Who Gets Identified as Autistic? Race, Ethnicity and Unequal Recognition

Who gets recognised as autistic still depends partly on race, ethnicity and migration. In the US, autism is now identified more often in Black, Hispanic and Asian or Pacific Islander children than in White children, the reverse of what surveillance found for years^[1][2]. But a much larger share of autistic Black children are identified with an intellectual disability as well, a sign that those without one may still be missed. In England, Australia and New Zealand, some groups remain under-identified.

Behind the numbers are ordinary moments. A family raises a concern and is told their child is just learning two languages. A child's distress at school is read as defiance. A parent does not know an assessment exists, fears what a label will mean in their community, or cannot find a clinician who speaks their language. None of these children is any less autistic for not being counted. They wait longer for understanding and support, or never get it.

In the US, the identification gap has reversed

The US Centers for Disease Control and Prevention (CDC) counts autistic 8-year-olds through its Autism and Developmental Disabilities Monitoring (ADDM) Network, which reviews health and education records at sites across the country. In its earliest years, the network “consistently reported lower overall ASD prevalence among Black and Hispanic versus White children”^[2]. In 2010 it identified 15.8 per 1,000 White children, 12.3 per 1,000 Black and Asian or Pacific Islander children, and 10.8 per 1,000 Hispanic children^[4].

By 2018 the rates were similar: 21.2 per 1,000 for White, 22.3 for Black, 22.2 for Asian or Pacific Islander and 22.5 for Hispanic children^[5]. In 2020, “for the first time”, the rate for White children (24.3) was lower than for Black, Hispanic and Asian or Pacific Islander children (29.3, 31.6 and 33.4)^[2]. In 2022, the rate among White children (27.7 per 1,000) was lower than among multiracial (31.9), Hispanic (33.0), Black (36.6), American Indian or Alaska Native (37.5) and Asian or Pacific Islander children (38.2). Asian or Pacific Islander children were identified at 1.4 times the rate of White children, and Black children at 1.3 times^[1].

The CDC does not attribute this to autism becoming more common. It says the reversal “is consistent with increased access to and provision of identification services among previously underserved groups”^[1]. A 2009 analysis of 2,568 8-year-olds who met a multisite surveillance network's criteria for autism on record review found that only 58% had a documented autism diagnosis. After adjusting for sex, IQ, birthweight and maternal education, Black (odds ratio 0.79) and Hispanic children (0.76) had lower odds than White children of having one, and for Black children the gap “persisted … regardless of IQ”^[6]. Many children, disproportionately children of colour, had the characteristics of autism described in their records without being identified as autistic.

Autistic children without an intellectual disability are still being missed

The overall rates have crossed, but a second gap has not closed. In 2022, among autistic children at ADDM sites who had cognitive test data, 52.8% of Black children had a co-occurring intellectual disability, as did 50.0% of American Indian or Alaska Native, 43.9% of Asian or Pacific Islander and 38.8% of Hispanic children. The figures for White and multiracial children were 32.7% and 31.2%^[1]. The pattern was the same in 2020 (50.8% of Black against 31.8% of White autistic children)^[2].

The CDC's cautious reading: “If Black children with ASD have less access to services than White children with ASD, as has been previously reported, the disproportionality in co-occurring intellectual disability might indicate an underascertainment of ASD among Black children without intellectual disability”^[2]. Put simply, Black children whose support needs are most apparent are being identified; others may not be. See also autism and intellectual disability.

Misdiagnosis is part of the story. In a US study of 406 Medicaid-eligible children, African American children were less likely than White children to be diagnosed autistic at their first specialist visit. Among children who were not, ADHD was the most common diagnosis, and African American children were more likely than White children to be given a diagnosis of adjustment disorder or conduct disorder rather than ADHD. The authors pointed to possible “variations in parents' descriptions of symptoms, clinician interpretations and expectations, or symptom presentation”^[7]. Later recognition also shows up in age at diagnosis, covered in age at autism diagnosis around the world.

In England, school records from 2016 and 2017 showed wide gaps between ethnic groups

England's school census records pupils with autism as a special educational need. A study of the January 2017 census covered 7,047,238 pupils aged 5 to 19 in state-funded schools, 119,821 of them recorded as autistic. Standardised for age and sex, 1.76% of pupils were recorded as autistic: 2.11% of Black pupils, 1.84% of White pupils, 1.06% of Asian pupils and 0.85% of Roma or Irish Traveller pupils, the lowest share^[8].

