Moderation, mediation - or even both? School climate and the association between peer and adolescent alcohol use.

INTRODUCTION Ample studies discuss the enhancing effects of peer drinking on student alcohol use. In addition, there is vast research on school climate impact on student alcohol use. Though these two areas are intertwined for most young adolescents, it is heretofore not completely clear, in what way these characteristics functionally interact and affect drinking behavior. METHODS In a longitudinal study, we analyzed a sample of 2490 German adolescents (Mage=13.32, SD=0.57, range=8-13) from 5th (fall 2010) to 8th (fall 2013) grade. We discerned mediating (class climate) and moderating (school organization variables) functions of school on the association between peer and adolescent alcohol use, and finally combined them in direct effect moderated mediation models for a variety of outcomes (lifetime alcohol use, frequency and amount of drinking, binge drinking), adjusting for possible confounders. RESULTS Class climate mediated a small significant part of the association between peer and adolescent alcohol use (1.8-2.4%), with the exception of lifetime drinking. Student-teacher ratio and percentage of at-risk students significantly moderated the peer-adolescent association, with the latter having an enhancing and the first having a buffering effect. CONCLUSIONS School life serves as an important context of adolescent development and as such, seems to have direct and indirect effects on behavior and health. Future research should pay attention to differentiating effects of school climate and include both forms of operationalization when analyzing school effects on student behavior.