Why is autism on the increase?

Why autism has increased is something many parents quietly search for — and quietly blame themselves for. I looked for every possible explanation. I blamed pregnancy sickness medication. I blamed stress. But the research points somewhere different, and understanding it properly is more useful than the guilt loop.

Here is what the evidence actually says.

Why autism diagnosis rates have risen since the 1990s

Autism diagnoses have increased dramatically over the past thirty years. In the UK, estimates now put autism prevalence at around 1 in 57 school-age children, based on English school census data. In 1990, reported rates were below 1 in 1,000. That is a substantial difference, and it asks for an explanation.

The clearest explanation — supported across the research — is that the way autism is defined, recognised, and diagnosed has changed enormously. The number of autistic people being born has not.

What explains the rise in autism diagnoses?

Estimated contributions — tap each to learn more

35%28%17%12%8%

Broader diagnostic criteria~35%Increased awareness~28%Girls now being identified~17%Adult diagnosis~12%Genuine increase / unknown~8%

Proportions are illustrative estimates based on research consensus, not precise measurements. The "unknown" slice reflects genuine scientific uncertainty.

Four changes explain most of the rise:

1. The diagnostic criteria widened. Before 1994, an autism diagnosis required significant language or intellectual delay. Asperger's syndrome — covering autistic people without those delays — wasn't in the diagnostic manual until that year. When it was added, a large population who had previously been missed suddenly had a route to identification. The DSM-5 in 2013 consolidated all presentations into a single autism spectrum, widening recognised presentation further still.

2. Awareness increased dramatically. Clinicians, teachers, and parents recognise autism now in ways they simply didn't thirty years ago. A child labelled "difficult," "anxious," or "in their own world" in 1985 is more likely to be referred for assessment today. This isn't overdiagnosis; it's recognition of something that was already there.

3. Girls and women started being diagnosed. Until recently, autism was understood almost entirely through research on boys and men. As awareness of the female autism phenotype has grown, diagnosis rates in girls have risen sharply. This represents a population that was always autistic, not a new one.

4. Adult diagnosis became possible. Many adults are now diagnosed for the first time, often after a child's diagnosis prompts recognition of their own profile. These are people who were always autistic; they're just finally counted.

Is autism actually becoming more common — or better diagnosed?

This is where it gets more nuanced, and where the research is honest about what it doesn't know.

Most researchers agree that changes in diagnosis and awareness account for a large portion of the increase. At the same time, scientists continue to study genetic, developmental, and environmental factors that may contribute to individual cases. The picture is multifactorial and remains genuinely uncertain in parts.

Genetics is the clearest factor. Between 50% and 90% of autism heritability is estimated to be genetic, based on twin and family studies. Many parents discover their own autism after a child is diagnosed, which is consistent with the genetic picture. Specific genes involved in brain development and synaptic function have been identified as associated with autism, though no single gene causes it.

Key numbers:

• 50–90% of autism heritability is estimated to be genetic
• Around 1 in 5 autistic children has a sibling who is also autistic
• Advanced parental age accounts for an estimated 5–10% of autism cases — real but modest

Common theories about why autism is on the increase — what the research says

These are the questions many parents search. Here is what the research shows, stated plainly.

Does stress during pregnancy cause autism?

This is one of the more complicated areas, because there is more research here than many people expect — but it needs careful reading.

A 2026 meta-analysis of 22 studies involving over three million participants found that significant maternal psychological distress during pregnancy was associated with a 72% increased likelihood of an autism diagnosis in offspring. Other studies have found associations between stressful life events during the later stages of pregnancy and autistic traits in children.

There are important caveats. Association is not causation: women who are more stressed during pregnancy may also share genetic factors with their children that are relevant to autism. Studies using genetically informed designs — where the mother is genetically unrelated to the child — show smaller or inconsistent effects, suggesting genetics plays a significant role in the observed associations. Researchers describe this as a "gene-environment interaction": maternal genetics that influence stress response may work alongside prenatal stress, rather than stress alone being the driver.

What the research does not support is the idea that everyday stress during a normal pregnancy causes autism.

Does having children later in life increase autism risk?

