Longitudinal examination of alcohol demand and alcohol‐related reinforcement as predictors of heavy drinking and adverse alcohol consequences in emerging adults

Background and Aims Behavioral economic theory predicts that high alcohol demand and high proportionate alcohol‐related reinforcement are important determinants of risky alcohol use in emerging adults, but the majority of research to date has been cross‐sectional in nature. The present study investigated prospective and dynamic relationships between alcohol demand and proportionate alcohol‐related reinforcement in relation to heavy drinking days and alcohol problems. Design Longitudinal cohort with assessments every 4 months for 20 months. Setting Ontario, Canada. Participants Emerging adults reporting regular heavy episodic drinking ( n = 636, M age = 21.44; 55.8% female). Measurements Heavy drinking days (HDD; Daily Drinking Questionnaire), alcohol problems (Brief Young Adult Alcohol Consequences Questionnaire), alcohol demand (Alcohol Purchase Task) and proportionate alcohol‐related reinforcement (Activity Level Questionnaire). Findings Linear mixed effects models revealed that behavioral economic indicators and alcohol‐related outcomes significantly decreased over the study, consistent with ‘aging out’ of risky alcohol use. Random intercept cross‐lagged panel models revealed significant between‐person relationships, such that higher alcohol demand and alcohol‐related reinforcement were positively associated with HDD and alcohol problems (random intercepts = 0.187–0.534, P s < 0.01). Moreover, alcohol demand indicators (particularly the rate of change in elasticity of the demand curve, as measured by α , and the maximum expenditure, O max ) and proportionate alcohol‐related reinforcement significantly forecasted changes in HDD at all time points (| β s| = 0.063–0.103, P s < 0.05) in cross‐lagged relationships, with bidirectional associations noted for the rate of change in elasticity ( β s = −0.085 to −0.104, P s < 0.01). Proportionate alcohol‐related reinforcement also significantly forecasted changes in alcohol problems at all time points ( β s = 0.072–0.112, P s < 0.01). Conclusions ...