Race, ethnicity and autism
How autism diagnosis and care differ by race and ethnicity: who gets counted and how the US counts changed, where the diagnostic gap sits, what autistic adults of color face in services, and where the research itself leaves people out.
How autism diagnosis and care can be different depending on a person's race and ethnicity. This entry looks at who gets counted, where the gaps are, and what is still not known.
The diagnosis gap, and how the counts changed Emerging
The groups least likely to have a documented autism diagnosis in US records have changed places. In the 2002 surveillance year, among 2,568 eight-year-olds who met the criteria for autism, Black children were less likely than White children to have a documented diagnosis (odds ratio 0.79), as were Hispanic children (0.76) and children of other race or ethnicity (0.65), after adjustment for sex, IQ, birthweight and maternal education.1 The shortfall for Black children persisted regardless of IQ, and for children of other ethnicities it was concentrated among those with an IQ below 70.1
Two decades on, the same network finds the opposite. In the 2020 surveillance year, autism was more common among Black, Hispanic and Asian or Pacific Islander 8-year-olds than among White or multiracial 8-year-olds.2 In 2022 the network counted 32.2 autistic 8-year-olds per 1,000 overall: White children 27.7, Asian or Pacific Islander 38.2, American Indian or Alaska Native 37.5, Black 36.6, Hispanic 33.0 and multiracial 31.9.3 The report describes this as "continuing a pattern first observed in 2020."3
A 2026 review cautions that convergence in prevalence "does not constitute equity," because differences in the age at which children are diagnosed, in co-occurring intellectual disability and in later outcomes remain.8 The higher counts do go with a different route to identification: among autistic 8-year-olds with cognitive-ability data, 52.8% of Black children and 32.7% of White children had a co-occurring intellectual disability, so a larger share of the Black children counted were also identified with higher support needs.3
Once, autism was written down less often for some children than for White children. In 2002, a US team looked at records for 2,568 children. Black, Hispanic and other children of color were less likely than White children to have autism listed.1
Today the picture is the other way round. In 2020, autism was counted more often in Black, Hispanic and Asian or Pacific Islander children than in White children.2 In 2022, 32.2 children out of every 1,000 were counted as autistic.3
Counting more children does not mean everything is fair. One review says the numbers coming closer together "does not constitute equity." Children are still diagnosed at different ages. More Black children were also found to have an intellectual disability, which means slower learning and thinking skills.3,8
US surveillance data from one network. The reports come from the same program and count children whose records contain a diagnosis, so they show who is identified, not how many autistic people there are in each group.
The network counts children whose records contain a diagnosis, a school classification or a billing code. A group counted more often may be reached better by services, assessed more often for another reason, or recognized more often when support needs are clear. The figures cannot show how many autistic people there are in each group, and the sites that make up the network differ widely, so part of the spread between groups is a spread between places.
Counting a condition across a whole population from records and assessments, rather than testing a sample of people.
This comes from US records. They count children who already had a diagnosis. They do not show how many autistic people are in each group.
The count includes children whose records mention autism, a school class or a billing code. A group may be counted more often because services reach them better. The numbers cannot tell us how many autistic people are really in each group.
Counting a condition in a whole population from records, instead of testing a sample of people.
Missed, then given another label Emerging
When the first specialist visit does not end in an autism diagnosis, children are given another label. In a study of 406 Medicaid-eligible children in one US city, Black children were 2.6 times less likely than White children to receive an autism diagnosis at their first specialty care visit.4 Among the children who did not receive one, ADHD was the most common diagnosis; relative to White children, Black children had 5.1 times the odds of being given adjustment disorder rather than ADHD, and 2.4 times the odds of conduct disorder rather than ADHD.4
Diagnosis also comes late. Parents of 584 African American autistic children enrolled in a US autism genetics registry recalled an average age at diagnosis of 64.9 months, on average 42.3 months after their first concerns about their child's development.5 Registries like it reach only part of the population: a 2010 report on one US region found that, even after traditional barriers to taking part were addressed and every family contacted agreed, 67% of reachable African American families were excluded from two national autism genetics registries by family structure alone.6 The 42-month gap describes the families who could enroll.
