLD Assessment Standards Committee, 2021. https://www.patoss- dyslexia.org/write/MediaUploads/ADHD_Guidance_JUNE_21.pdf

  1. ADHD Girls. ‘Staring Back At Me Campaign’ 2022. https://adhdgirls.co.uk/

645 de Vries W and others. ‘Exploring concept creep: youth’s portrayal of ADHD on TikTok’ SSM - Mental Health 2025: volume 8, page 100489. https://doi.org/10.1016/j.ssmmh.2025.100489

  1. Fitzpatrick M and others. ‘#OCD: a content analysis of obsessive-compulsive

disorder stereotype amplification and misinformation on TikTok’ Cyberpsychology, Behavior, and Social Networking 2025: volume 28, pages 607-615. https://doi.org/10.1177/21522715251370135

  1. Baranyi G and others. ‘Early life-course environmental exposures and

adolescent common mental illness in the UK: a 17-year follow-up study’ Environment International 2026: volume 214, page 110361. https://doi.org/10.1016/j.envint.2026.110361

  1. Sibieta L and Snape D. ‘Spending on special educational needs in England:

something has to change’ Institute for Fiscal Studies, 2024. https://ifs.org.uk/sites/default/files/2024-12/Spending-on-special-educational-needs- in-England.pdf

  1. Department for Education. ‘Education, health and care plans: reporting year

2025’ 2025. https://explore-education-statistics.service.gov.uk/find- statistics/education-health-and-care-plans/2025

  1. Moore A. ‘ICB with 30-fold increase in neurodiversity spend now paying 42

private providers’ HSJ 2026. https://www.hsj.co.uk/finance-and-efficiency/icb-with- 30-fold-increase-in-neurodiversity-spend-now-paying-42-private- providers/8124157.article

  1. Kostyrka-Allchorne K and others. ‘Online Parent Training for The Initial

Management of ADHD referrals (Optima): the protocol for a randomised controlled trial of a digital parenting intervention implemented to support parents and children on a treatment waitlist’ Trials 2022: volume 23, page 1003. https://doi.org/10.1186/s13063-022-06952-z

  1. Hinshaw SP and others. ‘Prospective follow-up of girls with attention-

deficit/hyperactivity disorder into early adulthood: continuing impairment includes elevated risk for suicide attempts and self-injury’ Journal of Consulting and Clinical Psychology 2012: volume 80, pages 1041-1051. https://doi.org/10.1037/a0029451

  1. Price A and others. ‘Seven steps to mapping health service provision: lessons

learned from mapping services for adults with attention-deficit/hyperactivity disorder (ADHD) in the UK’ BMC Health Services Research 2019: volume 19, page 468. https://doi.org/10.1186/s12913-019-4287-7

  1. NHS Business Services Authority. ‘Medicines used in mental health – England –

2015/16 to 2024/25’ 2025. https://www.nhsbsa.nhs.uk/statistical- collections/medicines-used-mental-health-england/medicines-used-mental-health- england-annual/medicines-used-mental-health-england-201516-202425

  1. University of Essex, Institute for Social and Economic Research. ‘Understanding

Society: Waves 1-14, 2009-2023 and Harmonised BHPS: Waves 1-18, 1991-2009. [data collection]. 19th edition’ UK Data Service. SN: 6614, 2024. https://doi.org/10.5255/UKDA-SN-6614-20

  1. Panayiotou M and others. ‘Longitudinal pathways between emotional difficulties

and school absenteeism in middle childhood: evidence from developmental cascades’ Development and Psychopathology 2023: volume 35, pages 1323-1334. https://doi.org/10.1017/S095457942100122X

  1. Peng A and others. ‘Longitudinal relations between school connectedness and

internalizing problems during early adolescence’ Journal of Youth and Adolescence 2026: volume 55, pages 1064-1083. https://doi.org/10.1007/s10964-025-02291-9

  1. Knowles C and others. ‘Mapping the missed days: latent patterns of attendance

and their predictors in English secondary schools’ PsyArXiv 2025. https://doi.org/10.31234/osf.io/6e2r4_v1

  1. Cheng Q and others. ‘Adolescent internalizing symptoms, school experiences,

and attendance: sex-specific within-person pathways’ Journal of the American Academy of Child and Adolescent Psychiatry 2026. https://doi.org/10.1016/j.jaac.2026.04.001

  1. Department for Education. ‘Pupil absence in schools in England: Academic year

2024/25’ 2026. https://explore-education-statistics.service.gov.uk/find-statistics/pupil- absence-in-schools-in-england/2024-25

  1. Department for Education. ‘Parent, pupil and learner voice: May 2025’ 2025.

https://www.gov.uk/government/publications/parent-pupil-and-learner-voice- omnibus-surveys-for-2024-to-2025/parent-pupil-and-learner-voice-may-2025

  1. Department for Education. ‘Parent, pupil and learner voice: December 2025’

2026. https://www.gov.uk/government/publications/parent-pupil-and-learner-voice- omnibus-surveys-for-2025-to-2026/parent-pupil-and-learner-voice-december-2025

  1. Department for Education. ‘Elective home education: Autumn term 2026/26’

2025. https://explore-education-statistics.service.gov.uk/find-statistics/elective-home- education/2025-26-autumn-term

  1. Diogu MI and others. ‘Risk of suicide following school persistent absence and

exclusion: an electronic cohort study in Wales, UK 2012-2019’ Journal of Affective Disorders 2025: volume 385, page 119394. https://doi.org/10.1016/j.jad.2025.119394

  1. Lereya ST and others. ‘Mental health difficulties, attainment and attendance: a

cross-sectional study’ European Child & Adolescent Psychiatry 2019: volume 28, pages 1147-1152. https://doi.org/10.1007/s00787-018-01273-6

  1. Parentkind. ‘The National Parent Survey 2024’ 2024.

https://www.parentkind.org.uk/assets/resources/NPS-24- Hyperlinked_compressed.pdf

  1. Department for Education. ‘Pupil attendance in schools, Week 29 2026 (End of

25/26 AY)’ 2026. https://explore-education-statistics.service.gov.uk/find- statistics/pupil-attendance-in-schools/2026-week-29-end-of-25-26-ay

  1. Department for Education. ‘Working together to improve school attendance:

statutory guidance for maintained schools, academies, independent schools and local authorities’ 2026. https://assets.publishing.service.gov.uk/media/6a4f9ec6a6586e258d371bd0/Working _together_to_improve_school_attendance_2026.pdf

  1. Allen K-A and Boyle C. ‘School belonging and student engagement: the critical

overlaps, similarities, and implications for student outcomes’. In: Christenson SL and others, editors. ‘Handbook of research on student engagement’ Springer, 2022, pages 133-154.

670 Aslan K and others. ‘The relationship between school belonging and societal belonging among ethnic minority adolescents in the Czech Republic’ Young 2025: volume 34, pages 275-302. https://doi.org/10.1177/11033088251379973

671 Guo X and others. ‘The association between academic pressure and adolescent depressive symptoms and self-harm: a longitudinal, prospective study in England’ Lancet Child & Adolescent Health 2026: volume 10, pages 265-272. https://doi.org/10.1016/S2352-4642(25)00342-6

672 Lowry C and others. ‘Teachers: the forgotten health workforce’ Journal of the Royal Society of Medicine 2022: volume 115, pages 133-137. https://doi.org/10.1177/01410768221085692

673 Jerrim J and Sims S. ‘When is high workload bad for teacher wellbeing? Accounting for the non-linear contribution of specific teaching tasks’ Teaching and Teacher Education 2021: volume 105, page 103395. https://doi.org/10.1016/j.tate.2021.103395

674 Kidger J and others. ‘Teachers’ wellbeing and depressive symptoms, and associated risk factors: a large cross sectional study in English secondary schools’ Journal of Affective Disorders 2016: volume 192, pages 76-82. https://doi.org/10.1016/j.jad.2015.11.054

675 Harding S and others. ‘Is teachers’ mental health and wellbeing associated with students' mental health and wellbeing?’ Journal of Affective Disorders 2019: volume 242, pages 180-187. https://doi.org/10.1016/j.jad.2018.08.080

676 Isos Partnership and County Councils Network. ‘Special educational needs and disabilities in England: how we got to crisis point, and why we need reform of the system’ 2025. https://uploads.countycouncilsnetwork.org.uk/uRtdfHIYombbrCKXzl6ei

677 Walker M and others. ‘The uneven distribution of pupils with SEND in mainstream schools’ Nuffield Foundation, 2026. https://www.nuffieldfoundation.org/project/the-uneven-distribution-of-pupils-with- send-in-mainstream-schools

678 Department for Education. ‘Education, health and care plans: reporting year 2026’ 2026. https://explore-education-statistics.service.gov.uk/find- statistics/education-health-and-care-plans/2026

679 House of Commons Public Accounts Committee. ‘Support for children and young people with special educational needs. First report of Session 2024–25’ 2025. https://publications.parliament.uk/pa/cm5901/cmselect/cmpubacc/353/report.html

680 National Audit Office. ‘Support for children and young people with special educational needs’ 2024. https://www.nao.org.uk/reports/support-for-children-and- young-people-with-special-educational-needs/

681 Education Policy Institute. ‘Annual Report 2025’ 2025. https://epi.org.uk/annual- report-2025/

682 Azpitarte F and Holt L. ‘Failing children with Special Educational Needs and Disabilities in England: new evidence of poor outcomes and a postcode lottery at the Local Authority level at Key Stage 1’ British Educational Research Journal 2024: volume 50, pages 414-437. https://doi.org/10.1002/berj.3930

683 Department for Education. ‘Special educational needs in England, academic year 2025/26’ Department of Education, 2026. https://explore-education- statistics.service.gov.uk/find-statistics/special-educational-needs-in-england/2025-26

684 Youth Justice Board. ‘Annual report and accounts, 2020 to 2021’ 2021. https://www.gov.uk/government/publications/youth-justice-board-annual-report-and- accounts-2020-to-2021

685 Coles EK and others. ‘Randomized trial of first-line behavioral intervention to reduce need for medication in children with ADHD’ Journal of Clinical Child and Adolescent Psychology 2020: volume 49, pages 673-687. https://doi.org/10.1080/15374416.2019.1630835

686 Aarnoudse-Moens CSH and others. ‘Meta-analysis of neurobehavioral outcomes in very preterm and/or very low birth weight children’ Pediatrics 2009: volume 124, pages 717-728. https://doi.org/10.1542/peds.2008-2816

687 Pettinger KJ and others. ‘Starting school: educational development as a function of age of entry and prematurity’ Archives of Disease in Childhood 2020: volume 105, pages 160-165. https://doi.org/10.1136/archdischild-2019-317124

688 Lysons J and others. ‘Measuring child development at the 2–2½-year health and development review in England: a rapid scoping review of available tools’ BMJ Open 2026: volume 16, page e102853. https://doi.org/10.1136/bmjopen-2025-102853

689 Department for Education. ‘Early years foundation stage profile results’ 2025. https://explore-education-statistics.service.gov.uk/find-statistics/early-years- foundation-stage-profile-results/2024-25

690 Reardon T and others. ‘Screening-to-intervention pathway for child anxiety problems alongside usual school practice versus usual school practice only (iCATSi2i): a cluster-randomised, controlled trial in primary schools in England’ Lancet Psychiatry 2026: volume 13, pages 396-412. https://doi.org/10.1016/S2215- 0366(26)00064-7

691 Bonell C and others. ‘Effects of the Learning Together intervention on bullying and aggression in English secondary schools (Inclusive): a cluster randomised controlled trial’ Lancet 2018: volume 392, pages 2452-2464. https://doi.org/10.1016/S0140-6736(18)31782-3

692 Ford T and others. ‘The relationship between exclusion from school and mental health: a secondary analysis of the British Child and Adolescent Mental Health Surveys 2004 and 2007’ Psychological Medicine 2018: volume 48, pages 629-641. https://doi.org/10.1017/S003329171700215X

693 Patalay P and others. ‘School characteristics and children’s mental health: a linked survey-administrative data study’ Preventive Medicine 2020: volume 141, page 106292. https://doi.org/10.1016/j.ypmed.2020.106292

694 Doyle L and others. ‘An empathy intervention reduces the gender gap in school discipline and facilitates belonging’ European Journal of Social Psychology 2025: volume 55, pages 327-341. https://doi.org/10.1002/ejsp.3146

695 Jerrim J. ‘How do pupil feelings of school safety relate to their attendance and wellbeing? Longitudinal evidence from England’ Cambridge Journal of Education 2025: volume 55, pages 499-524. https://doi.org/10.1080/0305764x.2025.2521512

696 White SR and others. ‘Networks of care for the modern adolescent’ Psychological Medicine 2024: volume 54, pages 4537-4550. https://doi.org/10.1017/S003329172400237X

697 Hayes D and others. ‘Universal, school-based, interventions to improve emotional outcomes in children and young people: a systematic review and meta- analysis’ Frontiers in Child and Adolescent Psychiatry 2025: volume 4, page 1526840. https://doi.org/10.3389/frcha.2025.1526840

698 Clarke A and others. ‘Adolescent mental health: a systematic review on the effectiveness of school-based interventions’ Early Intervention Foundation, 2021. https://foundations.org.uk/wp-content/uploads/2026/03/eif-adolescent-mental-health- review-report-plus-appendices.pdf

699 Deighton J and others. ‘Promoting mental health and wellbeing in schools: examining mindfulness-based exercises, relaxation practices and Strategies for Safety and Wellbeing in English primary and secondary schools (Inspire): a multi- school, cluster randomised controlled trial’ OSF 2025. https://doi.org/10.17605/Osf.Io/KXUG7

700 Brown J and others. ‘Clinical effectiveness and cost-effectiveness of a brief accessible cognitive behavioural therapy programme for stress in school-aged adolescents (BESST): a cluster randomised controlled trial in the UK’ Lancet Psychiatry 2024: volume 11, pages 504-515. https://doi.org/10.1016/S2215- 0366(24)00101-9

701 Reardon T and others. ‘Parent-led CBT delivered via online and telephone support alongside usual school practice versus usual school practice only for young children identified as at risk for anxiety disorders through screening in schools: a cluster randomised controlled trial’ Journal of Child Psychology and Psychiatry 2026. https://doi.org/10.1111/jcpp.70119

702 Kuyken W and others. ‘Effectiveness and cost-effectiveness of universal school- based mindfulness training compared with normal school provision in reducing risk of mental health problems and promoting well-being in adolescence: the Myriad cluster randomised controlled trial’ Evidence Based Mental Health 2022: volume 25, pages 99-109. https://doi.org/10.1136/ebmental-2021-300396

703 Montero-Marin J and others. ‘School-based mindfulness training in early adolescence: what works, for whom and how in the Myriad trial?’ Evidence Based Mental Health 2022: volume 25, pages 117-124. https://doi.org/10.1136/ebmental- 2022-300439

704 Ellins J and others. ‘Early evaluation of the Children and Young People’s Mental Health Trailblazer programme: a rapid mixed-methods study’ Health and Social Care Delivery Research 2023: volume 11, pages 1-137. https://doi.org/10.3310/XQWU4117

705 Payne S and others. ‘The effectiveness of low-intensity mental health interventions for children and young people in England: a presentation and analysis

of synthesised outcomes’ Child and Adolescent Mental Health 2026. https://doi.org/10.1111/camh.70130

706 Creswell C and others. ‘Digitally augmented, parent-led CBT versus treatment as usual for child anxiety problems in child mental health services in England and Northern Ireland: a pragmatic, non-inferiority, clinical effectiveness and cost- effectiveness randomised controlled trial’ Lancet Psychiatry 2024: volume 11, pages 193-209. https://doi.org/10.1016/S2215-0366(23)00429-7

707 Mundy J and others. ‘Evaluating the implementation of the Transforming Children and Young People’s Mental Health Provision Green Paper programme: findings from surveys of schools and colleges and Mental Health Support Teams (2024)’ National Children’s Bureau, 2025. https://www.ncb.org.uk/physicalmentalhealth/mental-health-and- wellbeing/evaluating-implementation-transforming-children

708 Norbury CF and others. ‘Language growth in children with heterogeneous language disorders: a population study’ Journal of Child Psychology and Psychiatry 2017: volume 58, pages 1092-1105. https://doi.org/10.1111/jcpp.12793

