What the secretary's own data say about his autism comments

Comparing RFK's statements at the press conference to what is in the actual CDC paper

Much has been written in the last month about the denigrating comments about autistic people that RFK made at a press conference where the most recent CDC report on autism prevalence was announced. There are many takes on this, but I think the best is my fellow autistic person and Tar Heel, Eric Michael Garcia, on this podcast. Eric explains both why the comments are wrong and how they affect an autistic person. Stick around for the end when he talks about actual people who need constant support and you can get a feel for the empathy that lives in our community.

At the press conference, the secretary stated with little evidence that increases in autism prevalence had to be ascribed to an “environmental toxin” and that genetics could not cause an epidemic. He dismissed those of us who claim that the growth is due to better diagnosis and awareness as spreading a “canard.” Here’s my piece in the New York Times where I discussed this. Here in all caps are some specific quotes from RFK at the press conference:

“MOST CASES NOW ARE SEVERE. 25% OF THE KIDS WHO ARE DIAGNOSED WITH AUTISM ARE NONVERBAL, NONTOILET TRAINED, AND HAVE OTHER STEREOTYPICAL FEATURES -- HEAD BANGING, TACTILE AND LIGHT SENSITIVITIES, ETC.”

Here, he contradicts himself by saying “most” and then “25%.”

“WHAT WE NEED TO MOVE AWAY FROM TODAY IS THIS IDEOLOGY THAT THIS -- THAT THE AUTISM DIAGNOSIS, THAT THE AUTISM PREVALENCE INCREASE, THE RELENTLESS INCREASES ARE SIMPLY ARTIFACTS OF BETTER DIAGNOSIS, BETTER RECOGNITION, OR CHANGING DIAGNOSTIC CRITERIA.”

and then he said again,

“ONLY A VERY, VERY SMALL PORTION OF IT CAN BE CHARGED TO BETTER RECOGNITION OR BETTER DIAGNOSTIC CRITERIA.”

OK, so the press conference was to announce a study by the CDC, which is part of HHS and therefore under his responsibility. Wonder what it says?

Quotes from the actual CDC study

Go through the study carefully yourself; it’s here. It is true that the study shows an increase in the number of autism cases, but essentially all of the analysis and discussion is about better diagnosis leading to the increase — and better access to services being the best response. There is a passing mention of the correlation of lead levels with autism prevalence, which is modest and has been known for some time. Quotes from the study below in italics.

Research has not demonstrated that living in certain communities puts children at greater risk for developing ASD. Differences in the prevalence of children identified with ASD across communities might be due to differences in availability of services for early detection and evaluation and diagnostic practices.

If it’s an environmental insult of some kind, you would expect different rates in different communities where certain toxins occur at different rates; quote also states differences are probably due to services and diagnosis. So, if the secretary is trying to move on from the ideology that there is no epidemic, he hasn’t convinced his own agency of that.

The reversal of these patterns in prevalence by race and ethnicity and SES (socioeconomic status) is consistent with increased access to and provision of identification services among previously underserved groups.

Differences by race and income also attributed to access to services.

Differences in health outcomes between racial and ethnic groups including higher rates of ASD and co-occurring intellectual disability could be related to differences in the frequency of social determinants of health (SDOH) characteristics (26). SDOH include individual and community-level factors such as low income, housing and food insecurity, and transportation barriers.

Higher rates of intellectual disability by race probably due to community level factors related to socioeconomic differences (not environment).

Higher prevalence of intellectual disability (along with other neurologic disorders) might be related to higher rates of preterm birth, which is associated with brain injuries and neurodevelopmental impairment and also is linked to SDOH (27–32). In 2022, a total of 12.3% of births to Black mothers were preterm, compared with 8.7% of births to Hispanic and 7.6% of births to White mothers (33). Other causes of intellectual disability associated with SDOH include lead poisoning and traumatic brain injuries (34,35). SDOH could also contribute to disparities in access to early autism therapies, which have been found to increase cognitive and language scores (36).

This is the one place where an environmental insult (lead poisoning) is mentioned, but still mostly social determinants of health.

Improvements over time in early identification have been apparent in the ADDM Network (9–11,37). In 2022, across the 13 sites with higher cumulative incidence of identification by age 48 months among children aged 4 years (born in 2018) compared with children aged 8 years (born in 2014), identification was 40%–300% higher in the younger group.

Stating explicitly that improvements in diagnosis have happened since the last study.

One reason suspected diagnoses could still be seen after comprehensive evaluation might be related to misperceptions about the age at which autism can be reliably diagnosed. For example, during review for the 2022 surveillance year one child with suspected ASD had the statement “autism suspected but cannot test until 4 years old” in their evaluation. However, autism can in certain cases be reliably identified as early as age 1 year (40).

Autism can be diagnosed as early as age 1. So, any environmental insult that is responsible for all of this would have to come on pretty quickly.

Differences in prevalence over time and across sites can reflect differing practices in ASD evaluation and identification and availability and requirements that affect accessibility of services (e.g., meeting financial or diagnostic eligibility requirements).

Again, identification services noted as responsible for differences.

And the final sentence of the paper:

Opportunities exist to learn from successful policies, systems, and practices in different communities and implement approaches for equitable identification or service eligibility to help families or persons receive the support they need as early as possible to improve outcomes for children with ASD.

Nothing about eliminating environmental insults. “Let’s treat this by getting people more access to services.” That’s how the authors of the study chose to end their paper, which was published by the CDC.

Aftermath of the press conference

After the press conference occurred, there was a massive backlash to the disrespectful comments that RFK made. But there was also lots of fact-checking where his comments did not hold up well.

Immediately after the press conference, new data were shared at an autism conference that refuted the statement that “most cases are now severe,” because these data showed that the growth was among the mild cases.

Importantly, one of the authors of the new study, Maureen Durkin, was an author on the CDC report that RFK was discussing. From the news report on this:

“One possible explanation for the rise in mild cases, Durkin said, is that doctors have gotten better at identifying mild symptoms. Developmental screenings became more common over the study period, and the American Academy of Pediatrics began recommending universal screenings specifically for autism in 2006.”

Crucial to RFK’s alarmism is the idea that there is a massive increase in the number of severe cases, i.e., children with autism who also have intellectual disabilities. But Durkin et al.’s study was of the CDC’s own data from the same network used for the paper discussed at the press conference. Going back to earlier studies, these data showed that from 2000 to 2016, the number of children with moderate to significant impairment DROPPED from 1.5 to 1.2 out of 10,000. In fact, just looking at the next most recent study from the CDC group in 2021 and comparing it to the one just announced, the percentage of children diagnosed with autism who had intellectual disabilities went from 52% in 2018 to 40% in 2022, also a decrease. The simplest explanation for this is that most of the growth is among children without intellectual disabilities who are better diagnosed now.

Go read all this stuff for yourself and decide. All the links that you need are here.

Added on Sunday, May 25. The Wall Street Journal is out with an excellent story today on adults who received autism diagnoses that features me. Thanks to Nidhi Subbaraman for highlighting this.