Genetics in Context: Social Factors Modify Polygenic Risk Across Multiple Cardiometabolic Diseases

Abstract

Polygenic risk scores (PRS) are increasingly used for disease prediction, but whether their associations vary across social contexts remains poorly understood. Here, we leveraged the large and diverse All of Us cohort to examine interactions between PRS and social determinants of health (SDoH) across eight prevalent chronic diseases and developed disease-specific polysocial risk scores (PsRS). We found consistent evidence of SDoH modification of PRS- disease associations for cardiometabolic diseases, including type 2 diabetes, coronary heart disease and chronic kidney disease, with PRS associations attenuated among more socially vulnerable populations. Joint modeling of PsRS and PRS significantly improved case status prediction across all eight diseases and PsRS were predictive for most diseases in external cohorts. These findings demonstrate that social context modifies associations between genetic risk and disease and suggest that integrating genetic and social information can refine disease prediction across diverse populations. TeaserBoth genetics and the environment influence chronic disease, but whether and how they interact remains understudied. We find evidence of genetic effect modification of polygenic risk scores (PRS) by social factors for multiple cardiometabolic traits and that jointly modeling social risk improves disease prediction across a range of chronic diseases.