709 West G and others. ‘Oral language enrichment in preschool improves children’s language skills: a cluster randomised controlled trial’ Journal of Child Psychology and Psychiatry 2024: volume 65, pages 1087-1097. https://doi.org/10.1111/jcpp.13947

710 Boyle J and others. ‘A randomised controlled trial and economic evaluation of direct versus indirect and individual versus group modes of speech and language therapy for children with primary language impairment’ Health Technology Assessment 2007: volume 11, pages 1-160. https://doi.org/10.3310/hta11250

711 Rogde K and others. ‘The effect of linguistic comprehension instruction on generalized language and reading comprehension skills: a systematic review’ Campbell Systematic Reviews 2019: volume 15, page e1059. https://doi.org/10.1002/cl2.1059

712 Ross A and others. ‘The relationship between camouflaging and mental health in autistic children and adolescents’ Autism Research 2023: volume 16, pages 190- 199. https://doi.org/10.1002/aur.2859

713 Bronfenbrenner U. ‘The ecology of human development: experiments by nature and design’ Harvard University Press, 1979

714 Department for Education. ‘Suspensions and permanent exclusions in England, academic year 2024/25’ 2026. https://explore-education- statistics.service.gov.uk/find-statistics/suspensions-and-permanent-exclusions-in- england/2024-25

715 Tejerina-Arreal M and others. ‘Child and adolescent mental health trajectories in relation to exclusion from school from the Avon Longitudinal Study of Parents and Children’ Child and Adolescent Mental Health 2020: volume 25, pages 217-223. https://doi.org/10.1111/camh.12367

716 Obsuth I and others. ‘The impact of school exclusion in childhood on health and well-being outcomes in adulthood: estimating causal effects using inverse probability of treatment weighting’ British Journal of Educational Psychology 2024: volume 94, pages 460-473. https://doi.org/10.1111/bjep.12656

717 Ijaz S and others. ‘Disciplinary behaviour management strategies in schools and their impact on student psychosocial outcomes: a systematic review [version 2; peer review: 2 approved]’ NIHR Open Research 2024: volume 4, page 13. https://doi.org/10.3310/nihropenres.13563.2

718 Guldberg K and others. ‘Educational need as relational and dynamic: towards a needs-led framework for neurodivergent learners’ EdArXiv 2026. https://doi.org/10.35542/osf.io/3bwaz_v1

719 Crenna-Jennings W and others. ‘Children missing from education’ Education Policy Institute, 2024. https://epi.org.uk/wp-content/uploads/2024/12/CME- report_final-1.pdf

720 Webster R and Blatchford P. ‘The Special Educational Needs in Secondary Education (Sense) study. Final report. A study of the teaching and support experienced by pupils with Statements and Education, Health and Care Plans in mainstream and special schools’ UCL Institute of Education, 2017. https://discovery.ucl.ac.uk/id/eprint/10096865/1/Webster_SENSE_study_final_report. pdf

721 Higher Education Statistics Agency (HESA). ‘Who’s studying in HE? Personal characteristics’ 2026. https://www.hesa.ac.uk/data-and-analysis/students/whos-in- he/characteristics

722 McManus S and Gunnell D. ‘Trends in mental health, non‐suicidal self‐harm and suicide attempts in 16–24-year old students and non-students in England, 2000–

2014’ Social Psychiatry and Psychiatric Epidemiology 2020: volume 55, pages 125- 128. https://doi.org/10.1007/s00127-019-01797-5

723 AMOSSHE: The Student Services Organisation. ‘Student Services benchmarking 2002: wellbeing, mental health and counselling’ 2023. https://www.amosshe.org.uk/resource/student-services-benchmarking-2022-- wellbeing--mental-health-and-counselling.html

724 Association of Colleges. ‘Pressure mounts for college staff as student mental health declines, warns AoC’ 2026. https://www.aoc.co.uk/news-campaigns- parliament/aoc-newsroom/pressure-mounts-for-college-staff-as-student-mental- health-declines-warns-aoc

725 Saunders R and others. ‘Effectiveness of psychological interventions for young adults versus working age adults: a retrospective cohort study in a national psychological treatment programme in England’ Lancet Psychiatry 2025: volume 12, pages 650-659. https://doi.org/10.1016/S2215-0366(25)00207-X

726 Bridgland VME and others. ‘A meta-analysis of the efficacy of trigger warnings, content warnings, and content notes’ Clinical Psychological Science 2023: volume 12, pages 751-771. https://doi.org/10.1177/21677026231186625

727 Bellet BW and others. ‘Trigger warning: empirical evidence ahead’ Journal of Behavior Therapy and Experimental Psychiatry 2018: volume 61, pages 134-141. https://doi.org/10.1016/j.jbtep.2018.07.002

728 PTSD UK. ‘Trigger warnings and PTSD: do they help or harm?’. https://www.ptsduk.org/trigger-warnings/

729 Fornauf BS and Erickson JD. ‘Toward an inclusive pedagogy through universal design for learning in higher education: A review of the literature’ Journal of Postsecondary Education and Disability 2020: volume 33, pages 183-199. https://files.eric.ed.gov/fulltext/EJ1273677.pdf

730 Ofsted. ‘The annual report of His Majesty’s Chief Inspector of Education, Children’s Services and Skills 2024/25’ 2025. https://www.gov.uk/government/publications/ofsted-annual-report-202425-education- childrens-services-and-skills/the-annual-report-of-his-majestys-chief-inspector-of- education-childrens-services-and-skills-202425

731 Dickinson J. ‘Universities need to get a grip on reasonable adjustments’. Wonkhe. 2026. https://wonkhe.com/blogs/universities-need-to-get-a-grip-on- reasonable-adjustments/

732 Borkin H. ‘The experiences of staff working within Disability Services in Higher Education’ Journal of Inclusive Practice in Further and Higher Education 2023: volume 15, pages 5-34. https://nadp-uk.org/wp-content/uploads/2023/07/1- Borkin.docx

733 Office of the Independent Adjudicator. ‘OIA Annual Report 2025’ 2025. https://www.oiahe.org.uk/media/xfljp10u/oia-annual-report-2025.pdf

734 Department for Education. ‘Every Child Achieving and Thriving’ 2026. https://www.gov.uk/government/publications/every-child-achieving-and-thriving

735 Karasz A and others. ‘What we talk about when we talk about depression: doctor-patient conversations and treatment decision outcomes’ British Journal of General Practice 2012: volume 62, pages e55-63. https://doi.org/10.3399/bjgp12X616373

736 Killaspy H and others. ‘Community-based social interventions for people with severe mental illness: a systematic review and narrative synthesis of recent evidence’ World Psychiatry 2022: volume 21, pages 96-123. https://doi.org/10.1002/wps.20940

737 Davies J and others. ‘Politicians, experts, and patient representatives call for the UK government to reverse the rate of antidepressant prescribing’ BMJ 2023: volume 383, page 2730. https://doi.org/10.1136/bmj.p2730

738 Rocks S and others. ‘Introducing a single point of access (SPA) to child and adolescent mental health services in England: a mixed-methods observational study’ BMC Health Services Research 2020: volume 20, page 623. https://doi.org/10.1186/s12913-020-05463-4

739 Royal College of Emergency Medicine. ‘What’s behind the increase in demand in Emergency Departments?’ 2021. http://president.rcem.ac.uk/index.php/2021/08/06/whats-behind-the-increase-in- demand-in-emergency-departments/

740 NHS England. ‘Mental health clinically-led review of standards: models of care and measurement’ 2021. https://www.england.nhs.uk/publication/mental-health- clinically-led-review-of-standards-models-of-care-and-measurement/

741 Iacobucci G. ‘New mental health waiting time targets to be introduced in England’ BMJ 2021: volume 374, page n1870. https://doi.org/10.1136/bmj.n1870

742 Health Services Safety Investigations Body. ‘Mental health inpatient settings: supporting safe care during transition from inpatient children and young people’s mental health services to adult mental health services’ 2024. https://www.hssib.org.uk/patient-safety-investigations/mental-health-inpatient- settings/third-investigation-report/pdf/

743 Adamou M and Jones SL. ‘Adult ADHD diagnostic pathways in the United Kingdom: a structured review and policy case for a national commissioning and quality framework for England’ Frontiers in Psychiatry 2026: volume 17, page 1860550. https://doi.org/10.3389/fpsyt.2026.1860550

744 Brede J and others. ‘“We Have to Try to Find a Way, a Clinical Bridge” - autistic adults’ experience of accessing and receiving support for mental health difficulties: a systematic review and thematic meta-synthesis’ Clinical Psychology Review 2022: volume 93, page 102131. https://doi.org/10.1016/j.cpr.2022.102131

745 Moore A and others. ‘Regional analysis of ADHD and autism assessments and pathways’ Report commissioned by and submitted to the Independent Review, 2026.

746 Dunn B and others. ‘Equity for the ‘missing middle’: closing the gap between primary and secondary care services for patients with complex mental health needs’ British Journal of General Practice 2024: volume 74, pages 292-294. https://doi.org/10.3399/bjgp24X738513

747 National Audit Office. ‘Progress on improving mental health services in England’ 2023. https://www.nao.org.uk/reports/progress-in-improving-mental-health-services- in-england/

748 de Souza R. ‘Children’s mental health services 2023-24’ Office of the Children’s Commissioner for England, 2025. https://assets.childrenscommissioner.gov.uk/wpuploads/2025/05/cco-childrens- mental-health-annual-briefing-may-2025.pdf

749 Education Policy Institute. ‘Access to child and adolescent mental health services in 2019’ 2020. https://epi.org.uk/publications-and-research/access-to-child- and-adolescent-mental-health-services-in-2019/

750 Richards TC and others. ‘System failures in care continuity and coordination for people with multiple long-term conditions: an umbrella review of patient, carer, and

professional perspectives’ Health Policy 2026: volume 172, page 105697. https://doi.org/10.1016/j.healthpol.2026.105697

751 NHS England. ‘The community mental health framework for adults and older adults’ 2019. https://www.england.nhs.uk/publication/the-community-mental-health- framework-for-adults-and-older-adults/

752 NHS England. ‘24/7 neighbourhood mental health centres’ 2026. https://www.england.nhs.uk/mental-health/24-7-neighbourhood-mental-health- centres/

753 Hobson-Merrett C and others. ‘What works, for whom, in what circumstances and why, when integrating voluntary and statutory community mental health services: a realist evaluation case study’ International Journal of Integrated Care 2025: volume 25, page 31. https://doi.org/10.5334/ijic.9060

754 Carr S and others. ‘Service users’ experiences of social and psychological avoidable harm in mental health social care in England: findings of a scoping review’ British Journal of Social Work 2023: volume 53, pages 1303-1324. https://doi.org/10.1093/bjsw/bcac209

755 Davis ALW and others. ‘Discharge from secondary care services to primary care for adults with serious mental illness: a scoping review’ BMC Psychiatry 2024: volume 24, page 614. https://doi.org/10.1186/s12888-024-06067-6

756 Cartwright C and others. ‘A meta-ethnography of shared decision-making in mental health care from the perspective of staff and service users’ BMC Health Services Research 2024: volume 24, page 1142. https://doi.org/10.1186/s12913- 024-11540-9

757 de Souza R. ‘Children’s and young people’s mental health services: 2024-25’ Office of the Children’s Commissioner for England, 2026. https://assets.childrenscommissioner.gov.uk/wpuploads/2026/06/Childrens-Mental- Health-Annual-Briefing-24-25.pdf

758 Care Quality Commission. ‘Community mental health survey 2025’ 2025. https://www.cqc.org.uk/publications/surveys/community-mental-health-survey

759 Royal College of Psychiatrists. ‘Hidden waits force more than three quarters of mental health patients to seek help from emergency services’ 2022. https://www.rcpsych.ac.uk/news-and-features/latest-news/detail/2022/10/10/hidden-

waits-force-more-than-three-quarters-of-mental-health-patients-to-seek-help-from- emergency-services

760 Ward JL and others. ‘Admission to acute medical wards for mental health concerns among children and young people in England from 2012 to 2022: a cohort study’ Lancet Child & Adolescent Health 2025: volume 9, pages 112-120. https://doi.org/10.1016/S2352-4642(24)00333-X

761 Syed S and others. ‘Systematic review and meta-analysis: effectiveness of intensive community care services and psychosocial interventions for adolescents with severe mental health problems’ Journal of the American Academy of Child and Adolescent Psychiatry 2026. https://doi.org/10.1016/j.jaac.2026.01.006

762 Cornes M and others. ‘Improving care transfers for homeless patients after hospital discharge: a realist evaluation’ Health and Social Care Delivery Research 2021: volume 9, pages 1-186. https://doi.org/10.3310/hsdr09170

763 British Psychological Society. ‘The emotional well-being and mental health of babies and young children from pregnancy to age five: psychologically-informed care, psychological interventions and the role of the psychological professions’ 2026. https://cms.bps.org.uk/sites/default/files/2026-01/Draft%20Guidance%20- %20Emotional%20Well- Being%20and%20Mental%20Health%20of%20Babies%20and%20Young%20Childr en%20from%20Pregnancy%20to%20age%20five.pdf

764 NHS North West Coast Clinical Networks. ‘Mental Health for infants and children aged 0-5 years old: a strategy for Lancashire and South Cumbria’ 2025. https://www.england.nhs.uk/north-west/wp-content/uploads/sites/48/2025/03/LSC-0- 5-Mental-Health-Strategy-2025_v1.0.pdf

765 Lyons-Ruth K and others. ‘The worldwide burden of infant mental and emotional disorder: report of the Task Force of the World Association for Infant Mental Health’ Infant Mental Health Journal 2017: volume 38, pages 695-705. https://doi.org/10.1002/imhj.21674

766 NHS England. ‘Performance report’ 2025. https://www.england.nhs.uk/long- read/performance-report-24-25/

767 Maternal Mental Health Alliance. ‘Maternal mental health services: progress report’ 2024. https://maternalmentalhealthalliance.org/media/filer_public/8c/f5/8cf5b1f4-c19b- 4554-bbf5-caf6ebf4a6d4/mmha_progress_report_on_mmhs_final.pdf

768 Rosan C and others. ‘Integrate: a guide to integrating psychological support for common perinatal mental health difficulties into universal services ’ Anna Freud, 2025. https://www.corc.uk.net/media/qhimzcqy/v4-447_integrate-report_jul-25.pdf

769 Amos V. ‘Independent investigation into maternity and neonatal services in England’ 2026. https://www.matneoinv.org.uk/wp-content/uploads/2026/08/NMNI- Final-Report-and-Recommendations.pdf

770 Reed J and Parish N. ‘Working for babies: lockdown lessons from local systems’ First 1000 Days Movement, 2021. https://parentinfantfoundation.org.uk/wp- content/uploads/2021/01/210121-F1001D_Working_for_Babies_v1.2-FINAL- compressed_2.pdf

771 Royal College of Psychiatrists. ‘Infant and early childhood mental health: the case for action’ 2023. https://www.rcpsych.ac.uk/docs/default-source/improving- care/better-mh-policy/college-reports/college-report-cr238---infant-and-early- childhood-mental-health.pdf?sfvrsn=1d8d5efd_15

772 Institute of Health Visiting. ‘Understanding the rise in 0-4-year-old Emergency Department (ED) attendances and changing health visiting practice’ 2023. https://ihv.org.uk/wp-content/uploads/2023/12/Understanding-the-rise-in-0-4-year- old-Emergency-Department-ED-attendances-and-changing-health-visiting-practice- FINAL-VERSION-18.12.23.pdf

773 Royal Foundation Centre for Early Childhood. ‘Big change starts small’ 2021. https://centreforearlychildhood.org/wp- content/uploads/2024/11/RFCEC_BCCS_Report_and_Appendices.pdf

774 Caspi A and others. ‘Behavioral observations at age 3 years predict adult psychiatric disorders. Longitudinal evidence from a birth cohort’ Archives of General Psychiatry 1996: volume 53, pages 1033-1039. https://doi.org/10.1001/archpsyc.1996.01830110071009

775 Kim-Cohen J and others. ‘Prior juvenile diagnoses in adults with mental disorder: developmental follow-back of a prospective-longitudinal cohort’ Archives of General Psychiatry 2003: volume 60, pages 709-717. https://doi.org/10.1001/archpsyc.60.7.709

776 Cullingham T and others. ‘Anxiety problems experienced by autistic children: child and parent perspectives ’ Autism In press.

