individuals, who may be in earlier stages of substance use and depression onset. We implement time-varying effect modeling (TVEM), an analytic approach that estimates the associations between predictors (e.g., smoking, marijuana use, and HED) and an outcome (e.g., depressive symptoms) as functions of continuous age. In all analyses, we examine whether gender moderates these age-varying associations. Given the increasing recognition of substance use and depression comorbidity and interest in concurrent treatment, this study was designed to elucidate critical age windows in which these associations are strongest and concurrent treatment may be most beneficial.