A 2026 systematic review of 20 US studies found that delayed diagnosis was the most common kind of diagnostic error, and named provider bias in taking histories and interviewing, and assessment measures built on mostly White samples, among the contributors for Black and Hispanic patients.7 A 2026 review of the wider literature makes a similar case about the measures themselves.8
Where on the pathway the gap opens is not settled. A clinical service evaluation at a London adult autism service found that the 350 clients referred and accepted for assessment were more likely to be male, aged 18 to 34, and of White or "Other" ethnicity than the local population; among the 269 who were assessed, no difference in the diagnostic rate was detected between groups.9 A null result in that small sample cannot rule out a smaller effect at assessment, and a comparison between referred clients and local population figures assumes that autism is equally common in each group, so the referral disparity rests on that assumption. In a participatory study of 12 caregivers of 13 autistic children, Spanish-speaking Latine families described a long lag between a first concern and diagnosis, with language barriers between families and clinicians.10
Sometimes a child is given a different label first. In one US city, a team studied 406 children on Medicaid, a US health program for people with low incomes. Black children were 2.6 times less likely than White children to get an autism diagnosis then. This was at their first specialist visit.4 Instead, they were much more likely to be given another label. The labels were adjustment disorder or conduct disorder, which describe behavior problems.4
Diagnosis can also come late. Parents of 584 African American autistic children were asked about diagnosis.5 Their child was diagnosed at about 64.9 months old. That was about 42 months after they first worried.5 Many families like theirs are left out of the studies that produce these numbers.6
A 2026 review of 20 US studies found that a late diagnosis was the most common mistake. It said some doctors' bias, and tests built on mostly White children, are part of the reason.7,8
Where the gap starts is not settled. At a London service for adults, the people sent for testing were often male and younger. They were also more often White or "Other" than the local population. Among the 269 people tested, the doctors did not find a difference between groups.9 A small study cannot rule out a small difference. In another study, Spanish-speaking Latine families said they waited a long time and had trouble with language.10
Single studies from one US city, one US registry and one London service, with a 2026 systematic review of US diagnostic inequities in mental disorders (17 of its 20 studies were on autism) behind the general pattern.
Each finding comes from one study, or one review of many US studies.
Racism and stigma together Emerging
Stigma and racism can work together. In 55 interviews with caregivers of Black autistic children at a public autism clinic in the US South, caregivers described punitive responses from people around them to their child's behavior, safety concerns on public transport, and difficulty getting services through the public school system. The authors call these intersectional stigma, where disability stigma and racism reinforce each other, and link them to reluctance to seek help and to delayed diagnosis.11 Both reviews of disparities in autism diagnosis and care make the same point, noting that caregivers from minoritized groups report lower-quality care.7,8
Stigma about disability and racism can happen at the same time. Caregivers of Black autistic children were interviewed 55 times.11 They talked about people being harsh about their child's behavior. They worried about safety on public transport. They had trouble getting help at school.11 Reviews say families like theirs often report worse care.7,8
One qualitative interview study of caregivers and two reviews. It describes what families said, not how often it happens.
Stigma that comes from more than one part of who a person is at once, such as racism and disability stigma together.
This comes from interviews with caregivers. It describes what they said, not how often it happens.
Stigma that comes from more than one part of who a person is, such as racism and disability stigma together.
Autistic adults of color Emerging
Disparities in care continue into adulthood. Records from a Northern California health system, covering 2008 to 2012, show that Black, Hispanic and Asian autistic adults were less likely than White autistic adults to have a recorded psychiatric diagnosis.12 In 2012 Medicare and Medicaid records for 172,071 publicly insured autistic adults, about half of White adults (49.9%) were eligible for both programs, against 37.5% of Black, 34.7% of Asian or Pacific Islander and 35.9% of Hispanic adults; after adjustment Black adults were significantly less likely to be dual-eligible, and median spending was higher for White than for Black beneficiaries among full-dual-eligible and Medicaid-only beneficiaries.13 Medicaid records from 2008 to 2012 show that Black, Hispanic and Asian autistic adults had higher odds of diabetes, heart disease and hypertension.14 The same 2012 public insurance records show that, among 122,250 autistic youth and young adults aged 14 to 29, Black, Asian or Pacific Islander and Hispanic people were less likely than White people to have a recorded mental health condition or to have had a community mental health visit.15
A recorded diagnosis may also understate need. In the REALS study of 179 autistic adults, Black participants reported similarly high current depression and anxiety whether or not they had a lifetime diagnosis of depression, while White participants differed by whether they had one.16 Among 44,074 autistic Medicaid enrollees aged 8 to 29 with a new depression claim, Black enrollees had about two-thirds the odds of continuous depression treatment (odds ratio 0.63) and Hispanic enrollees about three-fifths (0.58), against no or limited treatment.17
Little of this work reaches later life. A scoping review of racial and ethnic disparities among autistic adults aged 60 and over found 8 studies out of 4,179, and reported disparities in insurance eligibility, health conditions and service use.18
The differences continue into adult life. One California health system keeps records on autistic adults.12 Black, Hispanic and Asian adults were less likely than White adults to have a mental health diagnosis written down.12 In 2012, White autistic adults were more likely to get both Medicare and Medicaid than other groups.13 Medicaid records show other differences too. Black, Hispanic and Asian autistic adults were more likely to have diabetes, heart disease and high blood pressure.14 Among 122,250 autistic young people, those who were not White were less likely to get mental health care.15
Some people need care but have no diagnosis on record. In one study, Black autistic adults felt depressed and anxious. This was true whether or not a doctor had written it down.16 Another study followed 44,074 autistic people on Medicaid.17 Black and Hispanic people were less likely to keep getting treatment for depression.17
Older autistic adults have barely been studied this way. A review looked for studies about race in autistic adults aged 60 and over. It found only 8 out of more than 4,000.18
Several US data sources agree that disparities continue into adult services. Much of the data is from 2012 or earlier, all of it counts recorded diagnoses and service use, and several studies draw on the same public insurance records, so it shows who was treated, not who is autistic.
