We included participants’ age at T1 and sex as covariates given their demonstrated associations with alcohol use. We also included the first 3 genetic ancestry principal components (PC1-PC3) as covariates to account for potential population stratification. In addition, given the wide range of age in our sample, which offered a great opportunity for us to examine similarities and differences in pathways of risk across developmental periods, we created an age group variable to classify participants into three groups based at their age at T2: emerging adulthood (ages 18–29, n = 452), young adulthood (ages 30–44, n = 1,367), and middle adulthood (ages 45–65, n = 1,017). Our grouping of age groups is consistent with studies in the epidemiology literature (e.g., Grant et al., 2015).