Yes, there is a well-replicated association here. A meta-analysis of 27 studies found that every additional 10 years of maternal age was associated with an 18% higher risk, and every 10 years of paternal age with a 21% higher risk. A large multinational study of nearly 5.8 million children confirmed that both older maternal and paternal age independently increase risk.

The likely mechanism for paternal age is de novo mutations: genetic alterations that accumulate in sperm over time and can affect brain development. Maternal age mechanisms are less well understood.

The important context: while the relative risk increases with age, the effect size is modest. Parental age is estimated to account for around 5–10% of autism cases. The majority of children born to older parents are not autistic. And given that people are having children later partly because society now provides more choices and stability for that decision, this is not a reason to panic.

Can retained primitive reflexes cause autism?

Primitive reflexes are automatic movements present at birth that typically integrate — become voluntary — over the first year of life. Research does find that retained primitive reflexes are more common in autistic children than in neurotypical controls, and there is a theoretical link to the kind of brain connectivity differences associated with autism.

What the evidence does not support is the idea that retained reflexes cause autism, or that they explain why autism is on the increase. The relationship is more likely bidirectional: autism-related differences in brain maturation may result in some reflexes integrating more slowly, rather than reflex retention causing autism.

Interventions targeting primitive reflex integration have some preliminary research suggesting modest improvements in motor coordination and some behavioural outcomes. The evidence is described as preliminary; no major clinical body recommends reflex integration as a treatment for autism.

This is an area where the research is genuinely still developing. If you've been looking at it for your child, it's not a dead end, but treat commercial programmes making strong claims with appropriate scepticism.

Does Brain Balance work for autism?

Brain Balance is a US-based commercial programme that attributes autism and ADHD to "functional disconnection syndrome" — a theory that weak connections between brain hemispheres drive neurodevelopmental differences. The programme combines physical exercises, sensory activities, and dietary changes.

The theory is not medically recognised. A dozen experts interviewed by NPR described the hemispheric imbalance concept as "too simplistic" and built on "a popular, discredited myth." On the National Autism Center's Scientific Merit Rating Scale, none of the studies evaluating Brain Balance scored above 2 out of 5. The Association for Science in Autism Treatment has described the treatment components as meeting the definition of pseudoscientific or unproven.

Brain Balance does not explain why autism is on the increase. It also costs thousands of pounds per programme. The concern from researchers is not that exercises and sensory activity are harmful — they're not — but that the cost, time, and opportunity could be spent on support with a stronger evidence base.

Can pregnancy medication cause autism?

Several anti-nausea medications used during pregnancy, including ondansetron, have been studied for associations with autism. Some studies have found associations; others have not. The research is ongoing and hasn't produced consistent findings.

This is an area where the honest answer is: we don't know. Studies showing associations are finding small effects in large datasets where many other factors are also at play. If you took medication during pregnancy because you needed it to function, that was the right decision with the information available. The evidence does not support the conclusion that it caused your child's autism.

Why the "autism epidemic" framing gets the story wrong

The framing of autism as an epidemic implies that something external is causing a sudden surge in a harmful condition, and that it might be preventable. It also, by extension, implies that autism is a tragedy requiring prevention.

That framing does real harm. It has fuelled decades of parent guilt and led many families down expensive, unproven paths looking for a cause they could have avoided. It has also obscured the more accurate story: that hundreds of thousands of people have lived their whole lives without an explanation for their experience, because the systems around them weren't looking for the right things.

The rise in autism diagnoses is, in very large part, the closing of a longstanding diagnostic gap. More autistic people being identified means more autistic people accessing understanding, community, and support. That is progress.

An element of autism appears to be genetic — so instead of searching for something to blame, it can help to remember you're probably also wired similarly, and that there's a whole community out there that understands exactly how your family works.

---

This article is part of the Neuroequipped Autism guide. For how autism is diagnosed, see Autism assessment: NHS and private options. For autism in girls and the diagnostic gap, see Autism in girls: why it looks different. For what support is available after diagnosis, see the EHCP hub.

Neuroequipped provides research-grounded information for parents and educators. It is not medical advice.