These are counts from records in the United States. Much of the data is from 2012 or earlier. It shows who was treated, not who is autistic.
Police contact and safety Emerging
Autistic young people come into contact with police. In a nationally representative US sample of youth with autism, about 20% had been stopped and questioned by police by age 21 and nearly 5% had been arrested; the study had no non-autistic comparison group, so it shows how many autistic youth had contact, not whether that is more than other youth.19 In a UK cohort of 11,859 young people, autistic adolescents were more likely than non-autistic peers to have received a police warning before other factors were taken into account; after adjustment that difference did not hold, and in the adjusted comparison they were less likely to have been stopped, with no difference in arrests between the groups. Adolescents who were stopped and questioned were more likely to report short-term self-harm, and the link between a police warning and later depression was stronger for autistic than for non-autistic adolescents.20
How much of the risk comes from race is harder to pin down. Autistic writers describe the danger directly (see below), and the policing of autistic people is covered in Police, courts and prison.
The figure is widely repeated. It comes from a disability-rights report rather than a peer-reviewed study, and no peer-reviewed study found as of October 2026 measures it. It is not used as a finding here.
Autistic young people do come into contact with police. In one US study, about 20 in 100 autistic young people had been stopped by police by age 21. About 5 in 100 had been arrested. The study had no group of non-autistic young people to compare with.19 In a UK study, autistic teenagers were more likely to get a police warning at first. The study followed 11,859 young people. Once other things were taken into account, they were less likely to be stopped. Autistic teenagers who were stopped were more likely to hurt themselves soon after.20
This number is repeated a lot. It comes from a disability report, not a research study. No peer-reviewed study found as of October 2026 measures it.
Two cohort studies, in different countries, that measured police contact directly. Neither isolates the effect of race on the contact.
Two studies from two countries, each following a large group of young people.
Who the genetics studies include Established
The genetics of autism has been studied mostly in people of European ancestry, and that limits what the results can say about everyone else.21,22 A 2026 review describes the global distribution of autism genomics research as "profoundly imbalanced," with most non-European ancestry populations severely underrepresented, which constrains gene discovery and reduces how far the findings can be generalized.21
Studies in other populations are now appearing. A 2026 consortium sequenced more than 15,000 Latin American individuals, 4,717 of them autistic, and identified 35 autism-associated genes, with substantial overlap with European cohorts. For the damaging coding variants it studied, its authors concluded that the biology of autism is consistent across populations, with no detectable influence of ancestry.22 The 2026 review reads the same evidence differently, pointing to emerging evidence of ancestry-specific risk genes, which would mean the genetic architecture of autism is not uniform worldwide.21
Most studies of autism genes have looked at people of European ancestry. That limits what the results can say about everyone else.21,22 One review says the research is "profoundly imbalanced."21
New studies are starting to include other groups. In 2026, a large team studied more than 15,000 people in Latin America, 4,717 of them autistic. They found genes linked to autism that also show up in European studies.22 The review points to signs that some risk genes may differ between groups.21
A documented fact about the research literature, not a finding about autistic people: the genomes studied are mostly of European ancestry, stated in a 2026 review and in the primary studies themselves.
This is a known fact about the research itself. Reviews and the studies say it.
What would settle it Emerging
The studies so far describe who is identified and treated, from records. They cannot show how many autistic people there are in each group, or how much of the pattern comes from access to services, from how services reach families, or from how autistic traits are read.
