The campaign for the term “profound autism” features neurodiversity-averse parents of high-support autistic people. This cohort openly struggles to understand what autism and disability mean for their children, yet also tends to decline insights from the autistic and disability communities—preferring to come up with solutions uninformed by lived autistic experience.

Unsurprisingly, this “profound autism” effort dismisses critiques from autistic professionals, including those working with other parents of high-support autistics, on how the proposed label would interfere with accessing appropriate supports, while literally lobbying that the term is needed to access supports. The campaign also sidesteps feedback from autistic autism researchers on “‘profound autism’’s limited scientific validity and feasibility in everyday practice.”

The latest effort from the contingent is a new paper, The Prevalence and Characteristics of Children With Profound Autism, which applies “a newly proposed definition of profound autism to surveillance data to estimate the percentage of children with autism who have profound autism and describe their sociodemographic and clinical characteristics.” We spoke with autism researcher Dr. Kristen Bottema-Beutel about how aspects of the paper are misleading, which conflicts of interest it fails to mention, and why readers and researchers alike should be skeptical of the authors’ proposed criteria for defining “profound autism.”


Thinking Person’s Guide to Autism (TPGA): You are known for taking a broader look at disquieting trends in autism research, like non-disclosures of conflicts of interest, and low standards for evidence-based practices. Does this paper raise any red flags?

Dr. Bottema-Beutel: This paper does raise some red flags. I think it is notable that the conflict of interest statement reads “the authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.” However, one of the authors is a co-founder and president of the Autism Science Foundation, which has put forward position statements and funding initiatives on the concept of profound autism. This same author is also a member of the National Council on Severe Autism (NCSA), which is an advocacy organization that has been pushing for bifurcating autism into separate diagnoses for quite some time now. They also use a direct quote from the NCSA website as part of the rationale for their study, without indicating that an author of the study is a member of that organization.

The journal guidelines indicate that it follows COI policies outlined by the International Committee of Medical Journal Editors, which requires disclosure of these types of relationships. Being transparent about COIs allows readers to approach the research with an appropriate level of scrutiny in terms of the methods they used and their interpretations of the findings. Not disclosing those relationships undermines trust in the research process.

TPGA: As often happens with provocative autism studies, this paper made some headlines. Can you clarify why it’s not a great idea for media outlets to cite a single paper or exploratory study as evidence of a greater trend or phenomena?

Dr. Bottema-Beutel: Any single study is subject to limitations. I think it would be problematic to present this paper as evidence of consensus in the field, or even as evidence of the utility of the approach. The authors relied on record review to make classifications, which is subject to multiple forms of measurement error. Especially in this case, since the records the researchers had access to were not designed to classify subjects with “profound autism” (since the term did not exist at the time the records were created).

For example, one of the ways the authors’ determined if subjects were “non-verbal” or “minimally verbal” was to see if those terms appeared in clinical records. Researchers are currently not using these terms consistently or with clear definitions, and it is very doubtful that clinicians are either. It is therefore not at all clear to me what the language profiles of these children actually are.

TPGA: Speaking of headlines, Time magazine cites this paper as documenting a “rise” in autism, following the CDC’s recent update of the autism prevalence rate to 1 in 36. Can you talk about the difference between autism prevalence rising and the actual numbers of autistic people rising?

Dr. Bottema-Beutel: Prevalence refers to the number of people who are actually identified as autistic. These numbers are going up, but most experts attribute this increase to changes in diagnostic criteria and better access to healthcare among groups that have historically been excluded (e.g., Black and Latinx children). So these prevalence estimates mean that fewer autistic people are being gatekept from access to services, and not necessarily that there are any changes in the number of autistic people.

TPGA: As an autism researcher, do you have any other concerns about the ways in which the proposed term “profound autism” as defined in this paper will limit rather than support autistic people as a population?

Dr. Bottema-Beutel: My big concern is that there is no explanation as to how this classification system (I’m not exactly sure what to call it because the original proposal for the term published in the Lancet referred to the profound autism designation as an “administrative term,” but people seem to be using it in a wide variety of ways) will actually increase access to services or improve services planning.

The current system outlined in the DSM-5 is to classify autistic people according to their support needs. The authors of the paper published in Public Health Reports are correct that this classification system needs work, but they don’t provide even a logic model as to how the system they propose will improve providing services to people who need it. They rely on big assumptions that poorly defined criteria such as IQ scores and parent/clinician reports of verbal ability tells us something about the level and type of services that autistic people need. But others have pointed out (see for example Steven Kapp’s work on how scores on IQ tests vary across different tests and subdomains) that these relationships are not clear cut; many autistic people who do not have the characteristics the authors focus on have significant support needs, and support needs can be radically different across different domains, social contexts, and time.

If the goal is to make sure that everyone gets the support they need, why not improve our ability to directly assess support needs? Why focus on features that currently do not have well-characterized relationships with support needs? I haven’t seen any clear explanations about how this is meant to work, especially given that this “new” system looks a lot like the pre- DSM-5 system—and I don’t think there is any evidence that service provision was better under that system.


COI Disclosure: Dr. Bottema-Beutel has accepted speaker fees to discuss conflicts of interest in autism research.

Steampunk-style painting of two white women with aubergine hair and dark lab glasses, in a white lab coats, with their backs to each other.
Steampunk-style illustration of two white women scientists with their backs to each other. Image by DALL-E.