Has Autism Become Over-Diagnosed? What the Evidence Shows

Originally published March 2026 · Updated May 2026 with the March 2026 UK government independent review interim report on autism and ADHD, the April 2025 CDC update showing 1 in 31 US children, and the April 2025 Nature analysis of the RFK Jr "epidemic" claim

The number of people diagnosed with autism has risen steeply and consistently across the past three decades. In the United States, autism prevalence has risen from one in 150 eight-year-olds in 2000 to one in 31 in 2022 — a 382% increase. In the UK, there was a 787% exponential increase in recorded incidence of autism diagnoses between 1998 and 2018. By 2025, autism-related identified need accounted for around 3.1% of school-age children within the SEND system, with growth especially rapid among girls and pupils without learning disability.

These numbers have generated two very different responses.

In April 2025, US Health and Human Services Secretary Robert F. Kennedy Jr held a press conference pointing to the CDC data. He called it an "epidemic" caused by "an environmental toxin" and said he would soon be announcing a study to find the responsible agent. Scientists said he was ignoring answers from decades of research.

In the UK, a March 2026 government independent review interim report took a considerably more measured position, finding that stable epidemiological prevalence can coexist with rising symptom reporting and rising diagnostic activity — and that the two need not be in conflict.

The question "has autism become over-diagnosed?" is less straightforward than it sounds. Here is what the evidence actually shows.

TL;DR

• Autism prevalence in the US has risen from one in 150 in 2000 to one in 31 in 2022. In the UK, a 787% rise in recorded diagnoses occurred between 1998 and 2018.
• The scientific consensus is that the rise primarily reflects improved recognition, broader diagnostic criteria, and identification of previously missed groups — not a true increase in prevalence at the same rate as the diagnostic increase.
• The March 2026 UK government independent review interim report found that stable epidemiological prevalence can coexist with rising diagnostic activity — the underlying condition may not be more common while diagnoses rise.

Girls and women, adults, and people without intellectual disability are the groups most significantly driving recent diagnostic increases — all historically under-identified.

A 2023 study estimated that between 150,000 and 500,000 individuals aged 20 to 49 in England may be autistic but undiagnosed.

The RFK Jr "epidemic from environmental toxins" claim is not supported by the evidence. Genetic and neurodevelopmental factors account for the largest proportion of autism causation in research.

The over-diagnosis question matters because it shapes how society allocates resources and how autistic people are treated.

Why Diagnoses Are Rising: The Evidence-Based Explanations

The most important and most consistently supported explanation for rising autism diagnoses is improved recognition — not a true biological increase in the condition's prevalence.

A UK population-based cohort study tracking autism diagnoses between 1998 and 2018 concluded that increases are more likely due to increased reporting and application of diagnosis than to genuine growth in prevalence. Rising diagnosis among adults, females, and higher functioning individuals suggests augmented recognition underpins these changes.

Three specific mechanisms account for the majority of the increase.

Broadened Diagnostic Criteria

Earlier diagnostic frameworks focused on narrow, severe presentations — classic autism as originally described by Leo Kanner in 1943, characterised by significant communication impairment and marked behavioural restriction. Later, Asperger's syndrome and pervasive developmental disorder were added as separate categories.

The DSM-5, introduced in 2013, merged these separate categories into a single autism spectrum disorder diagnosis. This consolidation captured the reality that autistic traits exist on a continuous spectrum of presentation — but it also expanded the number of people meeting diagnostic criteria by including presentations that would previously have been classified separately or not at all.

This is not diagnostic inflation in a pejorative sense. It is a more accurate representation of how autism actually presents. But it does arithmetically increase diagnosis numbers.

Historical Under-Identification

Perhaps the most important and most overlooked explanation is that many of the people now being diagnosed were never missed by autism — they were missed by diagnostic systems not designed to identify them.

Girls and women represent the most significant group. The rise in autism diagnoses has been greater for females than males, moderated by age band, with the greatest rises in diagnostic incidence among adults. For decades, the diagnostic criteria, assessment tools, and clinical understanding of autism were based almost entirely on research conducted in male subjects. The social camouflaging strategies — "masking" — that autistic girls and women frequently develop to navigate social environments were not recognised as characteristic of autism. They were interpreted as evidence that autism was not present.

This has produced a significant diagnostic backlog. A 2023 study estimated that between 150,000 and 500,000 individuals aged 20 to 49 in England may be autistic but undiagnosed. These are not people who developed autism recently. They are people who were autistic throughout their lives and are now being recognised.

The same pattern applies to adults more broadly — the diagnostic rise among adults is proportionally larger than among children in UK data — and to people without intellectual disability, who were historically less likely to be referred for assessment.

