As hypothesized, depressed mood was negatively associated with growth, which replicates the finding by Fleming et al. (2008) for the high school years and is contrary to predictions based on the self-medication hypothesis of a positive association of depressed mood and future alcohol use. Depressed mood is related to poor peer relations or social withdrawal at this age (Capaldi, 1991) and thus may relate to lower levels of associations with deviant peers and exposure to social drinking. However, antisocial behavior/deviant peer association was included in the prediction model, indicating that the effect of depressed mood was not mediated by this construct. It is possible that less socializing with peers overall at adolescence is associated with lower levels of drinking, given the high prevalence of alcohol use among adolescents. Prediction from changes in levels of depressed mood across the high school years did not explain additional variance in growth.
