Understanding how rare genetic variants translate into real-world health outcomes has long been limited by the difficulty of recruiting carriers for in-depth study. This research offers one of the first empirical benchmarks for systematically identifying and recalling individuals with high-risk genomic variants from a diverse, healthcare-integrated biobank — a methodological step that could reshape how psychiatric genetics moves from discovery to clinical application.
Drawing on the BioMe biobank at Mount Sinai Health System, investigators recalled 892 participants stratified into three groups: carriers of rare copy number variants (CNVs) associated with neurodevelopmental disorders (NDDs), individuals with schizophrenia but no NDD CNVs, and neurotypical controls. Of those recontacted, 18% responded and just 8% completed the full protocol — yielding 73 participants total, with substantial multi-ancestry representation (37% African, 34% Hispanic, 26% European ancestry). Among the 30 NDD CNV carriers who completed assessments, 70% carried at least one neuropsychiatric or developmental diagnosis. Within the subset of 22 CNV carriers at loci specifically implicated in cognitive impairment, working memory scores on Digit Span Backward and Sequencing were significantly lower than controls (effect sizes of −1.76 and −2.01 respectively, both FDR-corrected p = 0.04). Interestingly, CNV carriers outperformed the schizophrenia group on verbal learning.
This work is methodologically incremental rather than paradigm-shifting, but its value lies precisely in the plumbing it exposes. Recall-by-genotype rates of only 8% completion underscore a persistent friction in genomic medicine: possessing genetic data does not readily translate into participant engagement, particularly in clinically diverse urban populations. The multi-ancestry composition is a genuine strength, addressing a well-documented bias in psychiatric genomics toward European-ancestry cohorts. However, the small final sample limits statistical power and generalizability of cognitive findings. Crucially, the observed neuropsychiatric burden among CNV carriers identified incidentally through a healthcare biobank — rather than clinical referral — suggests that many high-risk individuals remain undiagnosed. For health systems investing in genomic medicine infrastructure, these findings highlight both the feasibility and the logistical ceiling of population-level genotype-driven recall.