To examine variation by gender and race/ethnicity in associations of dichotomous and density FH measures with the alcohol variables, we ran all AUC analyses, linear, and logistic regression models separately for males, females, Whites, and Blacks. In all regression models, analyzed separately for males and females, we included assessment-age and self-reported race/ethnicity as covariates. Whereas, for all regression models analyzed separately for Whites and Blacks, we included assessment-age and gender as covariates. We also examined interactions between gender, race/ethnicity, and each of the FH measures across the outcomes. These findings are reported in the supplementary material.