Why are autism diagnoses rising?
Written by AI from saved sources. A script checks every quote and every figure against the saved sources, and a second AI compares each claim with its source. How we work and corrections
Better recognition, broader autism definitions and changes in records explain important parts of the large rise in recorded autism. These sources cannot tell us how much underlying autism risk has changed.
Checked: CDC; global systematic review; Danish reporting-practices study. October 2026.
Why people believe it
A California study describes concern about rising autism prevalence. England’s interim review describes competing views of rising counts of mental-health and neurodevelopmental conditions. These include worsening population mental health and wellbeing, or changes in recognition and diagnosis. Neurodevelopmental conditions begin during development and cause difficulties in personal, social, school or work life. Its final review stresses that people seeking assessment have real difficulties and need support.
“Surely the increase is real?”
Prevalence means the share of a population with a condition.
The increase in the share of children identified as autistic is real. In five monitoring areas, CDC researchers compared the share of eight-year-olds identified as autistic. It was 22.8% higher in 2022 than in 2020. Those areas covered the same places and used the same sources of records in both years. So changing which areas participate cannot explain the whole increase.
A separate US survey covered children outside institutions. It found that the share of 3–17-year-olds with a parent-reported autism diagnosis was 3.1% in 2019–2020. This rose to 4.5% in 2023–2024. It asked about ever receiving a professional diagnosis. It did not examine children directly or establish why the increase happened.
“Are we counting it differently?”
A global review identifies several contributors to the rise in recorded autism. They include broader autism definitions, greater awareness and identification in previously missed groups. For its 2022 count, CDC included children aged four or eight who lived in a monitoring area and whose records met any of three conditions. One is an autism diagnosis stated in a full assessment of the child’s development. Another is qualifying for autism special education. The third is having one of the specified codes for an autism diagnosis.
A Danish study examined changed diagnostic criteria and the addition of outpatient records for children born in 1980–1991. It estimated these changes explained 60% of the observed rise in reported autism prevalence in that group. The study reported a 95% confidence interval of 33–87%. With repeated samples, this method produces intervals that cover the true value 95% of the time. That estimate applies to that Danish population. It cannot tell us that recording changes explain 60% of today’s US increase.
“Does the rest prove something new is causing autism?”
A California study found that the factors it examined did not fully explain the rise. Other possible effects of how cases were identified and recorded had yet to be quantified. The dates when autism began were unknown. An unexplained part of the rise therefore cannot identify a particular cause. The global review found weak or conflicting evidence that differences in autism prevalence over time or place reflect differences in exposure to environmental risk factors. It also acknowledged that it had not formally checked study quality.
Thinking tool: changing measurement or changing risk?
Ask what the number counts before asking what caused it. CDC’s “one in 31” describes eight-year-olds identified in 16 US monitoring areas in 2022. The estimate does not represent the whole country. CDC’s measure of first identification uses whichever was recorded earlier: an autism diagnosis or qualifying for autism special education. It does not track when autism began. For other claims, check whether the age group, definition and way of finding cases stayed comparable. Do this before treating a changing count as changing risk.
Official positions, with dates
CDC, 17 April 2025: its monitoring report says diagnostic practices and access to early evaluation might explain community differences.
WHO, 17 September 2025: available evidence suggests there are probably multiple factors that make a child more likely to have autism. These include genetic and environmental factors. The measles, mumps and rubella (MMR) vaccine does not cause autism.
England’s commissioned independent review, 21 October 2026: it reports little evidence of substantial underlying prevalence change. It warns that there has been no recent England-wide study of autism prevalence. It discusses autism and attention deficit hyperactivity disorder together. It says rising recorded rates alone do not establish that widespread overdiagnosis is currently occurring. It says the risks of misdiagnosis and overdiagnosis may be becoming more important for policy and services.
What people also ask
Why do the US have the highest rates of autism?
That ranking is not established by these sources. CDC’s one-in-31 figure covers selected areas and one age. WHO’s global estimate covers all ages.
Why is the autism rate so high now?
Awareness, broader autism definitions and identification among previously missed groups contribute to rising measured rates. These sources do not provide a complete breakdown.
Why is autism so high in the USA?
CDC researchers suggest several factors might help explain differences in identified rates between US communities. These include differences in evaluation, insurance coverage and eligibility for services. These are possible contributors, not a complete explanation.
Does the USA have more autism than other countries?
These figures cannot settle that. A study found diagnosed autism in 2.94% of 10–14-year-olds in its sampled English primary-care records in 2018. CDC’s one-in-31 figure concerns eight-year-olds in selected US areas in 2022. Different ages, years and methods prevent treating these figures as a ranking of countries by autism risk.
Check it yourself (5 minutes)
- Open CDC’s 2025 monitoring report.
- Find “not nationally representative” in its Limitations section.
- Does its one-in-31 estimate represent every US child?
The report uses ASD for autism spectrum disorder.
What you should have found
“First, the populations within site-defined ADDM Network surveillance areas are not nationally representative and do not generate nationally representative ASD prevalence estimates.”
Before you forward this: does the box at the top say exactly what the sources below say?
Sources
CDC, 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 (2025); CDC/NCHS, Trends in Parent-Reported Developmental Disabilities: 2019–2024 (2026); Global systematic review, Global prevalence of autism: A systematic review update (2022); Danish reporting-practices study, Explaining the Increase in the Prevalence of Autism Spectrum Disorders: The Proportion Attributable to Changes in Reporting Practices (2015, public abstract only); California diagnostic-change study, Diagnostic change and the increased prevalence of autism (2009); California age-at-diagnosis study, The Rise in Autism and the Role of Age at Diagnosis (2009); English primary-care study, Autism in England: assessing underdiagnosis in a population-based cohort study of prospectively collected primary care data (2023); WHO, Autism (2025); England’s DHSC-commissioned independent review, Independent review into mental health conditions, ADHD and autism: interim report (2026) and executive summary (2026).