Profiles of confidence and commitment to change as predictors of moderated drinking: a person-centered approach

Identifying who, among problem drinkers, is best suited for moderation and has the greatest likelihood to control drinking has important public health implications. The current study aimed to identify profiles of problem drinkers who may be more or less successful in moderating drinking within the context of a randomized clinical trial of a brief treatment for alcohol use disorder. A person-centered approach was implemented, utilizing composite, baseline daily diary values of confidence and commitment to reduce drinking. Problem drinkers (N = 89) were assessed, provided feedback about their drinking, and randomly assigned to 1 of 3 conditions: 2 brief alcohol use disorder treatments or a third group asked to change on their own. Global self-report assessments were administered at baseline and Week 8 (end of treatment). Daily diary composites were created from data collected via an interactive voice recording system during the week prior to baseline. A K-means cluster analysis identified 3 groups: high, moderate, and low confidence and commitment to change drinking. Group differences were explored, and then group membership was entered into generalized estimating equations to predict drinking trajectories over time. Findings revealed that the groups differentially reduced their drinking, such that the high group had greater reduction in drinking and a faster rate of reduction than the other 2 groups, and the moderate group had greater reduction than the low group. Findings suggest that baseline motivation and self-efficacy are important for predicting prognoses related to successful moderated drinking. Limitations and arenas for future research are discussed.