777 Lockwood Estrin G and others. ‘Barriers to autism spectrum disorder diagnosis for young women and girls: a systematic review’ Review Journal of Autism and Developmental Disorders 2021: volume 8, pages 454-470. https://doi.org/10.1007/s40489-020-00225-8

778 Reardon T and others. ‘What do parents perceive are the barriers and facilitators to accessing psychological treatment for mental health problems in children and adolescents? A systematic review of qualitative and quantitative studies’ European Child & Adolescent Psychiatry 2017: volume 26, pages 623-647. https://doi.org/10.1007/s00787-016-0930-6

779 Bennett C and others. ‘Group-based parenting programs for improving parenting and psychosocial functioning: a systematic review’ Journal of the Society for Social Work and Research 2013: volume 4, pages 300-332. https://doi.org/10.5243/jsswr.2013.20

780 Saccà A and others. ‘Parents of children with autism spectrum disorder: a systematic review’ Journal of Clinical & Developmental Psychology 2019: volume 1. https://doi.org/10.6092/2612-4033/0110-2174

781 McDonald TP and others. ‘Predicting caregiver stress: an ecological perspective’ American Journal of Orthopsychiatry 1999: volume 69, pages 100-109. https://doi.org/10.1037/h0080385

782 Deloitte. ‘Mental health and employers: the case for employers to invest in supporting working parents and a mentally healthy workplace’ 2024. https://www.deloitte.com/uk/en/services/consulting/research/mental-health-and- employers-the-case-for-employers-to-invest-in-supporting-working-parents-and-a- mentally-health-workplace.html

783 Yatchmenoff DK and others. ‘Enrichment and stress in families caring for a child with a serious emotional disorder’ Journal of Child and Family Studies 1998: volume 7, pages 129-145. https://doi.org/10.1023/A:1022935014502

784 Tinelli M and others. ‘Economic analysis of early intervention for autistic children: findings from four case studies in England, Ireland, Italy, and Spain’ European Psychiatry 2023: volume 66, page e76. https://doi.org/10.1192/j.eurpsy.2023.2449

785 Segal L and others. ‘Estimated therapy costs and downstream cost consequences of iBASIS–Video Interaction to Promote Positive Parenting intervention vs usual care among children displaying early behavioral signs of autism

in Australia’ JAMA Network Open 2023: volume 6, page e235847. https://doi.org/10.1001/jamanetworkopen.2023.5847

786 El Baou C and others. ‘Effectiveness of primary care psychological therapy services for treating depression and anxiety in autistic adults in England: a retrospective, matched, observational cohort study of national health-care records’ Lancet Psychiatry 2023: volume 10, pages 944-954. https://doi.org/10.1016/S2215- 0366(23)00291-2

787 NHS England. ‘Patient and carer race equality framework’ 2023. https://www.england.nhs.uk/publication/patient-and-carer-race-equality-framework/

788 NHS England. ‘Culture of care standards for mental health inpatient services: including those for people with a learning disability and autistic people’ 2024. https://www.england.nhs.uk/publication/culture-of-care-standards-for-mental-health- inpatient-services/

  1. National Institute for Health and Care Excellence. https://www.nice.org.uk/
  2. Darzi A. ‘Independent investigation of the NHS in England’ Department of Health

and Social Care, 2024. https://www.gov.uk/government/publications/independent- investigation-of-the-nhs-in-england

  1. National Audit Office. ‘Improving children and young people’s mental health

services’ 2018. https://www.nao.org.uk/reports/improving-children-and-young- peoples-mental-health-services/

  1. Chisholm D and others. ‘Scaling-up treatment of depression and anxiety: a

global return on investment analysis’ Lancet Psychiatry 2016: volume 3, pages 415- 424. https://doi.org/10.1016/S2215-0366(16)30024-4

  1. Yardley E and others. ‘Data-driven exploration of National Health Service

Talking Therapies care pathways using process mining: retrospective cohort study’ JMIR Mental Health 2024: volume 11, page e53894. https://doi.org/10.2196/53894

  1. Sayal K and others. ‘The clinical and cost effectiveness of a STAndardised

DIagnostic Assessment for children and adolescents with emotional difficulties: the Stadia multi-centre randomised controlled trial’ Journal of Child Psychology and Psychiatry 2025: volume 66, pages 805-820. https://doi.org/10.1111/jcpp.14090

  1. NHS England Mental Health Services Data Set. ‘Community Mental Health

Dashboard - NHS England National Adult and Older Adult Mental Health Programme’ 2026. https://digital.nhs.uk/data-and-information/data-tools-and- services/data-services/mental-health-data-hub/dashboards

  1. National Institute for Health and Care Excellence. ‘Self-harm: assessment,

management and preventing recurrence (NG225)’ 2022. https://www.nice.org.uk/guidance/ng225

  1. Geulayov G and others. ‘Epidemiology and trends in non-fatal self-harm in three

centres in England, 2000-2012: findings from the Multicentre Study of Self-harm in England’ BMJ Open 2016: volume 6, page e010538. https://doi.org/10.1136/bmjopen-2015-010538

  1. Blissett C and others. ‘Why babies’ first relationships matter: the value of parent-

infant relationships in the UK’ Centre for Mental Health, 2026. https://www.centreformentalhealth.org.uk/publications/why-babies-first-relationships- matter/

  1. Mughal F and others. ‘Exploring general practitioners’ management of self-harm

in young people: a qualitative study’ Health Expectations 2024: volume 27, page e70026. https://doi.org/10.1111/hex.70026

  1. Delgadillo J and others. ‘Stratified care vs stepped care for depression: a cluster

randomized clinical trial’ JAMA Psychiatry 2022: volume 79, pages 101-108. https://doi.org/10.1001/jamapsychiatry.2021.3539

  1. Wolpert M and others. ‘Thrive framework for system change’ CAMHS Press,

2019. https://implementingthrive.org/wp-content/uploads/2024/08/THRIVE- Framework-for-system-change-2019NEW-2.pdf

  1. Farr J and others. ‘The impact of a needs-based model of care on accessibility

and quality of care within children's mental health services: a qualitative investigation of the UK i-THRIVE Programme’ Child: Care, Health and Development 2021: volume 47, pages 442-450. https://doi.org/10.1111/cch.12855

  1. Sippy R and others. ‘Effect of a needs-based model of care on the

characteristics of healthcare services in England: the i-THRIVE National Implementation Programme’ Epidemiology and Psychiatric Sciences 2025: volume 34, page e21. https://doi.org/10.1017/S2045796025000101

  1. Baker HJ and Waite P. ‘The identification and psychological treatment of panic

disorder in adolescents: a survey of CAMHS clinicians’ Child and Adolescent Mental Health 2020: volume 25, pages 135-142. https://doi.org/10.1111/camh.12372

  1. Conti G and others. ‘The state of health visiting in England: workforce

composition, caseloads and service delivery’ medRxiv 2026. https://doi.org/10.64898/2026.03.26.26349382

  1. NHS England. ‘Guidance on neighbourhood multidisciplinary teams for children

and young people’ 2025. https://www.england.nhs.uk/long-read/guidance-on- neighbourhood-multidisciplinary-teams-for-children-and-young-people/

  1. Bertotti M and others. ‘A two-year evaluation of the Young People Social

Prescribing (YPSP) pilot’ 2020. https://hostinger.sheffieldfutures.org.uk/wp- content/uploads/2024/08/SP-for-young-people-evaluation-final-report-for-publication- 1.pdf

  1. NIHR Community Mental Health Knowledge Exchange. ‘A whole systems

approach’ 2026. https://sites.google.com/nihr.ac.uk/cmhtknowledgeexchange/a- whole-systems-approach#h.36hqenh1e01

  1. Department for Education. ‘Mental Health Support Teams: coverage and school

and college experience’ 2026. https://www.gov.uk/government/publications/mental- health-support-teams-coverage-and-school-and-college-experience

  1. Gillberg C. ‘The Essence in child psychiatry: Early Symptomatic Syndromes

Eliciting Neurodevelopmental Clinical Examinations’ Research in Developmental Disabilities 2010: volume 31, pages 1543-1551. https://doi.org/10.1016/j.ridd.2010.06.002

  1. HM Inspectorate of Prisons. ‘Children in custody 2024–25: an analysis of 12–18-

year-olds’ perceptions of their experiences in secure training centres and young offender institutions’ 2025. https://hmiprisons.justiceinspectorates.gov.uk/hmipris_reports/children-in-custody- 2024-25/

  1. Department for Education. ‘Permanent exclusions and suspensions in England:

2019 to 2020’ 2021. https://www.gov.uk/government/statistics/permanent-exclusions- and-suspensions-in-england-2019-to-2020

  1. Cornish R and Brennan I. ‘Exclusion from school and risk of serious violence: a

target trial emulation study’ International Journal of Population Data Science 2025: volume 10 (4). https://doi.org/10.23889/ijpds.v10i4.3124

  1. Williams K and others. ‘Prisoners’ childhood and family backgrounds: results

from the Surveying Prisoner Crime Reduction (SPCR) longitudinal cohort study of prisoners’ Ministry of Justice, 2012. https://www.gov.uk/government/publications/prisoners-childhood-and-family- backgrounds

  1. Early Intervention Foundation. ‘The cost of late intervention: EIF analysis 2016’

2016. https://foundations.org.uk/wp-content/uploads/2024/08/cost-of-late- intervention-2016-report-v2-Feb-2024.pdf

  1. Beek M. ‘The secure base model: promoting attachment and resilience in foster

care and adoption’ British Association for Adoption and Fostering, 2014. https://hdl.handle.net/20.500.14394/43015

  1. Gibson R and Whitelaw R. ‘Reimagining secure care: a literature review’

Children and Young People's Centre for Justice (CYCJ), University of Strathclyde, 2024. https://www.cycj.org.uk/wp-content/uploads/2024/10/Reimagining-Secure- Care-2-Lit-Review.pdf

  1. Oostermeijer S and others. ‘The case for small-scale, community-integrated,

therapeutic facilities: utility and feasibility for policy transfer to the Victorian youth justice system’ Journal of Criminology 2024: volume 57, pages 100-120. https://doi.org/10.1177/26338076231193503

  1. Rogers A and others. ‘The Framework for Integrated Care: a catalyst for change

to enhance services for children and young people at risk’ Anna Freud, 2025. https://activepartnerships.org/wp-content/uploads/2025/05/The-Framework-for- Integrated-Care.pdf

  1. Edbrooke-Childs J and others. ‘Independent evaluation of the Framework for

Integrated Care (Secure Stairs)’ Anna Freud National Centre for Children and Families, 2022. https://d1uw1dikibnh8j.cloudfront.net/media/15924/secure-stairs- independent-evaluation-full-report.pdf

  1. Royal College of Paediatrics and Child Health. ‘Healthcare standards for

children and young people in secure settings’ 2023. https://www.rcpch.ac.uk/sites/default/files/2023- 04/rcpch_healthcare_standards_for_children_and_young_people_in_secure_setting s_2023.pdf

  1. Chitsabesan P and others. ‘Using an objective computer task (QbTest) to aid the

identification of attention deficit hyperactivity disorder (ADHD) in the Children and Young People Secure Estate (CYPSE): a feasibility randomised controlled trial’ BMJ Open 2022: volume 12, page e064951. https://doi.org/10.1136/bmjopen-2022- 064951

  1. HM Inspectorate of Probation. ‘The experiences of black and mixed heritage

boys in the youth justice system: a thematic inspection by HM Inspectorate of Probation’ 2021. https://hmiprobation.justiceinspectorates.gov.uk/document/a- thematic-inspection-of-the-experiences-of-black-and-mixed-heritage-boys-in-the- youth-justice-system/

  1. Lammy D. ‘The Lammy Review: final report. An independent review into the

treatment of, and outcomes for, Black, Asian and Minority Ethnic individuals in the criminal justice system’ 2017. https://www.gov.uk/government/publications/lammy- review-final-report

  1. Wolpert M and others. ‘Child- and parent-reported outcomes and experience

from child and young people’s mental health services 2011–2015’ CAHMS Press, 2016. https://www.corc.uk.net/media/1544/0505207_corc-report_for-web.pdf

  1. Child Outcomes Research Consortium (CORC). https://www.corc.uk.net/
  2. National Collaborating Centre for Mental Health. ‘Patient reported outcome

measures (PROMs) in community mental health care: implementation guide’ 2023. https://www.rcpsych.ac.uk/improving-care/nccmh/service-design-and- development/proms-cmh-ig

  1. NIHR School of Public Health Research. ‘Public Mental Health Programme

Overview’ 2026. https://sphr.nihr.ac.uk/research/public-mental-health-programme- overview/

  1. Kirkbride JB and others. ‘The social determinants of mental health and disorder:

evidence, prevention and recommendations’ World Psychiatry 2024: volume 23, pages 58-90. https://doi.org/10.1002/wps.21160

  1. Fusar-Poli P and others. ‘Preventive psychiatry: a blueprint for improving the

mental health of young people’ World Psychiatry 2021: volume 20, pages 200-221. https://doi.org/10.1002/wps.20869

  1. McGorry PD and others. ‘The Lancet Psychiatry Commission on youth mental

health’ Lancet Psychiatry 2024: volume 11, pages 731-774. https://doi.org/10.1016/S2215-0366(24)00163-9

  1. Knapp M and others. ‘Mental health promotion and mental illness prevention:

the economic case’ Department of Health, 2011. https://www.gov.uk/government/publications/mental-health-promotion-and-mental- illness-prevention-the-economic-case

  1. Thomson RM and others. ‘How do income changes impact on mental health and

wellbeing for working-age adults? A systematic review and meta-analysis’ Lancet Public Health 2022: volume 7, pages e515-e528. https://doi.org/10.1016/S2468- 2667(22)00058-5

  1. Costello EJ and others. ‘Relationships between poverty and psychopathology: a

natural experiment’ JAMA 2003: volume 290, pages 2023-2029. https://doi.org/10.1001/jama.290.15.2023

  1. Forget EL. ‘The town with no poverty: the health effects of a Canadian

guaranteed annual income field experiment’ Canadian Public Policy 2011: volume 37, pages 283-305. https://doi.org/10.3138/cpp.37.3.283

  1. Haushofer J and Shapiro J. ‘The short-term impact of unconditional cash

transfers to the poor: experimental evidence from Kenya’ Quarterly Journal of Economics 2016: volume 131, pages 1973-2042. https://doi.org/10.1093/qje/qjw025

  1. Kaufman JA and others. ‘Effects of increased minimum wages by

unemployment rate on suicide in the USA’ Journal of Epidemiology and Community Health 2020: volume 74, pages 219-224. https://doi.org/10.1136/jech-2019-212981

  1. Wickham S and others. ‘Effects on mental health of a UK welfare reform,

Universal Credit: a longitudinal controlled study’ Lancet Public Health 2020: volume 5, pages e157-e164. https://doi.org/10.1016/S2468-2667(20)30026-8

  1. Aubry T and others. ‘A multiple-city RCT of Housing First with assertive

community treatment for homeless Canadians with serious mental illness’ Psychiatric Services 2016: volume 67, pages 275-281. https://doi.org/10.1176/appi.ps.201400587

  1. Axford N and others. ‘The Effectiveness of the KiVa bullying prevention program

in Wales, UK: results from a pragmatic cluster randomized controlled trial’ Prevention Science 2020: volume 21, pages 615-626. https://doi.org/10.1007/s11121-020- 01103-9

  1. Gaffney H and others. ‘Effectiveness of school-based programs to reduce

bullying perpetration and victimization: an updated systematic review and meta- analysis’ Campbell Systematic Reviews 2021: volume 17, page e1143. https://doi.org/10.1002/cl2.1143

  1. Kellam SG and others. ‘Effects of a universal classroom behavior management

program in first and second grades on young adult behavioral, psychiatric, and social outcomes’ Drug and Alcohol Dependence 2008: volume 95 Suppl 1, pages S5-S28. https://doi.org/10.1016/j.drugalcdep.2008.01.004

  1. Wasserman D and others. ‘School-based suicide prevention programmes: the

SEYLE cluster-randomised, controlled trial’ Lancet 2015: volume 385, pages 1536- 1544. https://doi.org/10.1016/S0140-6736(14)61213-7

