Mechanisms of change for two brief alcohol interventions: Testing theoretical mediators for counter attitudinal advocacy and personalized feedback intervention effects.
Angelo M DiBello, Clayton Neighbors, Melissa R Hatch, Andrew Weinstein, Kate B Carey
Abstract
Open AccessBACKGROUND: Given the importance of identifying mechanisms of action for the development and dissemination of alcohol interventions, this study tests theory-based mechanisms of change for two brief alcohol interventions. METHODS: We conducted a secondary analysis of data from an efficacy trial that compared a novel intervention based on Counter Attitudinal Advocacy (CAA) to an evidence-based intervention using Personalized Normative Feedback (PNF) and an assessment-only control. Participants consisted of 585 heavy-drinking college students who reported experiencing alcohol-related consequences. Hypothesized mediators were linked to the theoretical underpinning of each intervention: perceived descriptive norms (PNF), dissonance (PNF and CAA), attitudes (CAA), and protective behavioral strategies (CAA). Negative binomial multilevel mediation analyses included data from baseline, posttest, and 1-, 3-, and 6-month follow-up assessments. RESULTS: Mediation analyses indicated that, with respect to drinks per week, PNF significantly reduced perceived norms compared to both the control and CAA conditions, which in turn were associated with decreased alcohol consumption. Similarly, CAA significantly reduced dissonance relative to both control and PNF, which was also associated with reduced drinking. Conversely, PNF increased dissonance relative to control, leading to greater alcohol consumption. Parallel patterns emerged for alcohol-related consequences: PNF reduced norms and CAA reduced dissonance, each associated with fewer consequences, whereas PNF increased dissonance contributing to greater alcohol-related consequences. CONCLUSIONS: Overall, these findings demonstrate that PNF and CAA operate through distinct mechanisms, emphasizing the complexity inherent in intervention effects. They further highlight the importance of empirically identifying and examining the processes underlying the efficacy of alcohol-related interventions.