Predictors of Responsible Drinking or Abstinence Among College Students Who Binge Drink: A Multitheory Model Approach

Context: Binge drinking is a salient problem on college campuses, with estimates as high as 40% of students engaging in it. Binge drinking is associated with numerous negative consequences among college students, such as suicide attempts, unsafe sex practices, property damage, and driving under the influence. Several behavioral approaches in this regard have had modest impact and only short-term effects, however. Objective: To use the multitheory model (MTM) of health behavior change to predict initiation and sustenance of responsible drinking or abstinence among binge-drinking college students in a sample drawn from a large southern public university. Methods: This cross-sectional survey study included a sample of college students who binge drank in the past 30 days. A 39-item face- and content-valid instrument was used. In addition, construct validity using confirmatory factor analysis and internal consistency reliability using the Cronbach a were established. Hierarchical regression modeling was used to build models. Results: A total of 289 students participated. The Cronbach alpha for the scale and all subscales ranged from 0.81 to 0.94 and demonstrated acceptable internal consistency reliability. Construct validity using confirmatory factor analysis yielded 1-factor solutions for each of the subscales. On hierarchical regression modeling, gender (P= .05). race/ethnicity (P=.004), behavioral confidence (P=.029), and changes in physical environment (P=.001) were associated with intended initiation for drinking responsibly/abstinence behavior change. The addition of MTM constructs led to a significant increase in R-2 of 0.20 (F-3,F-194 =18.1; P<.001). The sustenance constructs yielded a significant increase in R-2 of 0.20 (F-3,F-193 =19.4; P<.001). Conclusion: This study provides empirical justification for MTM constructs that can be used to inculcate the intention to drink responsibly or abstain among college students who binge drink. This predictive model may prove valuable in the design of interventions aiming to improve responsible drinking behavior in this population.