Looking for Autism in All the Wrong Places

The NIH's new autism research program is an exercise in throwing darts. Instead it should be following the signposts lit by decades of scientific discovery.

Following on health secretary Robert F. Kennedy, Jr.’s stated intent to accelerate research into the origins of autism, the National Institutes of Health has announced the $50 million Autism Data Science Initiative, inviting researchers to propose large studies based on integration of human health and exposure data. It flags “newly emerging and/or understudied factors that may contribute to autism” as a key area for inquiry.

It sounds good. But its approach is not sound science. It calls for an unnecessary dart throwing exercise, while overlooking clear priority paths lit by decades of investment in autism research.

Yes, autism’s causes for the most part remain mysterious, and yes inquiries across large datasets can help find new answers. But most questions have already been answered.

While autism is hardly a single condition with a single cause, there is overwhelming evidence that the disorder is largely innate, and not caused by exposures to the fetus in utero, nor to early childhood exposures like vaccines. Hundreds of studies on such factors have already failed to show significant associations, with rare exceptions such as maternal use of the anti seizure medication valproic acid.

Instead, it has become clear that autism arises due to abnormalities in how the brain wires up, beginning in the earliest weeks of embryonic brain development. These impairments are sometimes driven by defects in specific genes, but in the great majority of cases the genetic code is intact. What’s typically going awry is what is known as gene expression, meaning the genetic code is not being read correctly.

It is this rather unexplored question that should be the focus of our federal research dollars. Again, there is only weak evidence for the idea that RFK Jr. has promoted that we are “doing something” to otherwise typically developing kids to damage their brains. The mis-wiring of autism starts long before birth.

Because of defects in the expression of genes, brain neurons can fail to develop, proliferate and move to their proper locations, particularly in the outer part of the brain called the cerebral cortex. These micro-structural defects can lead to impairments in sensory processing, learning, and social behaviors, a set of phenotypes we label as autism.

An autistic child’s own cell biology develops gene activity of autism, notwithstanding any exposures. In cell culture models (think “mini-brains in a dish”), we can foretell a child’s level of symptoms in brain and behavior at two years old just by looking at the innate cellular signals.

But only a small percentage of autism patients have an “autism risk gene mutation.” The large majority of individuals with autism – an estimated 80 to 90% — seem to be free of these genetic defects, and the specific causes of autism in these individuals remain unknown. Further, many autism “risk” gene mutations occur as commonly in typical as in autism children, suggesting that the 80 to 90% of cases being non-genetic could be an underestimate.

One hypothesis that has grown popular in some corners is that a person’s “polygenic risk score” somehow leads to autism. What this means is that normally harmless and common genetic variants could conceivably conflate in certain people to push neurodevelopmental trajectories over a threshold, resulting in impaired brain development.

Unfortunately there is little actual evidence to support this idea. Instead, the studies show that autism is strongly heritable, a much broader concept that may have nothing to do with polygenic risk. We know autism is highly heritable because of the strong recurrence of autism among identical twins and siblings. The disconnect between the strong data on heritability and the weak findings from genetics is what the field calls “the missing heritability of autism.”

With that little science lecture in mind, we can see there are two major unanswered questions about the origins of autism. First, what is responsible for the errors in gene expression in the early autism brain? And second, what factors lie behind autism’s missing heritability?

Finding those answers to both those key questions requires us to focus on factors and exposures, including to toxicants, that happen before conception, even long before conception. Most notably, parents’ reproductive cells are not just simple “genes” — they are packed with other factors that help control gene expression. But such questions have been largely off the radar of autism research. As I have long argued it’s most likely exposures to the parents’ germ cells, not the fetal brain, that precipitate the faulty gene expression that is the hallmark of autism. And I have explained this with the key example of a parental history of procedures under general anesthesia.

Federal research dollars should address the big scientific elephants in the room like this one, but it seems we will see a $50 million fishing expedition instead.

Jill Escher is an autism research advocate and president of the National Council on Severe Autism. Learn more: jillescher.com