A syndrome thought vanishingly rare may affect more than 45,000 Americans

“Many individuals with developmental disabilities and autism are never offered genetic testing.” That sentence, from the team at the Seaver Autism Center for Research and Treatment at Mount Sinai, explains why a condition long counted among the rarest may be several times more common than assumed. Their new estimate for Phelan-McDermid syndrome is 1 in 7,300 people, or 13.7 per 100,000.

Scaled to the United States, that rate means more than 45,000 Americans could be living with the syndrome, most without a diagnosis. The number is a modelled projection, not a registry count, and the Seaver team presents it as the likely size of a hidden population rather than a tally of known patients.

Seaver’s Count of Nearly 180,000 Genetic Tests

Tess Levy, an assistant professor of psychiatry at Mount Sinai, is first author of the paper, and Joseph D. Buxbaum, director of the Seaver Autism Center, is senior author. Instead of testing a cross-section of the general public, which would take an impractical number of sequencing runs, the team went where the syndrome is most likely to surface: people with autism who had already had genetic testing. Phelan-McDermid syndrome is caused by changes to SHANK3, a gene on chromosome 22, and a lab that sequences a child with autism will often see it.

The pool was 179,837 individuals with autism drawn from ten independent sources, including the commercial laboratories GeneDx, Labcorp and Ambry Genetics, the research cohort SPARK, the Autism Sequencing Consortium and several children’s hospitals. Counting how many carried a SHANK3 disruption gave a rate inside that tested, autistic population.

Turning that into a figure for everyone required adjustments, which Mount Sinai’s account of the work and the paper describe. The analysis accounts for undiagnosed cases, for limits in the accuracy of genetic testing, and for people who have the syndrome but do not meet criteria for autism. Those corrections are the step that turns a laboratory tally into a population estimate, and they are also where the most judgment enters. A rate measured among tested people with autism says nothing by itself about how many untested people carry the same genetic change, so the model has to bridge that gap with assumptions about who gets tested and how reliably the tests find the gene.

From 1 in 15,000 to 1 in 7,300

The advocacy group CureSHANK, which follows the research closely, gives the earlier estimate as 1 in 15,000. The new figure roughly doubles it, which is a modest-sounding revision with large consequences: at the old rate, the same American population would hold roughly half as many people with the syndrome. The gap between that working number and the count of people actually diagnosed is the study’s central message: the syndrome has been rare on paper partly because it has rarely been looked for.

HealthDay’s summary of the paper describes the condition as a rare genetic disorder that causes medical, intellectual and behavioral challenges, and repeated the 13.7-per-100,000 rate for the general population. Without a genetic test, an affected child can carry a broader developmental label for years while the cause stays unnamed, and the syndrome then never enters the databases that depend on a specific diagnosis.

CureSHANK board chair Geraldine Bliss put the consequence bluntly: there are likely tens of thousands of people with the syndrome who have never received a genetic diagnosis, and finding them matters more now that several therapies are advancing into clinical trials.

Error Bars on the 45,000 Figure

Several features limit confidence in the 45,000 count. The underlying sample consists of people who were referred for testing, which is not the same as a random draw from the population, and the adjustments for undiagnosed and non-autistic cases are modelled assumptions rather than counts. The researchers also point to families who face insurance barriers or limited access to genetics services, and to tests that evaluate SHANK3 inadequately.

Buxbaum framed the practical stakes in terms of drug development. “Knowledge is power,” he said. “These genetic findings allow researchers to design more targeted clinical trials for potential therapies.” A company deciding whether enough patients exist to run a study needs exactly this kind of denominator, and so does a hospital weighing whether broader testing for children with unexplained developmental delay would find enough cases to be worth the cost.

The study appeared in Autism Research (2026; volume 19, issue 8). Whether any large share of the projected 45,000 will ever be identified is beyond what it can say, since that depends on how many families are offered a SHANK3 test and whether the test is adequate. Bliss’s own wording, tens of thousands never diagnosed, sits close to the Seaver projection.

This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.

More from Morning Overview