ABSTRACT
PURPOSE: Adolescence is associated with the onset of depressive symptoms as well as significant increases in health-risk behaviors. Potential explanations for the direction of effects in the association between depressive symptoms and health-risk behaviors include the self-medication/acting out hypothesis (i.e., early depressive symptoms predict increases in risk behaviors over time) and the failure hypothesis (i.e., early participation in health-risk behaviors predicts increases in depressive symptoms over time). The purpose of the present longitudinal study was to assess these competing hypotheses across the high school years, and to examine whether the direction of effects (and therefore the self-medication/acting out and failure hypotheses) may differ depending on the type of risk behavior under consideration. METHODS: The sample consisted of 4,412 adolescents (49% female) who were followed up from grade nine to 12. Adolescents reported on their depressive symptoms and six health-risk behaviors (frequency of alcohol use, amount of alcohol consumed per drinking episode, cigarette smoking, marijuana use, hard drug use, and delinquency). Analyses were conducted with dual trajectory growth curve modeling. RESULTS: Adolescents who had higher depressive symptoms in grade nine reported faster increases than their peers in smoking, marijuana, and hard drug use across the high school years, supporting the self-medication hypothesis. The failure hypothesis was not supported. CONCLUSION: The results are important because they suggest that by targeting depressive symptoms during early adolescence, treatment programs may prevent increases in the frequency of these risk behaviors later in adolescence.
