728. Environmental, psychiatric, and genetic predictors of alcohol use disorder criterion count
Abstract Background Genetic and environmental factors, and psychiatric traits, contribute to risk for alcohol use disorder (AUD) and other substance use disorders (SUDs) and traits. Weighing the importance of the different kinds of contribution to risk is important in understanding risk prediction – which may be important clinically -- and in formulating prevention and treatment strategies. Aims & Objectives We worked to quantify the contribution of environmental, psychiatric, and genetic factors to AUD across different ancestries. Although polygenic risk prediction represents an important future application, its utility may be enhanced when considered in the context of environmental predictors. We used deeply phenotyped African- and European-ancestry (AFR and EUR) samples to examine how genetic, psychiatric, and environmental factors predict AUD criterion count. Method We analyzed data from 11,021 individuals in the Yale-Penn Study sample, 5,843 AFR and 5,178 EUR. Polygenic risk scores (PRSs) were generated from genome-wide association studies (GWAS) of problematic alcohol use (PAU). Generalized linear regression and relative importance analyses determined the independent and interactive effects of environmental and genetic factors on AUD, considering criterion count rather than binary diagnosis to maximize power. Results PRSs for PAU were positively associated with AUD criterion count in both ancestries and predicted 1.9% of variance in AUD criterion count in the EUR sample and 1.3% in the AFR sample. A combination of education, substance use in the household before age 13, annual household income, and male sex explained 73.1% of the variance in AUD criterion count in the AFR sample and 58.9% in EUR. Among examined psychiatric disorders, posttraumatic stress disorder explained the most variance (10.0% in AFR, 9.4% in EUR), followed by anxiety disorders (3.4% in AFR, 6.2% in EUR) and major depressive disorder (1.3% in AFR, 2.1% in EUR). In the EUR sample, education level moderated the relationship between PRS for PAU and AUD criterion count. We completed similar analyses for several other SUDs, including opioid use disorder (OUD), which will also be discussed. Discussion & Conclusions In both AFR and EUR, environmental factors explained most of the variance in AUD criterion count, but polygenic risk was also a statistically significant predictor. The pattern of results for OUD was similar. These findings may help inform clinical, research, and policy efforts to mitigate AUD and OUD risk.