We acknowledge there are limitations in the study. First, while sample size is relatively large for alcohol challenge studies, it is small with regard to genetic analysis. Second, the sample size for genotype groups was insufficient for analysis of data disaggregated by sex or race/ancestral origin, or for correction for testing multiple hypotheses. Sex-related effects of OPRM1 G allele have been shown for experimental pain thresholds (Fillingim et al., 2005), magnitude of cortisol response to pharmacological blockade of the mu-opioid receptor by naloxone (Lovallo et al., 2015), and alcohol-induced DA release (Ramchandani et al., 2011). Third, the frequency of the OPRM1 G-allele differs substantially across different race/ethnicities (Oroszi and Goldman, 2004). To address this concern, sex and ancestral origin were covariates in our analyses. We also completed sensitivity analyses with the subsample of Caucasian participants. Fourth, focusing on two genotypes out of many does not account for the influence of other genes that affect alcohol sensitivity and consumption, or possible bias due to genome wide polygenic effects. For example, genetic polymorphisms of GABAA -α2 receptors, and GABAA-α6 receptors, nicotinic acetylcholine
