Schizophrenia-associated polygenic liability and structural genomic risk demonstrate broadly distributed neuropsychiatric associations in the All of Us Research Program

BackgroundSchizophrenia liability reflects both the aggregate effects of many common genetic variants and rare structural variation at recurrent neurodevelopmental copy-number variant (ND-CNV) loci. Although both components of genetic liability have been associated with schizophrenia and related phenotypes, it remains unclear whether they exhibit similar neuropsychiatric association profiles or show evidence of interaction when examined jointly. MethodsWe analyzed 71,992 participants from the All of Us Research Program Controlled Tier version 8 dataset, including 2,115 participants carrying structural variants overlapping curated recurrent ND-CNV loci. We integrated binary ND-CNV carrier status, an ancestry-adjusted and standardized reduced schizophrenia polygenic risk score (PRS), and longitudinal electronic health record-derived psychiatric and neurodevelopmental phenotypes. Logistic regression models evaluated pooled, deletion-overlapping, and duplication-overlapping carrier associations, schizophrenia PRS effects, and PRS x ND-CNV interactions. Dimensional analyses, confirmatory factor analysis (CFA), and structural equation modeling (SEM) provided complementary summaries of phenotype covariance and assessed the robustness of primary associations across alternative representations of the phenotype space. ResultsSchizophrenia PRS and ND-CNV carrier status showed distributed association profiles extending beyond schizophrenia diagnosis across psychosis-spectrum, neurodevelopmental, cognitive, and internalizing phenotypes. ND-CNV carriers had higher estimated odds of schizophrenia diagnosis than non-carriers (OR = 1.49, 95% CI 1.09-2.03), while each 1-SD increase in schizophrenia PRS was associated with higher odds of a broad psychosis-spectrum phenotype (OR per SD = 1.07, 95% CI 1.03-1.10). Deletion-overlapping carriers exhibited generally larger and more consistently positive association estimates than duplication-overlapping carriers. Psychosis-spectrum prevalence increased from 5.4% in the lowest PRS decile to 6.7% in the highest, while ND-CNV carriers maintained consistently higher prevalence than non-carriers across the PRS distribution. PRS x ND-CNV interaction estimates remained near the null, providing little evidence of widespread multiplicative interaction. Among tested covariance structures, bifactor models showed the lowest residual misfit and were therefore carried forward for SEM analyses. ConclusionsSchizophrenia-associated polygenic liability and structural variation overlapping recurrent ND-CNV loci exhibit distributed neuropsychiatric association profiles with limited evidence of widespread multiplicative interaction. Their joint evaluation in the large, ancestrally diverse All of Us cohort demonstrates that distinct forms of genomic liability can be examined within a shared, real-world phenotype space extending beyond schizophrenia diagnosis alone.