Greater Awareness and Reduced Stigma

Increased public awareness, online communities where autistic people share experiences, better GP and teacher training, and reduced stigma around neurodevelopmental differences have all contributed to more people recognising autistic traits in themselves and their children and seeking assessment.

Parent-reported prevalence rose by around 25% between 2022 and 2024, suggesting that public recognition and identification of autism has accelerated. This acceleration in self-referral and professional referral has driven waiting list growth rather than any sudden change in the underlying population.

The RFK Jr Controversy: What Scientists Actually Say

The April 2025 press conference by RFK Jr brought the autism prevalence debate to global attention in a way that has not been seen since the thoroughly discredited Wakefield MMR vaccine hypothesis of the 1990s.

RFK Jr pointed to CDC data showing autism prevalence rising from one in 150 to one in 31 since 2000, called it an "epidemic" caused by "an environmental toxin," and vowed to announce a study to find the responsible agent. Scientists said he was ignoring answers from decades of research.

The scientific response was consistent and clear. The claim that autism is an "epidemic" caused by an identifiable environmental toxin misrepresents the evidence in several important ways.

First, as established above, the rise in diagnosed prevalence reflects improved recognition of a condition that was always more prevalent than previously measured — not evidence of a new causative agent.

Second, autism has well-established genetic and neurodevelopmental foundations. Twin studies consistently show heritability estimates of 60 to 90%. Hundreds of genetic variants associated with autism have been identified. The neurodevelopmental origins of autism occur in utero — during brain development — not through post-natal environmental exposure of the kind RFK Jr was implying.

Third, the specific "environmental toxin" hypothesis has been investigated and has not produced the evidence base that would support his framing. Environmental factors — including advanced parental age, prenatal exposure to certain medications, preterm birth, and air pollution — have been associated with modestly increased autism risk in epidemiological studies. But these are contributing factors in a complex neurodevelopmental picture, not the explanation for a dramatic prevalence increase.

Despite political hostility and the spread of misinformation that disrupted some lines of research, autism scientists rose to the occasion in 2025 — continuing the careful, long-term research programme that the political intervention threatened to derail.

The March 2026 UK Government Independent Review

The most authoritative current UK assessment comes from the March 2026 interim report of the independent review into mental health conditions, ADHD, and autism — commissioned by the UK government and published on March 31, 2026.

The review found that stable epidemiological prevalence can coexist with rising symptom reporting and rising diagnostic activity. In the available survey data, parent-reported prevalence rose by around 25% between 2022 and 2024, suggesting that public recognition and identification of autism has increased. By 2025, autism-related identified need accounted for around 3.1% of school-age children within the SEND system, with growth especially rapid among girls and pupils without learning disability, indicating changing patterns of recognition.

The review's framing is important: a rise in diagnostic activity does not necessarily indicate over-diagnosis. It may indicate correction of previous under-diagnosis. Whether current diagnostic rates reflect appropriate identification, under-identification, or over-identification requires careful disaggregation — and the evidence points to different answers in different contexts.

Where over-diagnosis may be occurring: some research suggests inconsistency between diagnostic services — the same individual presenting to different services might or might not receive a diagnosis, reflecting variability in assessment practice rather than genuine clinical differences. This is a quality problem rather than evidence that autism itself is being over-diagnosed as a category.

Where under-identification persists: the 150,000 to 500,000 estimated undiagnosed autistic adults in England represents a significant ongoing under-identification problem, particularly affecting women, ethnic minorities, and people in more deprived areas with less access to diagnostic services.

The NHS Crisis: Waiting Times and System Pressure

Whatever the cause of rising diagnoses, the practical consequence is a system under enormous pressure.

As of September 2024, over 204,000 people in England awaited an autism assessment, with 89% waiting over 13 weeks — a 25% rise in a year and 96% in two years.

These waiting times have profound consequences. Adults and children waiting years for diagnosis cannot access the support, accommodations, and services that a formal diagnosis enables. Schools, universities, and employers cannot make appropriate adjustments. Mental health services see the downstream effects — undiagnosed autistic individuals presenting with anxiety, depression, burnout, and crisis — at considerable cost to individuals and systems alike.

40 to 50% of individuals with autism also have an anxiety disorder, and this prevalence is rising as more autism diagnoses occur, increasing demand for autism-related mental health services. Autistic individuals, especially those with co-occurring mental health conditions, are nine times more likely to die by suicide than the general population. The Anxiety Reset covers many of the daily-life strategies that overlap with autistic burnout and overwhelm.

The diagnostic waiting list problem is not a consequence of over-diagnosis. It is a consequence of the system's failure to scale to meet the long-standing unmet need that improved awareness and recognition has made visible.

Is Autism Being Over-Diagnosed? The Honest Answer

The honest answer is: probably not in the systematic way the question implies — but with important nuance.

