Social and individual-level predictors of alcohol use initiation and escalation: Replicating and extending tests of differential effects

Objective: Although alcohol use is considered a developmental phenomenon, there is a relative dearth of studies disaggregating predictors of alcohol use initiation versus early escalation of drinking. One perspective that has emerged is that social levels of influence may be relevant for the initiation of drinking, whereas individual levels of influence may be relevant for the early escalation in level of drinking among initiators, which we refer to as the specificity hypothesis. Method: A sample of alcohol-naive youth (n = 944; mean age = 12.16 years, SD = 0.96) was prospectively assessed for 3 years, spanning six waves of data collection. Results: Both social (parental conflict, perceived prevalence of peer drinking) and individual-level (higher sensation seeking) variables uniquely predicted increases in the likelihood of alcohol initiation. Likewise, both social (perceived descriptive norms of peer drinking) and individual-level (lower school grades, higher sensation seeking) variables uniquely predicted escalation in level of drinking among initiators (although only marginally for sensation seeking). Conclusions: Overall, there was little support for the specificity hypothesis. Our findings suggest that to assume that social and individual-level processes differentially predict drinking outcomes may be a false dichotomy. Theoretical work may benefit from drawing from developmental models emphasizing the interplay between individual and environmental factors in the prediction of the early development of drinking. The emergence of drinking behaviors is likely to result from a developmental cascade of interacting variables that make the ontogeny of drinking unlikely to emerge from a single class of variables.