Peer Mechanisms in the Internalizing Pathway to Alcohol Abuse

Introduction: This study examines adolescent developmental pathways to alcohol abuse within a social network context. The internalizing pathway into alcohol use posits a developmental process that results in the use of alcohol as a form of coping with ongoing emotional distress; some studies show that this tendency is first evident in adolescence. Notably, few studies consider this dynamic relative to the peer context. Peer relationships are central to healthy adolescent development and strongly implicated in both substance use and emotional distress. Social network research suggests that adolescents select friends with similar drinking behaviors, and are also influenced to become more similar to their friends with respect to alcohol use. Here we examine how these selection and influence processes are moderated by the experience of internalizing problems.

Methods: We estimate Stochastic Actor-Oriented (R-Siena) Models using longitudinal social network data from over 6,700 adolescents from six high schools over seven waves of data that span 4.5 years. Siena models social networks dynamically over time, allowing for the separate parameterization of peer influence and peer selection effects. We build on the basic influence vs. selection framework by testing interactions between peer behaviors and internalizing problems, specifically, we hypothesize that depressed youth are more likely to befriend alters who drink, and that depression renders adolescents more susceptible to peer influence processes.

Results: We find that adolescents tend to befriend peers who are similar to them with respect drinking, however, we also fined that adolescents who drink are more likely to be chosen as friends. Depression does not significantly moderate this selection process. We also find evidence of peer influence: adolescents whose friends drink are more likely to become drinkers themselves. In some schools, popularity was also positively associated with drinking. Depression is, in some schools, directly associated with increased drinking. However, there is no evidence that depression increases susceptibility to peer influence processes.

Conclusion: We conclude that peer influence is an important factor in the diffusion of alcohol abuse, and that depression, while it is itself a risk factor for drinking, does not significantly alter this process. A second key conclusion is that popularity appears, in some schools, to be reciprocally linked to drinking. Future prevention efforts should more fully consider direct peer influence and more distal influences through the linkage between drinking and popularity in school.