  1. Orben A and Przybylski AK. ‘The association between adolescent well-being

and digital technology use’ Nature Human Behaviour 2019: volume 3, pages 173- 182. https://doi.org/10.1038/s41562-018-0506-1

  1. Odgers CL. ‘The great rewiring: is social media really behind an epidemic of

teenage mental illness?’ Nature 2024: volume 628, pages 29-30. https://doi.org/10.1038/d41586-024-00902-2

  1. Braghieri L and others. ‘Social media and mental health’ American Economic

Review 2022: volume 112, pages 3660-3693. https://doi.org/10.1257/aer.20211218

  1. Chen Q and others. ‘Digital empowerment in mental health: a meta-analysis of

internet-based interventions for enhancing mental health literacy’ International Journal of Clinical and Health Psychology 2024: volume 24, page 100489. https://doi.org/10.1016/j.ijchp.2024.100489

  1. Villa C. ‘The effects of youth clubs on education and crime’ Institute for Fiscal

Studies, 2024. https://ifs.org.uk/sites/default/files/2025-06/WP202451-The-effects-of- youth-clubs-on-education-and-crime.pdf

  1. Fenn N and others. ‘Examining the relationship between civic engagement and

mental health in young adults: a systematic review of the literature’ Journal of Youth Studies 2024: volume 27, pages 558-587. https://doi.org/10.1080/13676261.2022.2156779

  1. Kusier AO and others. ‘Understanding mental health promotion in organized

leisure communities for young people: a realist review’ Frontiers in Public Health 2024: volume 12, page 1336736. https://doi.org/10.3389/fpubh.2024.1336736

  1. Hugh-Jones S and Munford L. ‘The effects of engagement in arts and cultural

activities on adolescent mental health: results from a large UK panel study’ Social Science and Medicine 2025: volume 382, page 118343. https://doi.org/10.1016/j.socscimed.2025.118343

  1. Rodriguez-Ayllon M and others. ‘Role of physical activity and sedentary behavior

in the mental health of preschoolers, children and adolescents: a systematic review and meta-analysis’ Sports Medicine 2019: volume 49, pages 1383-1410. https://doi.org/10.1007/s40279-019-01099-5

  1. Ciacchella C and others. ‘Impact of structured physical activity on youth’s mental

health: A review of umbrella reviews’ International Journal of Social Psychiatry 2025: volume 71, pages 1219-1234. https://doi.org/10.1177/00207640251342425

  1. Panza M and others. ‘Associations between adolescent organized sport

participation and self-reported symptoms of anxiety and depression: A systematic review and meta-analysis’ Journal of Exercise, Movement, and Sport (SCAPPS refereed abstracts repository) 2019: volume 51. https://www.scapps.org/jems/index.php/1/en/article/view/2359

  1. Engemann K and others. ‘Residential green space in childhood is associated

with lower risk of psychiatric disorders from adolescence into adulthood’ Proceedings of the National Academy of Sciences of the United States of America 2019: volume 116, pages 5188-5193. https://doi.org/10.1073/pnas.1807504116

  1. South EC and others. ‘Effect of greening vacant land on mental health of

community-dwelling adults: a cluster randomized trial’ JAMA Network Open 2018: volume 1, page e180298. https://doi.org/10.1001/jamanetworkopen.2018.0298

  1. Bickerdike L and others. ‘Social prescribing: less rhetoric and more reality. A

systematic review of the evidence’ BMJ Open 2017: volume 7, page e013384. https://doi.org/10.1136/bmjopen-2016-013384

  1. Mann F and others. ‘A life less lonely: the state of the art in interventions to

reduce loneliness in people with mental health problems’ Social Psychiatry and Psychiatric Epidemiology 2017: volume 52, pages 627-638. https://doi.org/10.1007/s00127-017-1392-y

  1. Harvey SB and others. ‘Exercise and the prevention of depression: results of the

HUNT cohort study’ American Journal of Psychiatry 2018: volume 175, pages 28-36. https://doi.org/10.1176/appi.ajp.2017.16111223

  1. Schuch FB and others. ‘Physical activity and incident depression: a meta-

analysis of prospective cohort studies’ American Journal of Psychiatry 2018: volume 175, pages 631-648. https://doi.org/10.1176/appi.ajp.2018.17111194

  1. Menting AT and others. ‘Effectiveness of the Incredible Years parent training to

modify disruptive and prosocial child behavior: a meta-analytic review’ Clinical Psychology Review 2013: volume 33, pages 901-913. https://doi.org/10.1016/j.cpr.2013.07.006

  1. Leijten P and others. ‘Meta-analyses: key parenting program components for

disruptive child behavior’ Journal of the American Academy of Child and Adolescent Psychiatry 2019: volume 58, pages 180-190. https://doi.org/10.1016/j.jaac.2018.07.900

  1. Wilson P and others. ‘How evidence-based is an ‘evidence-based parenting

program’? A PRISMA systematic review and meta-analysis of Triple P’ BMC Medicine 2012: volume 10, page 130. https://doi.org/10.1186/1741-7015-10-130

  1. Pickles A and others. ‘Parent-mediated social communication therapy for young

children with autism (PACT): long-term follow-up of a randomised controlled trial’ Lancet 2016: volume 388, pages 2501-2509. https://doi.org/10.1016/S0140- 6736(16)31229-6

  1. O’Connor E and others. ‘Interventions to prevent perinatal depression: evidence

report and systematic review for the US Preventive Services Task Force’ JAMA 2019: volume 321, pages 588-601. https://doi.org/10.1001/jama.2018.20865

  1. Robling M and others. ‘Effectiveness of a nurse-led intensive home-visitation

programme for first-time teenage mothers (Building Blocks): a pragmatic randomised controlled trial’ Lancet 2016: volume 387, pages 146-155. https://doi.org/10.1016/S0140-6736(15)00392-X

  1. Caspi A and others. ‘Longitudinal assessment of mental health disorders and

comorbidities across 4 decades among participants in the Dunedin Birth Cohort Study’ JAMA Network Open 2020: volume 3, page e203221. https://doi.org/10.1001/jamanetworkopen.2020.3221

  1. Ramchandani P and others. ‘A systematic review of treatments for settling

problems and night waking in young children’ BMJ 2000: volume 320, pages 209- 213. https://doi.org/10.1136/bmj.320.7229.209

  1. Madonna M and others. ‘Caregiver training improves child feeding behaviours in

children with paediatric feeding disorder and may reduce caregiver stress: a systematic review and meta-analysis’ International Journal of Speech-Language Pathology 2025: volume 27, pages 634-646. https://doi.org/10.1080/17549507.2024.2381459

  1. Creswell C and others. ‘Parent-centered interventions for childhood anxiety’

Annual Review of Developmental Psychology 2025: volume 7, pages 383-410. https://doi.org/10.1146/annurev-devpsych-111323-112637

  1. Gardner F and others. ‘The earlier the better? Individual participant data and

traditional meta-analysis of age effects of parenting interventions’ Child Development 2019: volume 90, pages 7-19. https://doi.org/10.1111/cdev.13138

  1. Minnis H. ‘Parenting groups can be effective across the social spectrum: the

next step is access’ Lancet Psychiatry 2019: volume 6, pages 458-459. https://doi.org/10.1016/S2215-0366(19)30165-8

  1. Jones SH and others. ‘Web-based integrated bipolar parenting intervention for

parents with bipolar disorder: a randomised controlled pilot trial’ Journal of Child Psychology and Psychiatry 2017: volume 58, pages 1033-1041. https://doi.org/10.1111/jcpp.12745

  1. Salari R and others. ‘An efficacy trial: positive parenting program for parents of

teenagers’ Behaviour Change 2014: volume 31, pages 34-52. https://doi.org/10.1017/bec.2013.31

  1. Creswell C and others. ‘Clinical outcomes and cost-effectiveness of brief guided

parent-delivered cognitive behavioural therapy and solution-focused brief therapy for treatment of childhood anxiety disorders: a randomised controlled trial’ Lancet Psychiatry 2017: volume 4, pages 529-539. https://doi.org/10.1016/S2215- 0366(17)30149-9

  1. Yu S and others. ‘Estimating the economic value of supported online parent-led

cognitive behavioural therapy to prevent child anxiety problems: findings from the MYCATS cluster randomised controlled trial’ Journal of Medical Internet Research in press. https://hdl.handle.net/10779/exe.33201207

  1. Murray L and others. ‘Effects of postnatal depression on mother–infant

interactions and child development’. In: Bremner J and Wachs, T, editors. ‘The Wiley-Blackwell handbook of infant development, volume 2, 2nd edition’ Wiley, 2010. https://doi.org/10.1002/9781444327588.ch8

  1. Cuijpers P and others. ‘The effects of psychological treatment of maternal

depression on children and parental functioning: a meta-analysis’ European Child & Adolescent Psychiatry 2015: volume 24, pages 237-245. https://doi.org/10.1007/s00787-014-0660-6

  1. Fleming WJ. ‘Employee well‐being outcomes from individual‐level mental health

interventions: cross‐sectional evidence from the United Kingdom’ Industrial Relations Journal 2024: volume 55, pages 162-182. https://doi.org/10.1111/irj.12418

  1. Frederick DE and VanderWeele TJ. ‘Supported employment: meta-analysis and

review of randomized controlled trials of individual placement and support’ Plos One 2019: volume 14, page e0212208. https://doi.org/10.1371/journal.pone.0212208

  1. Rapee RM. ‘The preventative effects of a brief, early intervention for preschool-

aged children at risk for internalising: follow-up into middle adolescence’ Journal of Child Psychology and Psychiatry 2013: volume 54, pages 780-788. https://doi.org/10.1111/jcpp.12048

  1. Hawton K and others. ‘Long term effect of reduced pack sizes of paracetamol on

poisoning deaths and liver transplant activity in England and Wales: interrupted time series analyses’ BMJ 2013: volume 346, page f403. https://doi.org/10.1136/bmj.f403

  1. Pirkis J and others. ‘Interventions to reduce suicides at suicide hotspots: a

systematic review and meta-analysis’ Lancet Psychiatry 2015: volume 2, pages 994- 1001. https://doi.org/10.1016/S2215-0366(15)00266-7

  1. Kreitman N. ‘The coal gas story. United Kingdom suicide rates, 1960-71’ British

Journal of Preventive and Social Medicine 1976: volume 30, pages 86-93. https://doi.org/10.1136/jech.30.2.86

  1. Office for National Statistics. ‘Adult smoking habits in the UK: 2021’ 2022.

https://backup.ons.gov.uk/wp-content/uploads/sites/3/2022/12/Adult-smoking-habits- in-the-UK-2021.pdf

  1. Wyper GMA and others. ‘Evaluating the impact of alcohol minimum unit pricing

on deaths and hospitalisations in Scotland: a controlled interrupted time series study’ Lancet 2023: volume 401, pages 1361-1370. https://doi.org/10.1016/S0140- 6736(23)00497-X

  1. Kristjansson AL and others. ‘Development and guiding principles of the Icelandic

Model for preventing adolescent substance use’ Health Promotion Practice 2020: volume 21, pages 62-69. https://doi.org/10.1177/1524839919849032

  1. Foulkes L and Andrews JL. ‘Are mental health awareness efforts contributing to

the rise in reported mental health problems? A call to test the prevalence inflation hypothesis’ New Ideas in Psychology 2023: volume 69, page 101010. https://doi.org/10.1016/j.newideapsych.2023.101010

  1. Harvey LJ and others. ‘Investigating the efficacy of a dialectical behaviour

therapy-based universal intervention on adolescent social and emotional well-being outcomes’ Behaviour Research and Therapy 2023: volume 169, page 104408. https://doi.org/10.1016/j.brat.2023.104408

  1. Rose S and others. ‘Psychological debriefing for preventing post traumatic

stress disorder (PTSD)’ Cochrane Database of Systematic Reviews 2002: page CD000560. https://doi.org/10.1002/14651858.CD000560

  1. Petrosino A and others. ‘‘Scared Straight’ and other juvenile awareness

programs for preventing juvenile delinquency’ Cochrane Database of Systematic Reviews 2013: volume 2013, page CD002796. https://doi.org/10.1002/14651858.CD002796.pub2

  1. Gould MS and others. ‘Evaluating iatrogenic risk of youth suicide screening

programs: a randomized controlled trial’ JAMA 2005: volume 293, pages 1635-1643. https://doi.org/10.1001/jama.293.13.1635

  1. Rai D and others. ‘Common mental disorders, subthreshold symptoms and

disability: longitudinal study’ British Journal of Psychiatry 2010: volume 197, pages 411-412. https://doi.org/10.1192/bjp.bp.110.079244

  1. Kellezi B and others. ‘Mental health and other factors associated with work

productivity after injury in the UK: multicentre cohort study’ Injury Prevention 2022: volume 28, pages 131-140. https://doi.org/10.1136/injuryprev-2021-044311

  1. Health and Safety Executive. ‘Work-related stress, depression or anxiety

statistics in Great Britain, 2025’ 2025. https://www.hse.gov.uk/statistics/assets/docs/stress.pdf

  1. Zaidi K and others. ‘Autism policy and services: employment’ House of

Commons Library, 2025. https://commonslibrary.parliament.uk/research- briefings/cbp-10389/

  1. Baloch A. ‘Closing the youth jobs gap: how to ensure the Youth Guarantee

reaches young people from disadvantaged backgrounds’ Impetus, 2026. https://impetus-org.files.svdcdn.com/production/assets/publications/Report/Impetus- Closing-the-Youth-Jobs-Gap-2026.pdf?dm=1784035999

  1. Coles B and others. ‘Estimating the life-time cost of NEET: 16-18 year olds not

in education, employment or training. Research undertaken for the Audit Commission’ University of York, 2010. http://www.york.ac.uk/inst/spru/research/pdf/NEET.pdf

  1. Wels J and others. ‘Rising rates of young people Not in Education, Employment,

or Training (NEET) explained by higher prevalence of physical and psychological ill health: a 15-year UK study’ medRxiv 2026. https://doi.org/10.64898/2026.08.11.26360217

  1. Report commissioned by and submitted to the Review using data from:

University of Essex Institute for Social and Economic Research. ‘Understanding Society: Waves 1-15, 2009-2024 and Harmonised BHPS: Waves 1-18, 1991-2009: Secure Access [data collection]. ’ UK Data Service. SN: 6676, 2026. https://doi.org/10.5255/UKDA-SN-6676-19

  1. Report commissioned by and submitted to the Review using data from the

Millennium Cohort Study. UCL Centre for Longitudinal Studies, 2026. https://cls.ucl.ac.uk/cls-studies/millennium-cohort-study/

  1. Lewis J and S. S. ‘Student mental health in England: statistics, policy, and

guidance’ House of Commons Library, 2025. https://commonslibrary.parliament.uk/research-briefings/cbp-8593

  1. Department for Science, Innovation & Technology. ‘A snapshot of entry-level

hiring in the UK’ 2026. https://www.gov.uk/government/publications/entry-level- hiring-in-the-uk-a-snapshot/a-snapshot-of-entry-level-hiring-in-the-uk

  1. Stea TH and others. ‘Interventions targeting young people not in employment,

education or training (NEET) for increased likelihood of return to school or employment—a systematic review’ Plos One 2024: volume 19, page e0306285. https://doi.org/10.1371/journal.pone.0306285

  1. Wale A and others. ‘What is the effectiveness of interventions to support the

mental and emotional health and wellbeing of young people who are not in education, employment or training (NEET)? A rapid evidence review’ medRxiv 2025: page 2025.2010.2029.25339049. https://doi.org/10.1101/2025.10.29.25339049

  1. Wattanawongwan S and others. ‘Interventions for improving social-

communication skills for adolescents and adults with ASD: a meta-analysis’ Review Journal of Autism and Developmental Disorders 2022: volume 10, pages 391-405. https://doi.org/10.1007/s40489-021-00300-8

  1. Office for National Statistics. ‘Labour Force Survey, 2012 to 2024’.

https://www.ons.gov.uk/

  1. Vlasveld MC. ‘Sickness absence and return to work in workers with major

depressive disorder: The Netherlands Depression Initiative in the occupational healthcare setting’ PhD thesis, Vrije Universiteit Amsterdam, 2012. https://pure.amsterdamumc.nl/en/publications/sickness-absence-and-return-to-work- in-workers-with-major-depress/