In the fully adjusted model, Chinese pupils were 1.38 times and Black pupils 1.26 times as likely as White pupils to be recorded as autistic, and Asian (0.83) and Roma or Irish Traveller pupils (0.42) less likely. Pupils whose first language was not English were 0.64 times as likely, and those ever eligible for free school meals 1.61 times as likely. Greater social disadvantage explained about 12% of the higher rate among Black pupils^[8].

The type of support tells its own story. Black and Chinese pupils were more likely than White pupils to have an autism-specific Education, Health and Care Plan (EHCP), the statutory plan for the highest needs, but “non-White pupils were less likely to receive support when compared with White pupils” at the lower level of school special educational needs support^[8]. That fits, though it does not prove, the US pattern of minority-group children being recognised mainly when their needs are highest. The authors list three possible explanations, “diagnostic biases, possible differences in detection and referral, or differential phenotypic prevalence”^[8], and census data cannot tell these apart.

A University of Oxford analysis of the 2016 census (pupils aged 5 to 16) found that Indian pupils had 0.46 times and Pakistani pupils 0.54 times the odds of White British pupils of being identified as autistic, and White Other pupils 0.60 times. Black Caribbean and Black Other pupils had 1.34 times the odds, a gap that largely disappeared once poverty and neighbourhood were taken into account. Asian pupils, though, were “still about half as likely” to be identified, and 79 of the 94 local authorities with enough data showed them under-represented. There was progress too: Bangladeshi pupils' odds ratio rose from 0.38 in 2005 to 0.79 by 2016^[3]. Among pupils newly recorded as autistic in 2014–2017, Asian, Black and Chinese girls not eligible for free school meals and speaking English as a first language had lower odds of identification than comparable White girls^[9] (see autistic women and girls).

Other countries show the same fault lines

The direction of the gap varies, but one thing recurs: autism with an intellectual disability is more often recognised in minority and migrant groups than autism without one.

Sweden. Among all children in Stockholm County in 2001–2007 (4,952 of them autistic), children of migrant parents had 1.5 times the odds of autism with an intellectual disability, but were less likely to be diagnosed with autism without one^[10]. The authors suggested environmental factors linked to migration, especially in pregnancy, might play a part. In our reading, under-recognition of autism without intellectual disability in migrant families would produce a similar pattern, and the authors' summary does not rule it out.

California. Among 1,626,354 births in Los Angeles County, children of foreign-born Black, Central and South American, Filipino and Vietnamese mothers, and of US-born Hispanic and Black mothers, were more likely than children of US-born White mothers to be diagnosed with autistic disorder at ages 3 to 5, both overall and with a co-occurring intellectual disability. The researchers framed the study around the possibility that such differences “may indicate etiologic heterogeneity or different thresholds for diagnosis”, and called for research into how migration bears on both the causes and the identification of autism^[11].

The Netherlands. Ethnic minority children were under-represented among 712 children referred to Dutch autism services, compared with the known community prevalence. When 81 paediatricians read case descriptions, they suggested autism more often for Dutch children than for Moroccan or Turkish children. When the paediatricians were asked to rate the probability of autism explicitly, the ethnic difference disappeared^[12].

Australia. Among children born in Western Australia in 1984–1999, those of Aboriginal mothers had lower odds of an autism diagnosis than children of Caucasian mothers (the study's term), but higher odds of a mild to moderate intellectual disability diagnosis. The researchers wrote that “Aboriginal children may be more readily assigned an ID than an ASD label”^[13]. In the Kimberley, when a psychiatrist took developmental histories from 14 selected Indigenous adults with a diagnosis of schizophrenia, 13 were considered autistic^[14]. A 2020 scoping review concluded that the findings “suggest similar prevalence rates for autism among Indigenous and non-Indigenous Australians”, although some Aboriginal and Torres Strait Islander people “may not receive a diagnosis or may be misdiagnosed”^[15]. A 2025 meta-analysis found no statistically significant difference in autism prevalence between Aboriginal and non-Aboriginal children, with an imprecise estimate (pooled odds ratio 0.43, 95% confidence interval 0.11 to 1.58); in Canada, studies found higher recorded prevalence among non-Indigenous children^[16].

Aotearoa New Zealand. A national study carried out with autistic Māori covered all 1,565,505 people aged 0 to 24 in 2018. Autism was identified in 70.9 per 10,000 Māori, against 78.3 per 10,000 non-Māori. Autistic Māori had 1.12 times the odds of also having an intellectual disability diagnosis, and 1.35 times the odds of receiving high-needs school funding. The authors say the lower rate “is likely to reflect ongoing inequities and systemic racism, not true prevalence differences”, and that autistic Māori with lower support needs may be missing out^[17].