The bulk of the evidence suggests that rising diagnoses primarily reflect improved recognition of a condition that was previously under-identified, particularly in girls, women, adults, and people without intellectual disability. The diagnostic increase has exposed historic under-identification rather than created artificial cases.

There are genuine quality and consistency concerns in diagnostic practice — the same person might receive different outcomes from different services, reflecting variability in training, resources, and assessment quality rather than over-diagnosis as such.

And there remains a meaningful debate about whether the broadening of the autism spectrum — particularly at its less severe end — captures genuinely autistic individuals or creates overlap with personality variation that does not reach clinical threshold. This is a legitimate research question. It does not support the "epidemic from environmental toxins" framing.

What is clearly true is that many autistic people remain undiagnosed, that diagnostic waiting lists are causing harm, and that the surge in public and political attention on autism in 2025 has been accompanied by both important research and significant misinformation.

Frequently Asked Questions

Why have autism diagnoses increased so dramatically?

The scientific consensus is that the rise primarily reflects improved recognition rather than a true biological increase. Key factors include broader diagnostic criteria following the introduction of the autism spectrum model, historical under-identification of girls and women, adults, and people without intellectual disability, and greater public awareness and reduced stigma prompting more people to seek assessment. A UK population study concluded that the rise is more likely due to increased reporting and application of diagnosis than to genuine growth in prevalence.

Is autism an epidemic caused by environmental toxins?

No — this claim, made by RFK Jr in April 2025, is not supported by the scientific evidence. Autism has well-established genetic and neurodevelopmental foundations, with heritability estimates of 60 to 90%. The rise in diagnoses reflects improved recognition of a condition that was always more prevalent than historical data captured. Environmental factors have been identified as modest contributors to autism risk in research, but do not explain the diagnostic increase. Scientists said RFK Jr was ignoring decades of research in making his claims.

Are girls and women under-diagnosed with autism?

Yes — historically, significantly so. Diagnostic criteria and assessment tools were based largely on research in male subjects. Autistic girls and women often develop masking strategies that obscure their autism from diagnostic observation. The rise in female autism diagnoses in recent years is proportionally larger than the rise in males, driven by belated recognition of these previously missed presentations. A 2023 study estimated 150,000 to 500,000 undiagnosed autistic adults in England.

How long is the NHS autism waiting list?

As of September 2024, over 204,000 people in England were waiting for an autism assessment, with 89% waiting more than 13 weeks. Waiting times had risen 25% in one year and 96% in two years. In practice, many people wait years rather than weeks for assessment. The waiting list reflects the combination of decades of unmet need becoming visible and a diagnostic system that has not scaled to meet it.

What is the current autism prevalence in the UK?

Precise UK prevalence estimates vary by study methodology, but the most commonly cited figure is approximately 1 in 100 people — though research suggests actual prevalence may be higher given the significant number of undiagnosed autistic adults. By 2025, autism-related identified need accounted for 3.1% of school-age children in the UK SEND system, with rapid growth particularly among girls.

Does an autism diagnosis help adults?

For many adults, a formal diagnosis provides validation of lifelong experiences, access to support and accommodations, and a framework for understanding their own cognition and social processing. It can reduce the anxiety and self-blame associated with years of struggling in neurotypical environments without understanding why. It can also enable access to workplace adjustments, support services, and appropriate mental health care. Whether to pursue a diagnosis is a personal decision — the process can be lengthy and emotionally demanding — but for many late-diagnosed adults, it is described as transformative.

The Bottom Line

Autism diagnoses have risen dramatically. The primary explanation — backed by the research consensus and the March 2026 UK government independent review — is improved recognition of a condition that was systematically under-identified for decades, particularly in girls, women, adults, and people without intellectual disability.

The "epidemic" framing advanced by RFK Jr in April 2025 misrepresents the evidence. The rise in diagnoses does not indicate a new causative agent. It indicates, largely, the correction of historical diagnostic failures.

What the rising diagnoses do indicate — unmistakably — is that autism is more prevalent than previously understood, that the systems designed to support autistic people are overwhelmed, and that a significant portion of autistic people in the UK and elsewhere remain unidentified and unsupported.

For understanding neurodevelopmental conditions and their relationship to mental health, anxiety, and daily functioning, the Anxiety Reset, ADHD Reset, and Stress Reset from the Reset Series™ address the overlapping conditions and daily management strategies most relevant to neurodivergent individuals. Pair any of these with the Reset Companion for ongoing, personalised guidance as you build the day-to-day systems that support neurodivergent wellbeing.

Related reading: Micro-Anxiety: The Tiny Stressors That Are Quietly Draining Your Mental Health · Why Therapy Works: The Neuroscience of Talking and How Digital Support Is Changing Everything · AI, Anxiety and the Future of Calm

Popular Articles