  1. Etuknwa A and others. ‘Sustainable return to work: a systematic review focusing

on personal and social factors’ Journal of Occupational Rehabilitation 2019: volume 29, pages 679-700. https://doi.org/10.1007/s10926-019-09832-7

  1. Lamb D and others. ‘Psychosocial impact of the COVID-19 pandemic on 4378

UK healthcare workers and ancillary staff: initial baseline data from a cohort study collected during the first wave of the pandemic’ Occupational and Environmental Medicine 2021: volume 78, pages 801-808. https://doi.org/10.1136/oemed-2020- 107276

  1. Kromydas T and others. ‘Comparing population-level mental health of UK

workers before and during the COVID-19 pandemic: a longitudinal study using Understanding Society’ Journal of Epidemiology and Community Health 2022: volume 76, pages 527-536. https://doi.org/10.1136/jech-2021-218561

  1. Lalloo D and others. ‘Comparing anxiety and depression in information

technology workers with others in employment: a UK Biobank cohort study’ Annals of Work Exposures and Health 2022: volume 66, pages 1136-1150. https://doi.org/10.1093/annweh/wxac061

  1. Farina E and others. ‘Zero hours contracts and self-reported (mental) health in

the UK’ British Journal of Industrial Relations 2024: volume 62, pages 50-71. https://doi.org/10.1111/bjir.12773

  1. Bhui KS and others. ‘A synthesis of the evidence for managing stress at work: a

review of the reviews reporting on anxiety, depression, and absenteeism’ Journal of Environmental and Public Health 2012: volume 2012, page 515874. https://doi.org/10.1155/2012/515874

  1. Nielsen MB and Knardahl S. ‘The impact of office design on medically certified

sickness absence’ Scandinavian Journal of Work, Environment & Health 2020: volume 46, pages 330-334. https://doi.org/10.5271/sjweh.3859

  1. Roelen CA and others. ‘Employees sick-listed with mental disorders: who

returns to work and when?’ Journal of Occupational Rehabilitation 2012: volume 22, pages 409-417. https://doi.org/10.1007/s10926-012-9363-3

  1. Brouwers EP and others. ‘Predicting return to work in employees sick-listed due

to minor mental disorders’ Journal of Occupational Rehabilitation 2009: volume 19, pages 323-332. https://doi.org/10.1007/s10926-009-9198-8

  1. Haslam C and others. ‘Anxiety and depression in the workplace: effects on the

individual and organisation (a focus group investigation)’ Journal of Affective Disorders 2005: volume 88, pages 209-215. https://doi.org/10.1016/j.jad.2005.07.009

  1. Duchaine CS and others. ‘Psychosocial stressors at work and the risk of

sickness absence due to a diagnosed mental disorder: a systematic review and meta-analysis’ JAMA Psychiatry 2020: volume 77, pages 842-851. https://doi.org/10.1001/jamapsychiatry.2020.0322

  1. Nieuwenhuijsen K and others. ‘Interventions to improve return to work in

depressed people’ Cochrane Database of Systematic Reviews 2020: volume 10, page CD006237. https://doi.org/10.1002/14651858.CD006237.pub4

  1. Reme SE and others. ‘Work-focused cognitive-behavioural therapy and

individual job support to increase work participation in common mental disorders: a randomised controlled multicentre trial’ Occupational and Environmental Medicine 2015: volume 72, pages 745-752. https://doi.org/10.1136/oemed-2014-102700

  1. Cullen KL and others. ‘Effectiveness of workplace interventions in return-to-work

for musculoskeletal, pain-related and mental health conditions: an update of the evidence and messages for practitioners’ Journal of Occupational Rehabilitation 2018: volume 28, pages 1-15. https://doi.org/10.1007/s10926-016-9690-x

  1. Bond GR and others. ‘A systematic review and meta-analysis of IPS supported

employment for young adults with mental health conditions’ Administration and Policy in Mental Health 2023: volume 50, pages 160-172. https://doi.org/10.1007/s10488-022-01228-9

  1. Gabbay M and others. ‘Sickness certification for common mental disorders and

GP return-to-work advice’ Primary Health Care Research & Development 2016: volume 17, pages 437-447. https://doi.org/10.1017/s1463423616000074

  1. Legha A and others. ‘Trajectories of work absence in England due to a

musculoskeletal or mental health condition: an electronic health record cohort study’ Journal of Occupational Rehabilitation 2025. https://doi.org/10.1007/s10926-025- 10342-y

  1. Reuschke D and others. ‘Impacts of long Covid on workers: a longitudinal

study of employment exit, work hours and mental health in the UK’ Plos One 2024: volume 19, page e0306122. https://doi.org/10.1371/journal.pone.0306122

  1. Li LZ and Wang S. ‘Do work-family initiatives improve employee mental health?

Longitudinal evidence from a nationally representative cohort’ Journal of Affective Disorders 2022: volume 297, pages 407-414. https://doi.org/10.1016/j.jad.2021.10.112

  1. Chandola T and Rouxel P. ‘The role of workplace accommodations in explaining

the disability employment gap in the UK’ Social Science and Medicine 2021: volume 285, page 114313. https://doi.org/10.1016/j.socscimed.2021.114313

  1. National Audit Office. ‘The Access to Work scheme’ 2026.

https://www.nao.org.uk/reports/the-access-to-work-scheme/

  1. Akhanemhe R and others. ‘An evaluation of REACTMH mental health training

for healthcare supervisors’ Occupational Medicine 2021: volume 71, pages 127-130. https://doi.org/10.1093/occmed/kqab023

  1. Logistics UK. ‘Skills and Employment Update: Q3 2024’ Logistics UK, 2024.

932 Health and Safety Executive. ‘Tackling work-related stress using the Management Standards approach: a step-by-step workbook’ 2019. https://www.hse.gov.uk/pubns/wbk01.htm

933 Institute for Employment Studies. ‘Flexible working for all: achieving greater equity for frontline and site-based workers’ 2026. https://www.employment- studies.co.uk/resource/flexible-working-for-all-achieving-greater-equity-for-frontline- and-site-based-workers/

934 Chartered Institute of Personnel and Development. ‘Flexible and hybrid working practices in 2025’ 2025. https://www.cipd.org/uk/knowledge/reports/flexible-hybrid- working/

935 Acas. ‘Implementing flexible working’ 2026. https://www.acas.org.uk/implementing-flexible-working

936 Mates in Mind. ‘Mental health in construction: joint code of practice (JCOP)’ 2025. https://www.matesinmind.org/what-we-do/research-and-advocacy/mental- health-in-construction-joint-code-of-practice-jcop

937 Skills for Care. ‘The state of the adult social care sector and workforce in England, 2025’ Skills for Care, 2025. https://www.skillsforcare.org.uk/Adult-Social- Care-Workforce-Data/workforceintelligence/resources/Reports/National/The-state-of- the-adult-social-care-sector-and-workforce-in-England-2025.pdf

938 UKHospitality. ‘The UK hospitality sector and sustainability in 2024’ 2024. https://www.ukhospitality.org.uk/the-uk-hospitality-sector-and-sustainability-in-2024/

939 British Retail Consortium. ‘Retail down by over 360,000 jobs in a decade’ 2025. https://brc.org.uk/news-and-events/news/corporate-affairs/2025/ungated/retail-down- by-over-360-000-jobs-in-a-decade/

940 Department for Transport. ‘Taxi and private hire vehicle statistics, England: 2026’ 2026. https://www.gov.uk/government/statistics/taxi-and-private-hire-vehicle- statistics-england-2026

941 Department for Transport. ‘Annual bus statistics: year ending March 2025 (revised)’ 2026. https://www.gov.uk/government/statistics/annual-bus-statistics-year- ending-march-2025/annual-bus-statistics-year-ending-march-2025

942 NHS England. ‘ADHD Taskforce’ 2026. https://www.england.nhs.uk/mental- health/adhd/

943 National Autistic Society. ‘A guide to supporting neurodivergent colleagues: reasonable adjustments and resources for employers’ 2026. https://media.autism.org.uk/asset/8df4c3dc-45de-4fee-9afc-54f7b544b0a3/National- Autistic-Society-WAAM26-Employment-Toolkit.pdf

944 NHS England. ‘Fit Notes Issued by GP Practices, England, December 2025’ 2026. https://digital.nhs.uk/data-and-information/publications/statistical/fit-notes- issued-by-gp-practices/december-2025

945 Viola S and Moncrieff J. ‘Claims for sickness and disability benefits owing to mental disorders in the UK: trends from 1995 to 2014’ BJPsych Open 2016: volume 2, pages 18-24. https://doi.org/10.1192/bjpo.bp.115.002246

946 Cribb J and others. ‘Living standards of working-age disability benefits recipients in the UK’ Institute for Fiscal Studies, 2022. https://ifs.org.uk/publications/living- standards-working-age-disability-benefits-recipients-uk

947 Latimer E and others. ‘Health-related benefit claims post-pandemic: UK trends and global context’ Institute for Fiscal Studies, 2024. https://ifs.org.uk/publications/health-related-benefit-claims-post-pandemic-uk-trends- and-global-context

948 Latimer E and others. ‘The role of changing health in rising health-related benefit claims’ Institute for Fiscal Studies, 2025. https://ifs.org.uk/publications/role-changing- health-rising-health-related-benefit-claims

949 World Health Organization. ‘From loneliness to social connection: charting a path to healthier societies. Report of the WHO Commission on Social Connection’ 2025. https://www.who.int/publications/i/item/978240112360

950 Jenkinson CE and others. ‘Is volunteering a public health intervention? A systematic review and meta-analysis of the health and survival of volunteers’ BMC Public Health 2013: volume 13, page 773. https://doi.org/10.1186/1471-2458-13-773

951 Singh B and others. ‘Effectiveness of physical activity interventions for improving depression, anxiety and distress: an overview of systematic reviews’ British Journal of Sports Medicine 2023: volume 57, pages 1203-1209. https://doi.org/10.1136/bjsports-2022-106195

952 Jacka FN and others. ‘A randomised controlled trial of dietary improvement for adults with major depression (the ‘Smiles’ trial)’ BMC Medicine 2017: volume 15, page 23. https://doi.org/10.1186/s12916-017-0791-y

953 White MP and others. ‘Spending at least 120 minutes a week in nature is associated with good health and wellbeing’ Scientific Reports 2019: volume 9, page 7730. https://doi.org/10.1038/s41598-019-44097-3

954 Henderson C and others. ‘Social inclusion of people with severe mental illness: a review of current practices, evidence and unmet needs, and future directions’ World Psychiatry 2026: volume 25, pages 56-82. https://doi.org/10.1002/wps.70031

955 Fenney D and others. ‘Caring in a complex world: perspectives from unpaid carers and the organisations that support them’ The King's Fund, 2023. https://assets.kingsfund.org.uk/f/256914/x/1f6607287b/caring_in_a_complex_world_ unpaid_carers_2023.pdf

956 NHS England. ‘Social prescribing’ 2023. https://www.england.nhs.uk/personalisedcare/social-prescribing/

957 Department for Education. ‘Children looked after in England including adoptions: Reporting year 2025’ 2025. https://explore-education-statistics.service.gov.uk/find- statistics/children-looked-after-in-england-including-adoptions/2025

958 Cummings A and Shelton K. ‘The prevalence of mental health disorders amongst care-experienced young people in the UK: a systematic review’ Children and Youth Services Review 2024: volume 156, page 107367. https://doi.org/10.1016/j.childyouth.2023.107367

959 Carter B and others. ‘The mental health and wellbeing of care-experienced young people during early and later adolescence’ Clinical Child Psychology and Psychiatry 2025: volume 30, pages 611-631. https://doi.org/10.1177/13591045251333028

960 Sprecher EA and others. ‘Sufficiency of current practice: how well does the Strengths and Difficulties Questionnaire detect clinically elevated posttraumatic stress, anxiety, and depression symptoms in children in care?’ JCPP Advances 2026: volume 6, page e70058. https://doi.org/10.1002/jcv2.70058

961 Baker C and Briheim-Crookall L. ‘Disability, disparity and demand: analysis of the numbers and experiences of children in care and care leavers with a disability or long- term health condition’ Coram, 2024. https://www.coram.org.uk/resource/disability-disparity-and-demand/

962 National Children’s Bureau. ‘Children in most need of specialist mental health support often turned away because their situation is too unstable’ 2024. https://www.ncb.org.uk/about-us/media-centre/news-opinion/children-most-need- specialist-mental-health-support-often-turned

963 Phillips AR and others. ‘Secondary data analysis of social care records to examine the provision of mental health support for young people in care’ JCPP Advances 2023: volume 3, page e12161. https://doi.org/10.1002/jcv2.12161

964 Evans R and others. ‘Interventions targeting the mental health and well-being of care-experienced children and young people: mixed-methods systematic review with stakeholder consultation to inform transportability and adaptability to UK context’ British Journal of Social Work 2024: volume 54, pages 3590-3618. https://doi.org/10.1093/bjsw/bcae061

965 Nuffield Family Justice Observatory. ‘New data shows a high of 1,440 deprivation of liberty applications in 2025’ 2026. https://www.nuffieldfjo.org.uk/news/new-data-shows-a-high-of-1440-deprivation-of- liberty-dol-applications-in-2025

966 Ministry of Justice. ‘Family Court Statistics Quarterly: October to December 2025’ 2026. https://www.gov.uk/government/statistics/family-court-statistics- quarterly-october-to-december-2025/family-court-statistics-quarterly-october-to- december-2025

967 National Audit Office. ‘Managing children’s residential care’ 2025. https://www.nao.org.uk/wp-content/uploads/2025/09/Managing-childrens-residential- care.pdf

968 Children’s Commissioner for England. ‘Children living in illegal children’s homes’ 2026. https://assets.childrenscommissioner.gov.uk/wpuploads/2026/01/cc-illegal- childrens-homes-Jan-26.pdf

969 Department for Education. ‘Government to examine deaths of vulnerable care leavers’ 2026. https://www.gov.uk/government/news/government-to-examine-deaths- of-vulnerable-care-leavers

970 Hiller R and others. ‘MyVoice: mental health assessments for children in care’ ISRCTN Registry 2026. https://doi.org/10.1186/ISRCTN18231994

971 O’Logbon J and others. ‘How does the increase in eating difficulties according to the Development and Well-Being Assessment screening items relate to the population prevalence of eating disorders? An analysis of the 2017 Mental Health in Children and Young People survey’ International Journal of Eating Disorders 2022: volume 55, pages 1777-1787. https://doi.org/10.1002/eat.23833

972 Micali N and others. ‘The incidence of eating disorders in the UK in 2000-2009: findings from the General Practice Research Database’ BMJ Open 2013: volume 3, page e002646. https://doi.org/10.1136/bmjopen-2013-002646

973 Trafford AM and others. ‘Temporal trends in eating disorder and self-harm incidence rates among adolescents and young adults in the UK in the 2 years since onset of the COVID-19 pandemic: a population-based study’ Lancet Child & Adolescent Health 2023: volume 7, pages 544-554. https://doi.org/10.1016/S2352- 4642(23)00126-8

974 Wilkins J and others. ‘Primary care use among adults with eating disorders in England: a population-based cohort study using electronic health records’ BMJ Open 2026: volume 16, page e119143. https://doi.org/10.1136/bmjopen-2026-119143

975 Ornstein RM and others. ‘Distribution of eating disorders in children and adolescents using the proposed DSM-5 criteria for feeding and eating disorders’ Journal of Adolescent Health 2013: volume 53, pages 303-305. https://doi.org/10.1016/j.jadohealth.2013.03.025

976 NHS England. ‘NHS staff to train teachers, school nurses, and GPs to spot eating disorders’ 2026. https://www.england.nhs.uk/2026/01/nhs-staff-to-train- teachers-school-nurses-and-gps-to-spot-eating-disorders/

977 Royal College of Psychiatrists. ‘National audit of eating disorders: service mapping report 2025’ 2025. https://www.rcpsych.ac.uk/docs/default- source/improving-care/ccqi/national-clinical-audits/national-audit-of-eating-disorders- (naed)/naed-audit-resources/naed-service-mapping-report- 2025.pdf?sfvrsn=cf162632_3

978 Austin A and others. ‘Duration of untreated eating disorder and relationship to outcomes: a systematic review of the literature’ European Eating Disorders Review 2021: volume 29, pages 329-345. https://doi.org/10.1002/erv.2745