Why identification differs: access, bias and assessment built for one culture

Identification is a chain of noticing, referral, assessment and recording. Each link can break differently for different communities.

• Assessment tools. A review of four diagnostic and six screening tools for autism found they needed improvement for culturally and linguistically diverse children, who “will likely continue to be misidentified (or not identified at all)” without better assessment^[18].
• Clinician judgement. Bias in the Dutch paediatricians' referrals disappeared with structured ratings^[12], and in the US Medicaid study African American children were less often diagnosed autistic at a first specialist visit^[7].
• Language. Subtle social-communication differences “may be more difficult to identify if the first language of the assessor and pupil are not congruent”^[3], which may help explain the lower rate among pupils in England whose first language is not English^[8].
• Stigma, trust and racism. A review of 24 US qualitative studies found “familial, cultural, and structural barriers”^[19]. A 2025 review of 56 studies found that racially and ethnically minoritised families struggled to find services meeting their cultural and language needs, and “experienced disability-based stigma and racism”^[20]. Aboriginal and Torres Strait Islander families described shame and isolation, but also strong traditions of acceptance^[21].

None of these studies shows that autism itself is more or less common because of race. In the US surveillance data, race and ethnicity “reflect what is documented in records rather than how families or persons might prefer to identify”^[1]; in England's school census they are broad self-reported categories^[8]. Other suggested causes, such as factors linked to migration, remain hypotheses. For the wider picture in UK health care, see our study of ethnicity and mental health care.

What the evidence asks of us

An identification rate is, first of all, a measure of how well services reach people. The evidence points to five things services, commissioners and funders can do.

1. Count it, by ethnicity and by support need. The Oxford team said councils, academy trusts and schools “should monitor ethnic disproportionality in the identification of SEN”^[3]. Totals are not enough: in the US and New Zealand the gap hides in who is identified without an intellectual disability^[1][17].
2. Use structured, culturally responsive assessment. Structured ratings reduced ethnic bias in the Dutch study^[12], and reviewers have called for better tools for culturally and linguistically diverse children^[18].
3. Train professionals in how differently autism can look. US researchers called for “continued professional education related to the heterogeneity of the presentation of ASD”^[6]. Another diagnosis, whether a behavioural one or an intellectual disability, should not rule autism out^[7][13].
4. Build outreach with communities, not at them. The Oxford researchers urged schools and councils to “raise awareness of ASD among Asian communities, improve outreach and review the extent to which the services are configured appropriately for access by ethnic minority groups”^[3]. The New Zealand team called for “Kaupapa Māori, autistic-led research”^[17].
5. Make services worth trusting. Reviewers call for practice that is “neurodiversity-positive, culturally affirming, and financially, geographically, physically, socially, and culturally accessible”^[20].

The US reversal suggests that identification gaps can close within about a decade^[1]. For how these counts fit into global estimates, see how many autistic people are there; for the full library, see Autism & Health research.

Language: this piece uses identity-first language ("autistic people"), the preference of most autistic-led organisations, and uses clinical terms such as "autism spectrum disorder" only when naming a study or diagnostic category. Ethnic group names follow each dataset's own categories.

Why we're publishing this

Health Insurance UK is a commercial health-insurance resource, not a charity and not a campaign. Families and organisations in minority ethnic, migrant and Indigenous communities have long raised concerns that autistic people among them are missed, misdiagnosed or recognised late, and the data backs them up. We have put the evidence in one place, cleanly sourced, for commissioners, schools and policymakers. The case belongs to those communities.

How to read this data

Almost every figure here measures identification: who has an autism diagnosis or school record, not how many people are autistic. US ADDM figures come from health and education records at a changing set of sites, so years are not a strict time series. England's figures come from school special educational needs records, which miss children not recognised at school and pupils outside state schools. Ethnic categories differ between countries and cannot be compared across them. Odds ratios and prevalence ratios are given as the studies report them; many are adjusted for other factors, which changes what they mean. The evidence is concentrated in the US and England, and studies of Indigenous communities are few and small.

Use this data

Free to cite with attribution to Health Insurance UK. Charities, campaigners and journalists are welcome to reuse these figures and charts. For the underlying figures as a spreadsheet, get in touch.

Important: This article summarises published research for general information. It is not medical advice and cannot be used to diagnose, treat or predict anything about an individual. If you have concerns about your own or someone else's health, please speak to a qualified health professional.

