These analyses of data from 1028 adolescents and young adults in the COGA protocol tested four hypotheses regarding how Baseline (Time 1) SRE5-LR and impulsivity related to the number of DSM-IV alcohol problems (ALC PROBS) in the 2 years before Time 3, 4 years after Baseline. A SEM evaluated both direct and indirect pathways from LR and impulsivity to the Time 3 outcome, with a focus on potential mediation by Time 2 latent variables reflecting perceived peer drinking (PEER), expectations of the effects of alcohol (EXPECT), and drinking to cope with stress (COPE), while controlling for sex, age, ethnic background, family histories of AUDs. Several findings reported here related to moderate to large effect sizes, indicating they might offer clinically relevant guidance for future efforts to prevent alcohol problems, as discussed below.