979 Department of Health and Social Care. ‘Eating disorders: Children and young people (Written Question UIN 105363)’ UK Parliament, 2026. https://questions- statements.parliament.uk/written-questions/detail/2026-01-13/105363

980 Royal College of Psychiatrists. ‘RCPsych uncovers £2.1 million planned spending cuts for children’s eating disorder services in England’ 2026. https://www.rcpsych.ac.uk/news-and-features/latest-news/detail/2026/02/26/rcpsych- uncovers--2.1million-planned-spending-cuts-for-children-s-eating-disorder-services- in-england

981 Health Innovation Network South London. ‘Early intervention eating disorders: AHSN national programme final report’ 2023.

https://healthinnovationnetwork.com/projects/national-programme-early-intervention- eating-disorders/

982 Sanchez-Cerezo J and others. ‘Incidence of avoidant/restrictive food intake disorder in children and adolescents across the UK and Ireland: a BPSU and CAPSS surveillance study’ BMJ Open 2024: volume 14, page e088129. https://doi.org/10.1136/bmjopen-2024-088129

983 Ministry of Justice. ‘Cutting youth crime. Changing young lives. The youth justice system reform and delivery plan’ 2026. https://committees.parliament.uk/publications/53118/documents/297178/default/

984 Youth Justice Board for England and Wales. ‘Record lows in Youth Justice Annual Statistics 2024-25’ 2026. https://www.gov.uk/government/news/record-lows- in-youth-justice-annual-statistics-2024-25

985 Youth Justice Board for England and Wales. ‘Youth Justice Statistics: 2023 to 2024’ 2025. https://www.gov.uk/government/statistics/youth-justice-statistics-2023- to-2024/youth-justice-statistics-2023-to-2024

986 Bryan K. ‘Preliminary study of the prevalence of speech and language difficulties in young offenders’ International Journal of Language & Communication Disorders 2004: volume 39, pages 391-400. https://doi.org/10.1080/13682820410001666376

987 Chester V and others. ‘Autistic people within forensic psychiatric services and the criminal justice system: a systematic review’ Journal of Forensic Psychiatry & Psychology 2025: volume 36, pages 711-773. https://doi.org/10.1080/14789949.2025.2558833

988 Bryan K and others. ‘Language and communication difficulties in juvenile offenders’ International Journal of Language & Communication Disorders 2007: volume 42, pages 505-520. https://doi.org/10.1080/13682820601053977

989 Ministry of Justice. ‘Safety in the Children and Young People Secure Estate: Update to March 2026’ 2026. https://www.gov.uk/government/statistics/safety-in-the- children-and-young-people-secure-estate-update-to-march-2026/safety-in-the- children-and-young-people-secure-estate-update-to-march-2026

990 Kien C and others. ‘Prevalence of mental disorders in young refugees and asylum seekers in European countries: a systematic review’ European Child & Adolescent Psychiatry 2019: volume 28, pages 1295-1310. https://doi.org/10.1007/s00787-018-1215-z

991 Joseph S and others. ‘Young carers research, practice and policy: an overview and critical perspective on possible future directions’ Adolescent Research Review 2019: volume 5, pages 77-89. https://doi.org/10.1007/s40894-019-00119-9

992 Fleitas Alfonzo L and others. ‘Mental health of young informal carers: a systematic review’ Social Psychiatry and Psychiatric Epidemiology 2022: volume 57, pages 2345-2358. https://doi.org/10.1007/s00127-022-02333-8

993 Centrepoint. ‘The mental health needs of homeless young people’ 2021. https://centrepoint.org.uk/sites/default/files/2023-06/prevalence-of-mental-health- need-report.pdf

994 Sharp M and others. ‘Health and Wellbeing Study of Serving and Ex-Serving UK Armed Forces Personnel: Phase 4. Office for Veterans’ Affairs Final Report’ King’s Centre for Military Health Research, 2024. https://kcmhr.org/pdf/Phase_4_Health_and_Wellbeing_Cohort_Study_Report.pdf

995 Cochrane H and Soni A. ‘Education, health and care plans: what do we know so far?’ Support for Learning 2020: volume 35, pages 372-388. https://doi.org/10.1111/1467-9604.12316

996 Castro-Kemp S and Samuels A. ‘Working together: a review of cross-sector collaborative practices in provision for children with special educational needs and disabilities’ Research in Developmental Disabilities 2022: volume 120, page 104127. https://doi.org/10.1016/j.ridd.2021.104127

997 Sales N and Vincent K. ‘Strengths and limitations of the Education, Health and Care plan process from a range of professional and family perspectives’ British Journal of Special Education 2018: volume 45, pages 61-80. https://doi.org/10.1111/1467-8578.12202

998 Barnett P and others. ‘‘Five hours to sort out your life’: qualitative study of the experiences of university students who access mental health support’ BJPsych Open 2021: volume 7, page e118. https://doi.org/10.1192/bjo.2021.947

999 Shanahan P and others. ‘Experiences of transition from children’s to adult’s healthcare services for young people with a neurodevelopmental condition’ Health & Social Care in the Community 2021: volume 29, pages 1429-1438. https://doi.org/10.1111/hsc.13198

1000 Hollenweger J. ‘Development of an ICF-based eligibility procedure for education in Switzerland’ BMC Public Health 2011: volume 11, page S7. https://doi.org/10.1186/1471-2458-11-S4-S7

1001 Sanches-Ferreira M and others. ‘Conditions for implementing the ICF-CY in education: the experience in Portugal’ Frontiers in Education 2018: volume 3, page

  • https://doi.org/10.3389/feduc.2018.00020

1002 Leonardi M and others. ‘20 Years of ICF—International Classification of Functioning, Disability and Health: uses and applications around the world’ International Journal of Environmental Research and Public Health 2022: volume 19, page 11321. https://doi.org/10.3390/ijerph191811321

1003 Sanches-Ferreira M and others. ‘Portugal's special education law: implementing the International Classification of Functioning, Disability and Health in policy and practice’ Disability and Rehabilitation 2013: volume 35, pages 868-873. https://doi.org/10.3109/09638288.2012.708816

1004 Silveira-Maia M and others. ‘How the use of the International Classification of Functioning, Disability and Health for Children and Youth changed the Individualized Education Programs in Portugal’ International Journal of Inclusive Education 2016: volume 21, pages 573-583. https://doi.org/10.1080/13603116.2016.1218950

1005 Selb M and others. ‘A guide on how to develop an International Classification of Functioning, Disability and Health Core Set’ European Journal of Physical and Rehabilitation Medicine 2015: volume 51, pages 105-117. https://www.minervamedica.it/en/journals/europa- medicophysica/article.php?cod=R33Y2015N01A0105

1006 Bölte S and others. ‘The Gestalt of functioning in autism revisited: First revision of the International Classification of Functioning, Disability and Health Core Sets’ Autism 2024: volume 28, pages 2394-2411. https://doi.org/10.1177/13623613241228896

1007 Bölte S and others. ‘Assessment of functioning in ADHD according to World Health Organization standards: First revision of the International Classification of Functioning, Disability and Health Core Sets’ Developmental Medicine & Child Neurology 2024: volume 66, pages 1201-1214. https://doi.org/10.1111/dmcn.15865

1008 Black MH and others. ‘From symptomatology to functioning - applying the ICF to autism measures to facilitate neurodiversity-affirmative data harmonization’

Journal of Autism and Developmental Disorders 2025: volume 55, pages 114-129. https://doi.org/10.1007/s10803-023-06204-2

1009 D’Arcy E and others. ‘Content validation of common measures of functioning for young children against the International Classification of Functioning, Disability and Health and Code and Core Sets relevant to neurodevelopmental conditions’ Autism 2022: volume 26, pages 928-939. https://doi.org/10.1177/13623613211036809

1010 Federici S and others. ‘World Health Organization disability assessment schedule 2.0: an international systematic review’ Disability and Rehabilitation 2017: volume 39, pages 2347-2380. https://doi.org/10.1080/09638288.2016.1223177

1011 Federici S and others. ‘Measurement characteristics of WHODAS 2.0 and WHODAS-Child: a systematic review of global psychometric studies in specific populations since 2010’ Frontiers in Psychiatry 2026: volume 17, page 1737452. https://doi.org/10.3389/fpsyt.2026.1737452

1012 Alehagen L and others. ‘Psychometric properties of the operationalized ICF Core Sets for autism and ADHD: item metrics, reliability, and validity’ Disability and Rehabilitation 2026: volume 48, pages 3871-3898. https://doi.org/10.1080/09638288.2025.2581019

1013 Hasslinger J and others. ‘Autism and ADHD strengths and needs in a nutshell – parental evaluation of a digital ICF-based assessment summary report’ Journal of Autism and Developmental Disorders 2026. https://doi.org/10.1007/s10803-026- 07376-3

1014 Freeth M and others. ‘Acceptability and usability of the International Classification of Functioning CoreSets for Autism platform: evaluation by U.K. autistic adults’ Neurodiversity 2026: volume 4. https://doi.org/10.1177/27546330261426684

1015 National Development Team for Inclusion. ‘“It’s Not Rocket Science”. Considering and meeting the sensory needs of autistic children and young people in CAMHS inpatient services’ 2020. https://www.ndti.org.uk/wp- content/uploads/2025/07/Its-not-rocket-science.pdf

1016 Barry TD and others. ‘The utility of home problem pervasiveness and severity in classifying children identified with attention-deficit/hyperactivity disorder’ Child Psychiatry and Human Development 2011: volume 42, pages 152-165. https://doi.org/10.1007/s10578-010-0205-2

1017 Luk SL and others. ‘Clinic observations in the assessment of pervasiveness of childhood hyperactivity’ Journal of Child Psychology and Psychiatry 1991: volume 32, pages 833-850. https://doi.org/10.1111/j.1469-7610.1991.tb01906.x

1018 Pilvar H and others. ‘Cost of emergency hospital admissions to acute general wards for mental health problems among children and young people in England, 2012–2022: a retrospective observational study’ BMJ Open 2026: volume 16, page e107143. https://doi.org/10.1136/bmjopen-2025-107143

1019 Ogundele MO and Morton MJS. ‘Subthreshold autism and ADHD: a brief narrative review for frontline clinicians’ Pediatric Reports 2025: volume 17, page 42. https://doi.org/10.3390/pediatric17020042

1020 Horstmann L and others. ‘Attention deficit hyperactivity disorder and other neurodevelopmental traits are associated with impact on functioning among children in the general population’ JCPP Advances 2025: volume 5, page e70004. https://doi.org/10.1002/jcv2.70004

1021 Baxter AJ and others. ‘Reducing excess mortality due to chronic disease in people with severe mental illness: meta-review of health interventions’ British Journal of Psychiatry 2016: volume 208, pages 322-329. https://doi.org/10.1192/bjp.bp.115.163170

1022 Taylor MJ and others. ‘Association between co-occurring psychiatric conditions and premature mortality in autistic people: population-based cohort study’ British Journal of Psychiatry 2026. https://doi.org/10.1192/bjp.2026.10696

1023 Welsh Government. ‘NEST framework: full report’ 2023. https://www.gov.wales/nest-framework-full-report-html

1024 Welsh Government. ‘The Mental Health and Wellbeing Strategy 2025–2035’ 2025. https://www.gov.wales/sites/default/files/publications/2026-04/the-mental- health-and-wellbeing-strategy-2025-to-2035.pdf

1025 Department for Work and Pensions. ‘Personal Independence Payment (PIP) assessment: second independent review’ 2016. https://www.gov.uk/government/publications/personal-independence-payment-pip- assessment-second-independent-review

1026 Jacobs R and others. ‘Funding approaches for mental health services: is there still a role for clustering?’ BJPsych Advances 2018: volume 24, pages 412-421. https://doi.org/10.1192/bja.2018.34

1027 Healthwatch York. ‘“I want to know, I want a diagnosis, I want help”. Pilot pathway for autism and ADHD: independent evaluation August 2023’ 2023. https://nds.healthwatch.co.uk/sites/default/files/reports_library/20231114_York%20In dependent-evaluation-pilot-pathway.pdf

1028 Male I and others. ‘Should clinical services for children with possible ADHD, autism or related conditions be delivered in an integrated neurodevelopmental pathway?’ Integrated Healthcare Journal 2020: volume 2, page e000037. https://doi.org/10.1136/ihj-2019-000037

1029 Autistica. ‘Earlier support and earlier outcomes: full report’ 2026. https://www.autistica.org.uk/downloads/files/Autistica-ICF-report-August-2026.pdf

1030 Maciver D and others. ‘Waiting times and influencing factors in children and adults undergoing assessment for autism, ADHD, and other neurodevelopmental differences’ Autism Research 2025: volume 18, pages 788-801. https://doi.org/10.1002/aur.70011

1031 Olsson E and others. ‘Stepped care versus stratified care for youth with anxiety disorders: a pilot randomized controlled trial’ Child Psychiatry and Human Development 2025. https://doi.org/10.1007/s10578-025-01908-1

1032 Kuipers E and others. ‘Cognitive model of caregiving in psychosis’ British Journal of Psychiatry 2010: volume 196, pages 259-265. https://doi.org/10.1192/bjp.bp.109.070466

1033 Radez J and others. ‘Why do children and adolescents (not) seek and access professional help for their mental health problems? A systematic review of quantitative and qualitative studies’ European Child & Adolescent Psychiatry 2021: volume 30, pages 183-211. https://doi.org/10.1007/s00787-019-01469-4

1034 Green J and others. ‘An integrated early care pathway for autism’ Lancet Child & Adolescent Health 2022: volume 6, pages 335-344. https://doi.org/10.1016/S2352- 4642(22)00037-2

1035 Goodman R and others. ‘Using the Strengths and Difficulties Questionnaire (SDQ) to screen for child psychiatric disorders in a community sample’ British Journal of Psychiatry 2000: volume 177, pages 534-539. https://doi.org/10.1192/bjp.177.6.534

1036 Reardon T and others. ‘Development of a brief assessment tool to identify children with probable anxiety disorders’ JCPP Advances 2025: volume 5, page e12265. https://doi.org/10.1002/jcv2.12265

1037 Barbaro J and others. ‘Diagnostic accuracy of the Social Attention and Communication Surveillance-Revised With Preschool Tool for early autism detection in very young children’ JAMA Network Open 2022: volume 5, page e2146415. https://doi.org/10.1001/jamanetworkopen.2021.46415

1038 UK National Screening Committee. ‘Evidence summary on screening for autism spectrum disorder in pre-school children under the age of 5 years’ 2022. https://view-health-screening-recommendations.service.gov.uk/review/evidence- summary/download-documents/cover_sheet/

1039 Rickwood D and others. ‘Sixteen years of innovation in youth mental healthcare: outcomes for young people attending Australia’s headspace centre services’ Plos One 2023: volume 18, page e0282040. https://doi.org/10.1371/journal.pone.0282040

1040 Singh SP and others. ‘Effect of managed transition on mental health outcomes for young people at the child–adult mental health service boundary: a randomised clinical trial’ Psychological Medicine 2023: volume 53, pages 2193-2204. https://doi.org/10.1017/S0033291721003901

1041 Priebe S and others. ‘The effectiveness of a patient-centred assessment with a solution-focused approach (Dialog+) for patients with psychosis: a pragmatic cluster-randomised controlled trial in community care’ Psychotherapy and Psychosomatics 2015: volume 84, pages 304-313. https://doi.org/10.1159/000430991

1042 Carruthers S and others. ‘Mediation of 6-year mid-childhood follow-up outcomes after pre-school social communication (PACT) therapy for autistic children: randomised controlled trial’ J Child Psychol Psychiatry 2024: volume 65, pages 233- 244. https://doi.org/10.1111/jcpp.13798

1043 Whitehouse AJO and others. ‘Effect of preemptive intervention on developmental outcomes among infants showing early signs of autism: a randomized clinical trial of outcomes to diagnosis’ JAMA Pediatrics 2021: volume 175, page e213298. https://doi.org/10.1001/jamapediatrics.2021.3298

1044 Gardner F and others. ‘Equity effects of parenting interventions for child conduct problems: a pan-European individual participant data meta-analysis’ Lancet

Psychiatry 2019: volume 6, pages 518-527. https://doi.org/10.1016/S2215- 0366(19)30162-2

