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

Seewald L.; Goldstick J.; Carter P.; Ngo Q.; Bourque C.; Pearson C.; Battisti K.; Walton M. · Addictive behaviors · 2026

**Background & Methods** This Sequential, Multiple Assignment, Randomized Trial (SMART) enrolled 400 adolescents and emerging adults (ages 14-20; 22% male) presenting to an emergency department with concurrent binge drinking and aggression. Participants received the SafERteens brief behavioral health intervention at baseline, were randomized to booster conditions, and completed 8 weekly surveys over 8 weeks. Within-person associations between weekly risk/protective factors and same-week aggression occurrence were examined using conditional logistic regression. **Key Findings** • Cannabis use (any use in a given week) was independently associated with increased likelihood of aggression in the same week, after adjusting for other factors. • Increased alcohol consumption and elevated mental health symptoms in a given week were significantly associated with higher odds of aggression that same week in the adjusted model. • Protective factors including hopefulness and social support were associated with reduced aggression risk in unadjusted analyses, though the adjusted model emphasized the importance of addressing substance use and mental health as primary intervention targets. **Dosage & Administration** Not reported. The study evaluated the SafERteens intervention as a brief behavioral health intervention with subsequent booster conditions, but specific dosage parameters were not detailed in the abstract. **Safety & Adverse Effects** Not reported. **Evidence Quality** This SMART design represents a methodologically rigorous approach to identifying temporal relationships between modifiable risk factors and aggression outcomes. However, the study has several limitations: the sample was predominantly female (78%), which may limit generalizability to male youth; the brief abstract does not provide sufficient detail regarding intervention fidelity, booster content, or effect sizes; and the within-person design, while strong for identifying temporal associations, cannot establish causality. The weekly assessment window may miss important behavioral patterns occurring between surveys. The findings are appropriate for generating hypotheses for adaptive interventions but should be supplemented with mechanistic investigations before implementation. This represents level II evidence suitable for informing intervention development targeting co-occurring substance use and aggression in youth.

More recent cannabis research