Drinking, cannabis use and mental health as predictors of within-person changes in aggression over eight weeks among youth involved with binge drinking.

INTRODUCTION: Sequential, multiple assignment, randomized trials (SMARTs) are used to deliver behavior interventions to adolescents and emerging adults (A/EAs) involved with alcohol use and aggression. We examine with-in person markers of aggression from a recent SMART to inform future interventions for this population. METHODS: A/EAs (n = 400; ages 14-20; 22 % male) in the emergency department who reported binge drinking and aggression were enrolled in the SMART, completed a baseline survey, received a brief behavioral health intervention (SafERteens), were randomized to booster conditions over 8 weeks, and completed 8 weekly surveys to inform intervention components. We examined same week aggression occurrence in relation to weekly risk (i.e., alcohol consumption, any cannabis use, mental health symptoms, acting impulsively) and protective factors (i.e., excitement, hopefulness, fun activities, social support). Unadjusted and adjusted conditional logistic regressions examine within-person effects of weekly factors on an individual's risk of aggression. RESULTS: In the unadjusted model weekly factors associated with increased likelihood of aggression in the same week included earlier survey week and any cannabis use; more alcohol consumption, mental health symptoms, and impulsive acts; and, less hopefulness and social support. In the adjusted model, earlier survey week, cannabis use, and more alcohol use and mental health symptoms were associated with increased likelihood of same week aggression. CONCLUSIONS: Given these associations, findings suggest that aggression prevention interventions may benefit from focusing on substance use and mental health symptoms, while enhancing social support. Findings extend prior work by identifying temporally proximate within-person associations with aggression to inform adaptive interventions.