1045 van Aar J and others. ‘Sustained, fade-out or sleeper effects? A systematic review and meta-analysis of parenting interventions for disruptive child behavior’ Clinical Psychology Review 2017: volume 51, pages 153-163. https://doi.org/10.1016/j.cpr.2016.11.006

1046 Sandbank M and others. ‘Autism intervention meta-analysis of early childhood studies (Project AIM): updated systematic review and secondary analysis’ BMJ 2023: volume 383, page e076733. https://doi.org/10.1136/bmj-2023-076733

1047 O’Hara L and others. ‘Video feedback for parental sensitivity and attachment security in children under five years’ Cochrane Database of Systematic Reviews 2019: volume 11, page CD012348. https://doi.org/10.1002/14651858.CD012348.pub2

1048 van Ginneken N and others. ‘Primary-level worker interventions for the care of people living with mental disorders and distress in low- and middle-income countries’ Cochrane Database of Systematic Reviews 2021: volume 8, page CD009149. https://doi.org/10.1002/14651858.CD009149.pub3

1049 World Health Organization. ‘mhGAP intervention guide for mental, neurological and substance use disorders in non-specialized health settings. Version 2.0’ 2016. https://www.who.int/publications/b/31377

1050 Sarvet B and others. ‘Improving access to mental health care for children: the Massachusetts Child Psychiatry Access Project’ Pediatrics 2010: volume 126, pages 1191-1200. https://doi.org/10.1542/peds.2009-1340

1051 Asarnow JR and others. ‘Integrated medical-behavioral care compared with usual primary care for child and adolescent behavioral health: a meta-analysis’ JAMA Pediatrics 2015: volume 169, pages 929-937. https://doi.org/10.1001/jamapediatrics.2015.1141

1052 Goodyer IM and others. ‘Cognitive behavioural therapy and short-term psychoanalytical psychotherapy versus a brief psychosocial intervention in adolescents with unipolar major depressive disorder (Impact): a multicentre, pragmatic, observer-blind, randomised controlled superiority trial’ Lancet Psychiatry 2017: volume 4, pages 109-119. https://doi.org/10.1016/s2215-0366(16)30378-9

1053 Lock J and others. ‘Randomized clinical trial comparing family-based treatment with adolescent-focused individual therapy for adolescents with anorexia nervosa’ Archives of General Psychiatry 2010: volume 67, pages 1025-1032. https://doi.org/10.1001/archgenpsychiatry.2010.128

1054 Mufson L and others. ‘A randomized effectiveness trial of interpersonal psychotherapy for depressed adolescents’ Archives of General Psychiatry 2004: volume 61, pages 577-584. https://doi.org/10.1001/archpsyc.61.6.577

1055 Mehlum L and others. ‘Dialectical behavior therapy for adolescents with repeated suicidal and self-harming behavior: a randomized trial’ Journal of the American Academy of Child and Adolescent Psychiatry 2014: volume 53, pages 1082-1091. https://doi.org/10.1016/j.jaac.2014.07.003

1056 McCauley E and others. ‘Efficacy of dialectical behavior therapy for adolescents at high risk for suicide: a randomized clinical trial’ JAMA Psychiatry 2018: volume 75, pages 777-785. https://doi.org/10.1001/jamapsychiatry.2018.1109

1057 Rossouw TI and Fonagy P. ‘Mentalization-based treatment for self-harm in adolescents: a randomized controlled trial’ Journal of the American Academy of Child and Adolescent Psychiatry 2012: volume 51, pages 1304-1313 e1303. https://doi.org/10.1016/j.jaac.2012.09.018

1058 Fuggle P and others. ‘The Adolescent Mentalization-based Integrative Treatment (AMBIT) approach to outcome evaluation and manualization: adopting a learning organization approach’ Clinical Child Psychology and Psychiatry 2015: volume 20, pages 419-435. https://doi.org/10.1177/1359104514521640

1059 Ung D and others. ‘A systematic review and meta-analysis of cognitive- behavioral therapy for anxiety in youth with high-functioning autism spectrum disorders’ Child Psychiatry and Human Development 2015: volume 46, pages 533- 547. https://doi.org/10.1007/s10578-014-0494-y

1060 Sharma S and others. ‘Cognitive behavioural therapy for anxiety in children and young people on the autism spectrum: a systematic review and meta-analysis’ BMC Psychology 2021: volume 9, page 151. https://doi.org/10.1186/s40359-021- 00658-8

1061 Cottrell DJ and others. ‘Effectiveness of systemic family therapy versus treatment as usual for young people after self-harm: a pragmatic, phase 3, multicentre, randomised controlled trial’ Lancet Psychiatry 2018: volume 5, pages 203-216. https://doi.org/10.1016/S2215-0366(18)30058-0

1062 Royal College of Paediatrics and Child Health. ‘Children in crisis: mental health pressures in emergency departments’ 2026. https://www.rcpch.ac.uk/resources/children-crisis-mental-health-pressures- emergency-departments-quick-read-report

1063 Ougrin D and others. ‘Comparison of effectiveness and cost-effectiveness of an intensive community supported discharge service versus treatment as usual for adolescents with psychiatric emergencies: a randomised controlled trial’ Lancet Psychiatry 2018: volume 5, pages 477-485. https://doi.org/10.1016/S2215- 0366(18)30129-9

1064 Herpertz-Dahlmann B and others. ‘Day-patient treatment after short inpatient care versus continued inpatient treatment in adolescents with anorexia nervosa (ANDI): a multicentre, randomised, open-label, non-inferiority trial’ Lancet 2014: volume 383, pages 1222-1229. https://doi.org/10.1016/s0140-6736(13)62411-3

1065 Edwards D and others. ‘Review: Crisis responses for children and young people – a systematic review of effectiveness, experiences and service organisation (CAMH-Crisis)’ Child and Adolescent Mental Health 2024: volume 29, pages 70-83. https://doi.org/10.1111/camh.12639

1066 Ougrin D and others. ‘Trial of Therapeutic Assessment in London: randomised controlled trial of Therapeutic Assessment versus standard psychosocial assessment in adolescents presenting with self-harm’ Archives of Disease in Childhood 2011: volume 96, pages 148-153. https://doi.org/10.1136/adc.2010.188755

1067 Ougrin D and others. ‘Therapeutic interventions for suicide attempts and self- harm in adolescents: systematic review and meta-analysis’ Journal of the American Academy of Child and Adolescent Psychiatry 2015: volume 54, pages 97-107.e102. https://doi.org/10.1016/j.jaac.2014.10.009

1068 Witt KG and others. ‘Interventions for self-harm in children and adolescents’ Cochrane Database of Systematic Reviews 2021: volume 3, page CD013667. https://doi.org/10.1002/14651858.CD013667.pub2

1069 Asarnow JR and others. ‘An emergency department intervention for linking pediatric suicidal patients to follow-up mental health treatment’ Psychiatric Services 2011: volume 62, pages 1303-1309. https://doi.org/10.1176/ps.62.11.pss6211_1303

1070 Inagaki M and others. ‘Active contact and follow-up interventions to prevent repeat suicide attempts during high-risk periods among patients admitted to

emergency departments for suicidal behavior: a systematic review and meta- analysis’ BMC Psychiatry 2019: volume 19, page 44. https://doi.org/10.1186/s12888- 019-2017-7

1071 Bickley H and others. ‘Suicide within two weeks of discharge from psychiatric inpatient care: a case-control study’ Psychiatric Services 2013: volume 64, pages 653-659. https://doi.org/10.1176/appi.ps.201200026

1072 Foster D and Bunn S. ‘Mental Health Act reform – impacts on autistic people and people with a learning disability (POSTnote 722)’ UK Parliament, 2024. https://post.parliament.uk/research-briefings/post-pn-0722/

1073 Clark A and others. ‘Severe mental illness in the UK. Public perceptions and healthcare priorities’ King's College London, 2025. https://www.kcl.ac.uk/policy- institute/assets/severe-mental-illness-in-the-uk.pdf

1074 Huckshorn KA. ‘Reducing seclusion restraint in mental health use settings: core strategies for prevention’ Journal of Psychosocial Nursing and Mental Health Services 2004: volume 42, pages 22-33. https://doi.org/10.3928/02793695- 20040901-05

1075 Stanley B and others. ‘Comparison of the Safety Planning Intervention with follow-up vs usual care of suicidal patients treated in the emergency department’ JAMA Psychiatry 2018: volume 75, pages 894-900. https://doi.org/10.1001/jamapsychiatry.2018.1776

1076 Grist R and others. ‘Technology delivered interventions for depression and anxiety in children and adolescents: a systematic review and meta-analysis’ Clinical Child and Family Psychology Review 2019: volume 22, pages 147-171. https://doi.org/10.1007/s10567-018-0271-8

1077 Baumel A and others. ‘Objective user engagement with mental health apps: systematic search and panel-based usage analysis’ Journal of Medical Internet Research 2019: volume 21, page e14567. https://doi.org/10.2196/14567

1078 National Institute for Health and Care Excellence. ‘Guided self-help digital cognitive behavioural therapy for children and young people with mild to moderate symptoms of anxiety or low mood: early value assessment (HTE3)’ 2023. https://www.nice.org.uk/guidance/hte3

1079 McBain RK and others. ‘Evaluation of alignment between large language models and expert clinicians in suicide risk assessment’ Psychiatric Services 2025: volume 76, pages 944-950. https://doi.org/10.1176/appi.ps.20250086

1080 Heard K. ‘From analogue to digital: tackling inequality and digital exclusion in the future NHS’ Good Things Foundation, 2025. https://www.goodthingsfoundation.org/discover/digital-inclusion-insights/digital- inclusion-insights-2025/tackling-digital-exclusion-in-the-future-nhs

1081 Bowyer G and others. ‘Closing the digital divide for good: an end to the digital exclusion of children and young people in the UK’ Carnegie UK Trust and Unicef UK, 2021. https://www.unicef.org.uk/wp-content/uploads/2021/06/Closing-the-Digital- Divide-for-Good_FINAL.pdf

1082 Titov N and others. ‘User characteristics and outcomes from a national digital mental health service: an observational study of registrants of the Australian MindSpot Clinic’ Lancet Digital Health 2020: volume 2, pages e582-e593. https://doi.org/10.1016/S2589-7500(20)30224-7

1083 MQ: Transforming Mental Health. ‘UK Mental Health Research Funding 2014– 2017’ 2019. https://www.mqmentalhealth.org/wp- content/uploads/UKMentalHealthResearchFunding2014-2017digital.pdf

Additional evidence considered by the Review In addition to the published sources referenced here, the Review drew on a range of supporting evidence papers prepared by topic groups and commissioned contributors to inform its evidence base. These papers are intended for separate publication subject to future peer review.

Annex A. A practical framework for assessing need

This Annex sets out an illustrative framework for a needs-led approach to support. It is offered for further development and independent evaluation. It is not a validated assessment instrument, a diagnostic tool, an eligibility framework or a national service-allocation algorithm.

Its purpose is to provide a practical framework to support planning, commissioning and service design. It describes:

  • a common way of recording need across six broad dimensions; and
  • a small number of indicative levels of response, linking need to appropriate support, workforce and interventions.

The Review does not recommend national adoption of this framework at this stage. Any implementation should begin through co-designed demonstrator sites and be evaluated for validity, safety, equity, acceptability, workforce implications, cost-effectiveness and unintended consequences.

A.1 Purpose

Chapters 18 and 19 propose separating the decision to support a person from the decision to establish a diagnosis. This Annex addresses the two practical questions that follow and which commissioners and service planners repeatedly raised during the Review:

How can we describe a person’s functioning, participation, risk and support needs in a way that different services can understand and use?

How can that description guide decisions about the type and intensity of support required?

This Annex is intended primarily as a practical framework for discussion and development by commissioners, service planners and clinicians. It should be read alongside Chapter 18 (the needs-led plus diagnosis model) and Chapter 19 (the pathway).

A.2 The principle

Describe need. Match support to need. Keep that description available as needs change.

Everything in this Annex depends on keeping two questions distinct.

What support does this person need now, and at what intensity? This is answered through a structured assessment of need across the six dimensions described below. The resulting shared description informs the timing, intensity and type of support alongside clinical formulation, risk assessment and professional judgement, whether or not a diagnosis has yet been established.

Does this person require diagnosis for confirmation of condition, condition-specific treatment, medication or specialist care? Diagnostic assessment and support should proceed in parallel where appropriate. Neither should be made unnecessarily dependent on completion of the other.

In time, this approach could be supported by a single digital platform into which validated digital tools could be incorporated, with AI-assisted navigation helping individuals, families and professionals identify appropriate support according to need.

Supporting this would be a secure, portable digital record that follows the individual across services, geographical areas and life stages, maintaining a shared description of need, personal goals and relevant clinical information. Any such system should be interoperable with existing records and protected against vendor lock-in so that tools remain evidence-based and interchangeable.

A.3 Assessing need: the six dimensions

Need is described across six dimensions. Table A.1 summarises each dimension, what it captures and how it can be evidenced. Together they inform three related, but distinct, decisions:

  • What support should begin now?
  • How urgent is the response?
  • What level of expertise is needed for further assessment or treatment?

These questions should not be collapsed into a single score or level. Immediate risk may require urgent intervention, while diagnostic uncertainty may require specialist assessment without constituting an immediate crisis. The six dimensions are intended to support clinical formulation, not replace it. Their purpose is to provide a shared, portable description of need that can be understood across services.

Table A.1 The six dimensions of need

Dimension of needWhat it capturesHow would a clinician assess?
Current functioning and participationThe impact of current difficulties on everyday life. For children this includes attendance, learning, friendships, family life, self-care and, where developmentally relevant, sleep, feeding and toileting. For young adults it includes education, employment, relationships and independent living. This is the primary dimension because participation is the outcome the pathway aims to protect. It also records the distress associated with these difficulties, including where participation is maintained only at considerable personal cost, and distress that has not yet shown itself in functioning is still need. Assessment should distinguish capacity (what someone can do with appropriate support) from performance (what they currently manage in everyday life).ICF participation domains; WHODAS 2.0 or a validated child equivalent; information from the person, family and school or college.
Risk of deteriorationThe likely course of the person’s difficulties, including declining functioning, risks to self or others, safeguarding concerns and the likelihood of deterioration without timely support.Clinical risk assessment (a professional judgement, never delegated to an automated tool); history of change; safeguarding information.
Complexity and co-occurrenceThe interaction between mental health, neurodevelopmental, communication, physical health and social difficulties. Complexity arises from how these difficulties interact rather than simply how many diagnoses are present. Physical health should always be considered explicitly because it is frequently overlooked, particularly where communication is limited.Clinical formulation across domains; structured assessment of co-occurring conditions; developmental history.
Environmental contextThe family, educational, occupational, housing, community and socioeconomic factors that either support or hinder participation. Recording both barriers and facilitators helps distinguish difficulties that are best addressed by changing the environment from those requiring individual clinical intervention.ICF environmental domains; contextual and social history; information from family, education and other relevant settings.
Purpose and likely benefit of supportThe intended purpose of intervention and the outcomes that are realistic at this stage. These may include recovery, improved functioning, prevention of deterioration, safety, better communication, improved quality of life or sustaining family care. A lower likelihood of measurable improvement should never reduce entitlement to necessary support.Clinical and developmental judgement; response to previous interventions; shared decisionmaking with the person and family.
Capacity to manage engagementWhat services need to do to make support accessible, including communication adjustments, executive-function support, outreach, continuity and culturally appropriate care. Previous non-attendance or disengagement should prompt adaptation by services rather than reduced support.Assessment of communication and engagement needs; previous experiences of services; information from people who know the individual well.

A.4 The assessment process

The assessment should be developed with the person, and with their family where appropriate, rather than about them. What the person wants, values and believes is achievable should sit alongside professional observations. Where the person’s account and professional judgement differ, both should be recorded.

This is essential if the approach is to work. A description that the person does not recognise or trust is unlikely to be shared or used across services.

  • A professional formulation supported by structured tools, not an

administrative form. The starting point should be a concise, age-appropriate description of functioning, participation and environmental context, using validated measures where appropriate (for example, the ICF, WHODAS 2.0, or a validated child equivalent). No validated child equivalent is currently in routine use in England, and Chapter 18 explains why one should be developed and evaluated rather than adopted uncritically. The specific tools should be determined through evaluation rather than prescribed in this Annex. Autism and ADHD ICF Core Sets may guide more detailed enquiry where those conditions are diagnosed or are strong clinical possibilities, but they are not diagnostic instruments.

