Mount Sinai Revises Phelan-McDermid Prevalence to 1 in 7,300 Using 180k-Person Genetic Dataset

The team aggregated de-identified variant data from commercial labs (GeneDx, Ambry, Labcorp) and cohorts (SPARK, Autism Sequencing Consortium, pediatric hospitals). They modeled ascertainment bias by estimating the fraction of PMS cases missed due to absent or incomplete SHANK3 sequencing and the share of non-autistic PMS cases. This produced the upward revision from prior 1-in-15,000 figures. The analysis highlights that current diagnostic pipelines systematically under-sample developmental-disability populations without autism labels.

Broader access to exome or targeted SHANK3 testing would link thousands more individuals to emerging trials of IGF-1 mimetics and other precision agents already in Phase 2. Because PMS accounts for up to 1 % of autism cases, improved detection also sharpens stratification in larger autism trials. Insurance barriers and incomplete gene coverage remain the dominant bottlenecks identified by the authors.

The key limitation is reliance on secondary genetic records without uniform clinical validation or population-based sampling; a prospective newborn-screening or biobank study with 500,000 participants would strengthen causal claims about true prevalence and treatment access.