The current analyses extended our work to a more socioeconomically heterogeneous and larger sample of adolescents and young adults, individuals who by virtue of their age are at risk for heavy drinking and alcohol problems. We used three stages of data gathering, each of which was separated from the next by approximately two years. Based on the existing literature, the analyses tested four hypotheses: 1) Both a baseline lower LR and higher Barratt Impulsivity scores will predict a greater number of later alcohol problems, even after considering family history of AUDs and age at baseline, sex, and ethnicity; 2) LR and impulsivity will be only weakly related to each other and will not interact in predicting later alcohol problems; 3) LR and impulsivity will each demonstrate both direct and partially mediated relationships to later alcohol problems; and 4) Both SRE5-LR and impulsivity will increment over the other in hierarchical regression analyses such that consideration of the two Baseline endophenotypes better predicts future alcohol problems than either predictor alone.
