Background: Polygenic Risk Scores (PRS) estimate genetic predisposition for disorders like bipolar disorder (BD), but much of the risk remains unexplained, suggesting environmental influences. The interplay of genetic and social risk factors is understudied in large, diverse populations. Here, we examine the combined and interactive effects of a PRS for BD (PRSBD) and social risk factors on BD status in the All of Us Research Program.

Methods: We used data from genotyped participants in the All of Us cohort. BD cases and controls were identified using an electronic health record (EHR)-based algorithm. We applied a pre-calculated PRSBD from the Psychiatric Genomics Consortium. Social risk factors—including adverse childhood experiences, discrimination, and perceived stress—were derived from participant surveys. We used logistic regression to test the independent and interactive effects of standardized PRSBD and social risk on BD status, controlling for age, sex/gender, and genetic ancestry. Results: The PRSBD was validated in the All of Us cohort, showing a significant association with BD status that increased across higher risk deciles. Social risk factors also demonstrated a strong, independent association with BD.

Conclusion: Integrating PRS with comprehensive EHR and survey data helps clarify BD's complex etiology by underscoring the importance of social context in risk evaluation. These findings can inform more holistic prediction models and lay the groundwork for targeted prevention strategies addressing both genetic and environmental risk.