This review was updated from the original version in June 2023.

I am approximately 100% certain that you will be 0% surprised to learn that there is bias in our systems for autism diagnosis and services. Autism was initially defined—and still is, to a large extent—through the lens of whiteness. Bias is kind of baked into the premise. Lately, we’ve seen some reflection about how we can reduce bias against autistic traits (such as here, here, here, and here), but we also know that racial bias is an intersecting concern (see here and here, as well as our previous reviews here, here, and here). We’re going to take some time this month to unpack what’s going on in the research on race and autism. 

Here’s the problem in a nutshell: Autism research and services are incredibly focused on white people.
 

Steinbrenner et al. did a large-scale systematic review of over 1000 autism intervention studies to give us a picture of what racial and ethnic groups are included in the research, and that picture is about as white as an overexposed polaroid of a polar bear in a snowdrift on a cloudy afternoon. In the studies that reported race and ethnicity information, 65% of study participants were white, vastly outnumbering participants who were Black (7.7%), Latinx (9.4%), or Asian (6.4%). Some groups are essentially excluded completely—a grand total of one (1) Native American and three (3) Native Hawaiian or Pacific Islanders were identified as participants across all of the gajillion studies included in this review. Perhaps even more concerning, only 26% of studies reported race/ethnicity data at all, and the rate of reporting is barely improving over time. Considering that in white-centering societies whiteness tends to be treated as a non-racialized “neutral” identity (meaning we usually only talk about race when we’re discussing people who aren’t white), it’s probable that the percentage of white participants is even more disproportionate in the studies that didn’t report that data. 

If the research disproportionately describes autism in white study participants, and those descriptions then inform how autism is diagnosed IRL, it’s a bit of a no-brainer that we’ll end up identifying autism from a white perspective. In a study of educational disability diagnoses of school-aged children in North Carolina, Kim et al. found that students with a primary educational classification of ASD were disproportionately white, male, and proficient in English. Higher rates of educational autism diagnosis were also associated with urban areas, lower rates of community poverty, and a greater number of health professionals per capita. 

The diagnosis gap isn’t just about clinicians. Azad et al. studied the beliefs and concerns that parents reported on an intake questionnaire for an initial diagnosis at an autism clinic, and found that not only were white parents more likely to believe that their child was autistic prior to diagnosis, their belief that their child was autistic was correlated with fewer/less frequent signs of autism as rated by clinicians. You could interpret this to mean that white parents are more likely to be unnecessarily preoccupied with autism, but also consider that white parents tend to be given more access to autism information and resources. Maybe they were better able to identify autistic traits in their children prior to diagnosis and use strategies that helped alleviate some of their difficulties.

There’s a snowball effect from all of these factors. Esler et al. found that children of white parents were more likely to receive an initial autism diagnosis from a clinic rather than through the public school system and that on average autism is diagnosed nearly two years earlier in clinics than in schools. Dallman et al. conducted a systematic review of 21 studies to identify which populations are more likely to access OT, PT, SLP, and ABA services following an autism diagnosis, and found that younger children were significantly more likely to receive more services than older children, so that two-year delay is likely to make a big difference in what services they are ultimately able to access. They also found preliminary evidence correlating better access to services with having a white caregiver, as well as with higher levels of parental income and education. 
 

It’s important to identify how systemic bias towards whiteness shapes our knowledge about autism and access to services (or lack thereof) for everyone else. Minoritized groups having less access to autism diagnosis and services isn’t the underlying problem; the problem is that our systems weren’t designed to serve everyone in the first place. That said, we did learn some things about how other racial groups are affected:
 

Kim et al. found that Black students were significantly more likely to be identified with an intellectual disability rather than autism as a primary educational disability label. Golson et al. provide us with some insight as to how that happens—they developed a vignette of a student based on the description of autism in the DSM-V and asked 229 school psychologists to rate the likelihood of the student receiving an autism diagnosis and how confident they were about their opinion. School psychologists expressed the least confidence in an autism diagnosis when the student was identified as Black.

The opposite situation seems to be the case for Asian children. Azad et al. found that Asian parents were the least likely of all racial groups to express a belief that their child was autistic prior to diagnosis, but clinicians tended to rate Asian children as having the most pronounced autistic characteristics of all racial groups. In the Golson et al. study, school psychologists were most confident in identifying autism in a hypothetical Asian student. Sociologists have identified a Model Minority myth that stereotypes Asians as being universally high-achieving and academically gifted within white-centering cultures, which can lead clinicians to predict that this population “should” be doing just-fine-to-fantastic at school. And, of course, “Asian” refers to a massively diverse group of people and cultures, some of which have tiny populations outside of their home countries (see here for a detailed overview). So for clinicians, good luck getting an interpreter, let alone an expert on cultural norms to help override or clarify your own biases. 

Yikes, right?! It’s a lot. The good news is that we do have some power to improve the situation—we’ll get into what we can do about all this in Part 2. 

Update May 2023:

The research continues to show us that these inequities are present across diagnoses. Casseus investigated the prevalence of co-occurring autism and ADHD in US children using data from the 2016–2018 National Survey of Children’s Health (NSCH). Results indicated that 1.2% of children between the ages of 3 and 17 were reported to be autistic with coexisting ADHD. The data revealed some compelling disparities, showing that children who were Black non-Hispanic or multi-racial/other non-Hispanic were significantly less likely to be identified as autistic with coexisting ADHD than white non-Hispanics.

Members can read more about family-centered care for Black autistic children here and about equity and autism as it relates to genderqueer people here.

Meredith Harold, Ph.D., CCC-SLP; V. Tisi, MS, CCC-SLP, Kyomi Gregory-Martin, Ph.D., CCC-SLP; Susan Douglas, Ph.D., CPSP, and Valerie Pitaluga, MS, CCC-SLP also contributed to this review.

 This review is part of our “Start Here” content guide for Language Assessment in EI.