Predictors of the prospective associations between depressive symptoms and problematic substance use in youth

One particularly prevalent psychiatric comorbidity in youth occurs between depression and substance use. Youth with this comorbidity frequently present in the primary-care setting, but little research has examined this comorbidity within this setting. Moreover, few studies have examined longitudinal relations between depressive symptoms and problematic substance use, and it remains necessary to evaluate mechanisms that predict prospective associations. The present study addresses these needs, and examines the prospective relations between depressive symptoms and problem use and tests the effects of covariates in predicting severity and change over time.;Participants in the present study were from a longitudinal investigation examining treatments for youth depression in the primary care setting. Youth aged 13 to 22 years (N=451, 71% female) were assessed over four time points. Data were analyzed using parallel-process latent growth modeling to test prospective associations, and separate growth models tested the predictive effects of co-occurring symptom constellations (anxiety, attention, aggression, and delinquency), demographic factors (age and gender), and family-relationship (perceived criticism and parental support) variables.;Data suggest that higher initial depressive symptoms predicted growth in problematic substance use; initial problematic substance use was not associated with growth in depressive symptoms. Anxiety, gender, and perceived criticism were predictive of initial severity of both processes. Delinquency was only predictive of initial problem use, and interestingly accounted for the relationship between attention and problem use. Important age and gender effects were also found. Change was harder to predict, but certain covariates (anxiety and attention) evidenced indirect effects in predicting problem use change via higher initial depressive symptoms.;Findings suggest that depressive symptoms and problematic use are associated both initially and longitudinally, but effects appear uni-directional. These data highlight that certain co-occurring symptom, demographic, and family-relationship factors likely impact the associations between these two disorder processes. These data contribute to our understanding of comorbidity theory, and suggest the potential theoretical and clinical utility of integrating causal and shared models of comorbidity. Further research is needed to identify mechanisms that impact risk for the development of comorbid disorders, as well as improve assessment of and interventions for youth within the primary-care setting.