Sources

1. Shaw 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. cdc.gov/mmwr/volumes/74/ss/ss7402a1.htm
2. Maenner 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. cdc.gov/mmwr/volumes/72/ss/ss7202a1.htm
3. Strand S, Lindorff A. Ethnic disproportionality in the identification of Special Educational Needs (SEN) in England: Extent, causes and consequences. Executive summary. University of Oxford Department of Education, 2018. education.ox.ac.uk/wp-content/uploads/2018/08/Executive-Summary_2018-12-20.pdf
4. Autism and Developmental Disabilities Monitoring Network Surveillance Year 2010 Principal Investigators. Prevalence of Autism Spectrum Disorder Among Children Aged 8 Years — Autism and Developmental Disabilities Monitoring Network, 11 Sites, United States, 2010. MMWR Surveillance Summaries, 2014. cdc.gov/mmwr/preview/mmwrhtml/ss6302a1.htm
5. Maenner MJ, Shaw KA, Bakian AV, et al. Prevalence and Characteristics of Autism Spectrum Disorder Among Children Aged 8 Years — Autism and Developmental Disabilities Monitoring Network, 11 Sites, United States, 2018. MMWR Surveillance Summaries, 2021. cdc.gov/mmwr/volumes/70/ss/ss7011a1.htm
6. Mandell 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. europepmc.org/article/MED/19106426
7. Mandell 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. europepmc.org/article/MED/17160456
8. Roman-Urrestarazu A, van Kessel R, Allison C, Matthews FE, Brayne C, Baron-Cohen S. Association of Race/Ethnicity and Social Disadvantage With Autism Prevalence in 7 Million School Children in England. JAMA Pediatrics, 2021. europepmc.org/article/MED/33779707
9. Roman-Urrestarazu A, Yang JC, van Kessel R, et al. Autism incidence and spatial analysis in more than 7 million pupils in English schools: a retrospective, longitudinal, school registry study. The Lancet Child & Adolescent Health, 2022. europepmc.org/article/MED/36302393
10. Magnusson C, Rai D, Goodman A, et al. Migration and autism spectrum disorder: population-based study. British Journal of Psychiatry, 2012. europepmc.org/article/MED/22361019
11. Becerra TA, von Ehrenstein OS, Heck JE, et al. Autism spectrum disorders and race, ethnicity, and nativity: a population-based study. Pediatrics, 2014. europepmc.org/article/MED/24958588
12. Begeer S, Bouk SE, Boussaid W, Terwogt MM, Koot HM. Underdiagnosis and referral bias of autism in ethnic minorities. Journal of Autism and Developmental Disorders, 2009. europepmc.org/article/MED/18600440
13. Leonard H, Glasson E, Nassar N, et al. Autism and intellectual disability are differentially related to sociodemographic background at birth. PLoS ONE, 2011. europepmc.org/article/MED/21479223
14. Roy M, Balaratnasingam S. Missed diagnosis of autism in an Australian indigenous psychiatric population. Australasian Psychiatry, 2010. europepmc.org/article/MED/21117839
15. Bailey B, Arciuli J. Indigenous Australians with autism: A scoping review. Autism, 2020. europepmc.org/article/MED/31928063
16. Lau SC, Czuczman NM, Dennett L, Hicks M, Ospina MB. Systematic Review and Meta-Analysis: Prevalence of Neurodevelopmental Disorders Among Indigenous Children. JAACAP Open, 2025. europepmc.org/article/MED/40922776
17. Tupou J, Kewene F, Morrison-Young I, et al. Identification of autistic Māori in Aotearoa New Zealand: A national cross-sectional study. Autism, 2026. europepmc.org/article/MED/41578668
18. Harris B, Barton EE, Albert C. Evaluating autism diagnostic and screening tools for cultural and linguistic responsiveness. Journal of Autism and Developmental Disorders, 2014. europepmc.org/article/MED/24186120
19. Singh JS, Bunyak G. Autism Disparities: A Systematic Review and Meta-Ethnography of Qualitative Research. Qualitative Health Research, 2019. europepmc.org/article/MED/30474491
20. Sterman J, Wagland Z, Scott-Cole L, Spassiani N, Njelesani J. Navigating racism, stigma, and autism services: A scoping review of the lived experiences of racially and ethnically minoritized families. PLOS Mental Health, 2025. europepmc.org/article/MED/41662117
21. Lilley R, Sedgwick M, Pellicano E. Inclusion, acceptance, shame and isolation: Attitudes to autism in Aboriginal and Torres Strait Islander communities in Australia. Autism, 2020. europepmc.org/article/MED/32529835