A study that followed the same children over time, with autism assessed directly rather than taken from records, would show whether the counts drawing closer together are closing a gap or moving it. The adult picture rests almost entirely on records and public insurance data from one country; what it leaves out is autistic adults of color describing what happened in diagnosis, at work and with services, in their own words rather than through claims. For genetics, the fix is a fairer share of samples from under-studied ancestries.21,22
Today's studies count who gets a diagnosis and who gets care. They cannot show how many autistic people are in each group. They cannot show how much comes from access to services. They cannot show how much comes from how autistic traits are read.
A study could follow the same children over time. It would check for autism directly. That would show whether the numbers coming together is closing a gap or moving it. It would also help to hear from autistic adults of color in their own words.21,22
This section states what the current studies cannot answer, from the limits of the evidence rather than from a finding.
This section says what today's studies cannot answer.
In the community Community
Autistic writers describe carrying two masks at once. Megan Anna Neff gives code-switching as "a second layer of adapting one's language and mannerisms that BIPOC Autistic people navigate alongside masking."c1 Devon Price quotes Catina Burkett, a Black autistic woman, on how the same traits read differently through race and gender: "When I am inflexible, I am sometimes called unfriendly, insubordinate, lazy, aggressive or uncontrollable."c2 Price also reports Timotheus Gordon Jr., sent to speech therapy and taught to speak "standard" English, who was then mocked at his mostly Black school for speaking differently.c2 The clinicians Donna Henderson, Sarah Wayland and Jamell White call the combination double camouflage: hiding autistic traits while also code-switching to fit cultural expectations.c3
Eric Garcia writes that autism is imagined as White, and that the "perpetual child" image that can protect White autistic people is not extended to autistic people of color.c4 Price reports Aisha, a Black autistic woman, whose mother and sister held to the careful, respectable manner that had helped them survive, and who, in Aisha's words, "keep bullying me to do exactly what helped them survive."c5
Some autistic writers say they wear two masks at once. Megan Anna Neff calls this code-switching, a second layer of changing your words and manner to fit in.c1 Devon Price quotes Catina Burkett, a Black autistic woman. When she is just being herself, people call her unfriendly, lazy or aggressive.c2 Price also writes about Timotheus Gordon Jr. He was taught to speak "standard" English. Then he was teased at his mostly Black school.c2
Eric Garcia writes that people picture autism as White. He says autistic people of color are not given the same kindness.c4 Price tells of Aisha, a Black autistic woman. Her family want her to keep the careful manner that helped them survive.c5
How autistic writers describe being autistic and a person of color. It describes experience and argument; the research is in the earlier sections.
How autistic writers describe being autistic and a person of color. This is experience and argument, not research findings.
Sources
Research
- 1Mandell DS, Wiggins LD, Carpenter LA, et al. Racial/Ethnic Disparities in the Identification of Children With Autism Spectrum Disorders. American Journal of Public Health, 2009. https://doi.org/10.2105/AJPH.2007.131243
- 2Maenner MJ, Warren Z, Williams AR, et al. Prevalence and Characteristics of Autism Spectrum Disorder Among Children Aged 8 Years — Autism and Developmental Disabilities Monitoring Network, 11 Sites, United States, 2020. MMWR Surveillance Summaries, 2023. https://doi.org/10.15585/mmwr.ss7202a1
- 3Shaw KA, Williams S, Patrick ME, et al. 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. MMWR Surveillance Summaries, 2025. https://doi.org/10.15585/mmwr.ss7402a1
- 4Mandell DS, Ittenbach RF, Levy SE, Pinto-Martin JA. Disparities in Diagnoses Received Prior to a Diagnosis of Autism Spectrum Disorder. Journal of Autism and Developmental Disorders, 2007. https://doi.org/10.1007/s10803-006-0314-8
- 5Constantino JN, Abbacchi AM, Saulnier C, et al. Timing of the Diagnosis of Autism in African American Children. Pediatrics, 2020. https://doi.org/10.1542/peds.2019-3629
