Abstract
AIMS: To determine whether diminished alternative reinforcement (i.e. engagement and enjoyment from substance-free activities) mediated the longitudinal association of conduct problems with substance use in early-mid-adolescence. DESIGN: Structural equation modeling tested whether the association between wave 1 (baseline) conduct problems and wave 3 (24-month follow-up) substance use outcomes was mediated by diminished alternative reinforcement at wave 2 (12-month follow-up). Additional analyses tested whether sex and socio-economic status moderated this association. SETTING: Ten high schools in Los Angeles, CA, USA, 2013-15. PARTICIPANTS: Students (n = 3396, 53.5% female, mean [standard deviation (SD)] age at wave 1 baseline = 14.1 (0.42) years). MEASUREMENTS: Self-reported conduct problems (11-item questionnaire), alternative reinforcement (44-item questionnaire) and use of alcohol, marijuana and combustible cigarettes during the past 6 months (yes/no) and the past 30 days (nine-level ordinal response based on days used in past 30 days). RESULTS: Significant associations of wave 1 conduct problems with wave 3 marijuana use during the past 6 months (β = 0.25) and past 30 days (β = 0.26) were mediated by wave 2 diminished alternative reinforcement (β(indirect effect) : 6 months = 0.013, 30 days = 0.017, Ps < 0.001). Associations of conduct problems with alcohol or combustible cigarette use were not mediated by alternative reinforcement. All associations did not differ by sex and socio-economic status. CONCLUSIONS: Diminished alternative reinforcement may be a modifiable mechanism linking early adolescent conduct problems and subsequent marijuana use that could be targeted in prevention programs to offset the adverse health and social sequelae associated with comorbid conduct problems and marijuana use in early-mid adolescence.