  • Experienced clinicians at first contact. Placing the least experienced staff

at the point of entry is a false economy. It is here that professional judgement matters most.

  • Multi-informant. Information should come from the person, their family, the

GP and, where appropriate, education or other relevant settings, so that the assessment reflects functioning across everyday life rather than a single perspective.

  • Continuous rather than one-off. The description should be reviewed at key

points using structured measures of functioning together with a person- centred review, allowing people to move between levels of support without re- entering the system from the beginning.

  • Digital as decision support, never decision-maker. Structured information

can be gathered before appointments to support richer clinical conversations, but decisions about suicide risk, self-harm, safeguarding and clinical management must always remain with appropriately qualified professionals. Non-digital routes should remain available throughout.

  • A single shared formulation and support plan. The outcome should be one

shared description setting out what support begins immediately, what requires diagnostic or specialist assessment, who is responsible for coordination, and when review will take place. It should not become another assessment that records need without leading to action. The description guides everyday support; diagnosis identifies condition-specific interventions. Together they form the portable record of need.

  • Relationship to diagnosis. Functional assessment complements rather than

replaces diagnosis. It guides the level of everyday support, while the certainty of diagnosis required increases with the specificity, intensity and risk of the intervention (Table A.2).

Table A.2 How diagnostic certainty relates to the decision (the graduated rule)

DecisionDiagnostic certainty requiredExamples
Everyday support that is broadly reversibleNone at the outsetEducational and workplace adjustments; parenting and family support; psychoeducation; sensory and environmental adaptations; guided self-help; active monitoring.
Higher-specificity or higher-risk interventionDiagnosis established promptly, in parallel with supportMedication (including stimulants); condition-specific psychological therapies; specialist community services; inpatient care where required.
Legal rights, statutory decisions and formal planningEvidence proportionate to the legal or statutory decision. Formal diagnosis is required only where legislation or statutory guidance explicitly requires it.Reasonable adjustments should normally be based on evidenced disability, disadvantage and functional need. Diagnosis may contribute to EHCPs, Care Act assessments and other statutory decisions, but should not be treated as universally necessary.

A shared description of need is valuable only if it can travel with the person, and it will only travel if the person trusts how it is used. The following principles are therefore not optional safeguards but essential features of the approach.

  • Consent should underpin information sharing. People should be informed

about how information is used and shared, and involved in decisions wherever possible, with parental responsibility, competence, capacity, safeguarding duties and the relevant legal framework taken into account. Information sharing should be lawful, necessary, proportionate, purpose- limited and transparent. Choosing not to share optional information should never be interpreted as disengagement or lead to reduced support.

o The system should be federated rather than centralised. The Review recommends exploring whether a lightweight, federated digital architecture could support portability, service coordination and commissioning without creating a new central repository of sensitive personal information. Such a system would require national standards for interoperability, cybersecurity, access, correction, audit and information governance.

o Information should come from routine care. The principal source should be the structured information generated through the pathway itself, including validated questionnaire scores, the shared description of need, pathway decisions, engagement with treatment, patient- reported outcomes and symptom trajectories. The aim is that information gathered during care supports future care rather than being lost within separate systems.

  • Population information should support improvement. Governed within

established data-safe-haven arrangements, aggregated information could provide local systems with a consistent picture of demand, outcomes and inequalities while supporting service evaluation and research.

  • The description should travel; the clinical narrative should not. What is

shared should be the description of need and the adjustments it implies, not the detailed diagnostic report or clinical history from which it was derived. People should not have to disclose intimate personal histories simply to obtain practical support.

  • Sharing should be proportionate and purpose-specific. Employers should

receive information relevant to workplace adjustments; schools, information relevant to learning and participation; clinicians, the clinical information required for care. Sharing should never be all-or-nothing.

  • Access should be role-based and auditable. The same Caldicott principles

should govern any shared or single-record arrangement, and people should be able to see who has accessed their information.

  • People should be able to read, correct and contribute to their own

record. Where professional and personal accounts differ, both should be recorded.

  • Non-digital routes must remain available. Digital tools can enrich assessment but should never become the only route into support.

A.6 Who assesses, and what each part requires

A shared description of need has important workforce implications. It involves professionals across health, education, local authorities, the voluntary sector, early years services and employment support. The approach is only safe if responsibilities are clearly defined. This is what distinguishes a shared description of need from an unqualified assessment.

Recording observations – Any trained practitioner in contact with the person can contribute observations about functioning, participation, environmental context and engagement. This requires training in the shared language of need rather than in clinical assessment.

Formulation across domains – A practitioner with appropriate professional training and supervision should bring these observations together into a coherent formulation, considering complexity, co-occurrence and the person’s own account.

Risk assessment and decisions about the level of response – These remain the responsibility of a qualified professional with clinical accountability and should never be delegated to automated systems or the least experienced staff. Decisions about suicide risk, self-harm, safeguarding and urgency belong here.

  • Physical health review – Recording physical health is not enough. It should be reviewed by a clinician able to recognise problems and act on what is

found. The value of including physical health in the assessment is lost if it is never translated into care.

A.7 The four levels of need

Three design principles underpin the four levels and are as important as the levels themselves.

First, the levels describe the intensity of response, not categories of people. Individuals move between them as their needs change, in either direction, and many will experience more than one level over time. What changes is the intensity of support; the shared understanding of the person remains continuous. That is what makes this a pathway rather than a tiered system.

  • Level 1 – Emerging or mild need: early support
  • Level 2 – Moderate need (the ‘missing middle’): enhanced support
  • Level 3 – Severe or complex need: specialist coordinated care
  • Level 4 – Very severe or high-risk need: highly specialised support

Second, the level reflects the pattern across all six dimensions of need and is never calculated from a score. Table A.3 illustrates how the dimensions are typically interpreted at each level. It is intended to support professional judgement, not replace it. A high level of need in one domain – particularly risk – may justify a higher level of response regardless of the others.

Duration of unmet need should also be considered. Need that has persisted through a prolonged wait should be recognised as need that has increased rather than need that has remained stable. Otherwise, the system risks rewarding deterioration as the route to support.

Third, the descriptors themselves (Table A.3) should be regarded as hypotheses for evaluation, not validated thresholds. They should not be used to determine eligibility, remove people from diagnostic pathways or drive contractual payment until they have been prospectively tested and independently evaluated.

The level then determines the response. Table A.4 illustrates how settings, workforce and interventions differ across the four levels. The examples are presented for ages 0–25, reflecting the strongest evidence available to the Review. The six dimensions of need, and the graduated relationship between diagnosis and

support, apply across the life course. Extending the framework to adults would require the same work to be undertaken for adult settings, workforce and statutory responsibilities.

The four levels are intended to encompass the full range of need, including the most severe and enduring. One group merits particular attention: autistic people whose difficulties emerge early and who require substantial, often lifelong support, commonly including severe or profound intellectual disability, very limited functional speech, marked dependency in daily living and complex physical health needs.

Three principles follow.

First, movement between levels relates to the intensity and composition of support, not to whether support continues. For this group, support is lifelong. Stability at a high level of need is not a reason to reduce support; it is often evidence that the support is working.

Second, Table A.3 should not flatten profound and enduring disability into the same range as difficulties that can be addressed through reasonable adjustments alone. Where communication is very limited, dependency is high, behaviour presents significant risk, or physical health is complex, several dimensions of need will appropriately remain at Levels 3 or 4.

Third, capacity to manage engagement requires particular care. A person who cannot describe pain or who communicates distress through behaviour does not have a reduced capacity to engage. They have communication needs that the system must respond to. The same applies to families and carers who provide information the person cannot communicate directly.

Whether the proposed descriptors achieve these aims is one of the questions that demonstrator sites should address and is listed among the evaluation questions in Section A.8.

Information should be gathered from the young person, their family and education using a structured template. A trained practitioner would prepare the initial description, while an appropriately qualified clinician would review the formulation, risk assessment and pathway decision. Where needs span health, education and social care, the final response should be agreed through multidisciplinary discussion rather than by a single practitioner.

Table A.3 Reading the six dimensions across the four levels of need

DimensionLevel 1 – emerging / mildLevel 2 – moderate (‘missing middle’)Level 3 – severe / complexLevel 4 – very severe / high-risk
Current functioning and participationEmerging or mild; everyday participation beginning to change but largely maintained. Distress may be marked; where it is the response is lifted regardless of participation reading.Definite, meaningful impairment to attendance, learning, relationships or family life.Marked impairment across several areas of life.Severe impairment; participation in ordinary life substantially prevented.
Risk of deteriorationCurrently low evidence of deterioration or immediate risk; review required if difficulties persist or circumstances change.Moderate; difficulties liable to entrench if unsupported.Significant; persistent or escalating despite earlier help.High; acute risk to safety, or crisis.
Complexity and co-occurrenceLimited complexity or predominantly one area of difficulty, with no substantial interaction between multiple needs currently identified.Some co-occurrence; interacting difficulties emerging.Substantial co-occurrence of mental-health, neurodevelopmental and social needs.Highest complexity; multiple interacting conditions and circumstances.
Environmental contextBroadly supportive, or barriers readily modifiable.Contributory barriers needing active attention.Significant adversity or unmet environmental need compounding difficulty.Severe adversity, safeguarding or placement instability often present.
Purpose and likely benefit of supportHigh from low-intensity, developmental support.Good from structured, matched intervention.Requires specialist intervention to benefit.Requires intensive, highly specialised, sustained input.
Capacity to manage engagementGenerally intact; light-touch access sufficient.May need adjustments and outreach to engage.Often needs assertive coordination to engage and sustain.Frequently needs intensive outreach and keyworking; disengagement itself a risk.

Table A.4 The response at each level (settings across the age bands, workforce and interventions)

LevelSettings (age bands 0–5 · 5–11 · 11– 16 · 16–25, illustrated at Level 1)WorkforceInterventions
Level 1 – emerging / mildHealth visiting, family hubs, parent– infant and early-years services (0–5); schools, primary care, community organisations (5–11); secondary schools, youth services, community and digital support (11–16); colleges, universities, early-support hubs, primary care, neighbourhood services (16–25).Health visitors, family-support and earlyyears practitioners, school nurses, education mental-health and children’s wellbeing practitioners, youth workers, social-prescribing link workers, peer supporters, school and college pastoral and SENCO staff, and the voluntary sector – with specialist consultation and supervision.Parenting and parent–infant programmes; psychoeducation; support for sleep, feeding and regulation; guided self-help; brief interventions; coaching; peer support; social prescribing; youth work; environmental and educational adjustments; early language support.
Level 2 – moderate (‘missing middle’)Enhanced school-linked mental health and Mental Health Support Teams; early-support hubs; neighbourhood multidisciplinary teams; community therapies and community mental health; for older youth, primary-care mental health with participation and employment support.Mental-health nurses, counsellors, clinical and educational psychologists, psychological wellbeing practitioners, family therapists, speech-and-language and occupational therapists, specialist teachers – alongside the Level 1 workforce.A range of therapies matched to presentation (CBT; family and systemic; interpersonal; DBT-A; MBT-A; AMBIT for complex cases); executive-function and social-communication support; sensory adaptations; OT and SLT; educational adjustments; mentoring and attendance support; employment support (IPS) for older youth; adapted, parity-of-access care for neurodivergent young people.
Level 3 – severe / complexCAMHS, community paediatrics, integrated neurodevelopmental services, specialist therapy and educational support, neighbourhood multidisciplinary teams, local-authority and social care – coordinated, not sequential.Fully multidisciplinary: child and adolescent psychiatrists, paediatricians, clinical psychologists, mental-health nurses, speech-and-language and occupational therapists, pharmacists, family therapists, social workers, education professionals.Specialist psychological therapies; ADHD and autism assessment and treatment; medication management; intensive family intervention; specialist SLT and OT; early intervention in psychosis; key-working and coordinated multi-agency planning – organised around one plan, not separate episodes.
LevelSettings (age bands 0–5 · 5–11 · 11– 16 · 16–25, illustrated at Level 1)WorkforceInterventions
Level 4 – very severe / high-riskCrisis assessment and response; intensive home treatment; community eating-disorder and early-intervention-inpsychosis services; day services; specialist neurodevelopmental services; and, where necessary for safety, inpatient care.Consultant psychiatrists and paediatricians, clinical psychologists, specialist nurses and therapists, social workers, crisis practitioners and intensive community teams.Crisis assessment and intervention; specialist therapies; medication optimisation; intensive family work; highly specialised behavioural support; intensive care coordination; and, where clinically necessary, inpatient treatment – with transition planning from the outset.

A.8 Applying this approach – a worked example

The tables above are easier to understand through an example. This section follows a single description of need through the framework: the six dimensions, the level of need they indicate, and the response that follows. It uses the same constructed example as Box 18.1 in Chapter 18 so that the two can be read together. The chapter shows the description travelling between the young person, her family and different services; this Annex shows how the same description guides decisions about support. The example is illustrative and is not based on an identifiable individual or patient record.

A seventeen-year-old in the second year of a college course has begun missing classes. She has been waiting fourteen months for an autism and ADHD assessment. A shared description of need is developed with her, her mother and her personal tutor.

Current functioning and participation

She is still attending college but missing increasing numbers of sessions. She is managing the coursework but struggling with deadlines and unstructured group work. She is sleeping poorly and describes a persistent anxiety she has not previously disclosed. Participation has begun to deteriorate rather than collapse, but the distress is greater than participation alone suggests, which is why both are recorded.

Risk of deterioration

Attendance has declined over the course of a term. She has withdrawn from friendships she previously valued and has twice disengaged from services after being asked to repeat her history. There is no immediate crisis, but the trajectory is downward and the history of disengagement increases the importance of timely support.

Complexity and co-occurrence

Autism and ADHD remain under consideration, but neither diagnosis has yet been established. Anxiety and disturbed sleep are interacting with these emerging difficulties. No physical health concerns have been recorded, but neither have they been explored. The absence of information should not be assumed to indicate the absence of need.

Environmental context

Her college is supportive and able to make adjustments; her part-time employer is not. Her social network has reduced. The fourteen-month wait is itself now part of her environment rather than simply a delay before it.

Purpose and likely benefit of support

The likely benefit of support is high and immediate. Structured support with deadlines, access to a quiet space, and treatment for anxiety can all begin now and are likely to improve participation.

Capacity to manage engagement

Her capacity to engage is reduced. Previous experiences of repeatedly recounting her history without receiving effective help have undermined trust. The implication is not that less should be offered, but that support should be coordinated, consistent and sustained.

The level

Taken together, the six dimensions indicate Level 2. No single dimension determines the level. Functioning is meaningfully impaired but not markedly so; risk is moderate rather than high; complexity is emerging rather than substantial; and environmental barriers are important but potentially modifiable. The prolonged wait is interpreted as evidence of increasing unmet need rather than successful coping. Importantly, the level does not depend on whether autism or ADHD is eventually diagnosed.

The response

Table A.4 indicates a Level 2 response for a young person aged 16–25. This includes structured support with deadlines and access to a quiet space through the college’s existing arrangements; treatment for anxiety in primary care without waiting for the neurodevelopmental assessment; workplace adjustments agreed with her consent; and a named professional responsible for maintaining continuity. The neurodevelopmental assessment continues in parallel. If a diagnosis is made, it adds to this shared understanding rather than replacing it and enables access to any interventions or legal entitlements that depend upon diagnosis.

What would change the level?

Escalating self-harm, further deterioration in attendance or the loss of support from college would increase the ratings for risk and environmental context and indicate Level 3. Equally, if anxiety improves and the adjustments prove sufficient, the level would reduce to Level 1, with review at the next major transition. Both movements are expected. Neither requires the young person to leave the pathway and begin again.

What this example illustrates

Every element of the description was already known to somebody – the young person, her family or her college. None required a diagnosis before it could be recorded. The response is matched to the overall pattern of need rather than calculated from a numerical score.

Demonstrator sites should evaluate whether the framework shortens time to first support, improves timely diagnosis where needed, increases access to treatment, strengthens continuity across transitions, improves participation in education and employment, enhances quality of life and family wellbeing, and reduces inequalities. They should also monitor unintended consequences, including inappropriate diversion from specialist assessment, missed risk, delayed diagnosis, digital exclusion, repeated assessment and movement into lower-cost services without improved outcomes.