- 6Hilton CL, Fitzgerald RT, Jackson KM, et al. Brief Report: Under-Representation of African Americans in Autism Genetic Research: A Rationale for Inclusion of Subjects Representing Diverse Family Structures. Journal of Autism and Developmental Disorders, 2010. https://doi.org/10.1007/s10803-009-0905-2
- 7Srivarathan A, Bradford A, Shearkhani S, et al. Bridging diagnostic safety and mental health: a systematic review highlighting inequities in autism spectrum disorder diagnosis. BMJ Quality & Safety, 2026. https://doi.org/10.1136/bmjqs-2025-018723
- 8Bala A, Deban C, Hoskinson H, et al. Understanding disparities in autism diagnosis and care: A socioecological perspective. PLOS Mental Health, 2026. https://doi.org/10.1371/journal.pmen.0000703
- 9Woolhouse R, Nicholson SL. Exploring Biases in Autism Diagnostic Pathways. Autism Research, 2026. https://doi.org/10.1002/aur.70252
- 10Martin AM, Perez G, Battle EF, et al. Autism service barriers for Latine children and their families: a participatory approach to adapting autism diagnostic care. Frontiers in Psychiatry, 2026. https://doi.org/10.3389/fpsyt.2026.1834721
- 11Singh JS, Nicholas J. "You can't whoop autism": Intersectional stigma in the Black community. Social Science & Medicine, 2026. https://doi.org/10.1016/j.socscimed.2026.119218
- 12Ames JL, Morgan EH, Giwa Onaiwu M, et al. Racial/Ethnic Differences in Psychiatric and Medical Diagnoses Among Autistic Adults. Autism in Adulthood, 2022. https://doi.org/10.1089/aut.2021.0083
- 13Benevides TW, Carretta HJ, Rust G, Shea L. Racial and ethnic disparities in benefits eligibility and spending among adults on the autism spectrum: A cohort study using the Medicare Medicaid Linked Enrollees Analytic Data Source. PLOS ONE, 2021. https://doi.org/10.1371/journal.pone.0251353
- 14Schott W, Tao S, Shea L. Co-occurring conditions and racial-ethnic disparities: Medicaid enrolled adults on the autism spectrum. Autism Research, 2022. https://doi.org/10.1002/aur.2644
- 15Benevides TW, Jaremski JE, Williams E, et al. Racial and Ethnic Disparities in Community Mental Health Use Among Autistic Adolescents and Young Adults. Journal of Adolescent Health, 2024. https://doi.org/10.1016/j.jadohealth.2024.01.036
- 16Williams EG, Sivathasan S, Anthony N, Eack SM, Mazefsky CA. Racial disparities in depression and mental health service use among black and white autistic adults. Scientific Reports, 2026. https://doi.org/10.1038/s41598-025-34696-8
- 17Carey ME, Graham Holmes L, Shea LL, et al. Therapy and Antidepressant Use in 8- to 29-Year-Old Autistic Medicaid Enrollees With Depression. Autism, 2026. https://doi.org/10.1177/13623613261453106
- 18Yuan J, Guo T, Rozear H, et al. Racial and ethnic disparities among older adults with autism spectrum disorder: a scoping review. Aging & Mental Health, 2026. https://doi.org/10.1080/13607863.2026.2612742
- 19Rava J, Shattuck P, Rast J, Roux A. The Prevalence and Correlates of Involvement in the Criminal Justice System Among Youth on the Autism Spectrum. Journal of Autism and Developmental Disorders, 2017. https://doi.org/10.1007/s10803-016-2958-3
- 20Webb L, Castillo IE, Holingue C, et al. Police Contact and Mental Health Among Adolescents With Autism: Findings From the UK Millennium Cohort Study. Journal of Adolescent Health, 2025. https://doi.org/10.1016/j.jadohealth.2025.05.022
- 21Yang K, Geng Y, Zhou W, et al. Closing the Gap in Autism Genetics: Population-Specific Variants and the Imperative for Global Inclusion. Biological Psychiatry, 2026. https://doi.org/10.1016/j.biopsych.2026.03.999
- 22Natividad Avila M, Jung S, Satterstrom FK, et al. Deleterious coding variation associated with autism is shared across ancestries. Nature Medicine, 2026. https://doi.org/10.1038/s41591-026-04228-6
Community sources
- c1Neff MA. Self-Care for Autistic People: 100+ Ways to Recharge, De-Stress, and Unmask! Adams Media, 2024. Chapter "Part Ways with Masking". Megan Anna Neff's site
- c2Price D. Unmasking Autism: Discovering the New Faces of Neurodiversity. Harmony Books, 2022. Chapter 2. Devon Price's site
- c3Henderson D, Wayland S, White J. Is This Autism? A Guide for Clinicians and Everyone Else. Routledge, 2023. Chapter "Connecting culture".
- c4Garcia E. We're Not Broken: Changing the Autism Conversation. Houghton Mifflin Harcourt, 2021. Chapter "Race". Eric Garcia's site
- c5Price D. Unmasking for Life: The Autistic Person's Guide to Connecting, Loving, and Living Authentically. Harmony Books, 2025. Chapter "The Rewards and Challenges of Unmasking". Devon Price's site