Defining and characterizing differences in college alcohol intervention efficacy: A growth mixture modeling application.

OBJECTIVE While college alcohol misuse remains a pervasive issue, individual-level interventions are among the most efficacious methodologies to reduce alcohol-related harms. Growth mixture modeling (GMM) was used as an exploratory moderation analysis to determine how many types of college drinkers exist with regard to intervention efficacy over a 12-month period. METHOD Data from 3 randomized controlled trials were combined to yield a sample of 1,040 volunteer and mandated college students who were given 1 of 3 interventions: a brief motivational intervention, Alcohol Edu for Sanctions, or Alcohol 101 Plus. Participants were assessed at baseline, and 1, 6, and 12 months postintervention. RESULTS Through the examination of heavy drinking behaviors, piecewise GMM identified 6 subpopulations of drinkers. Most of the sample (76%) was lighter drinkers who demonstrated a strong intervention response, but returned to baseline behaviors over the subsequent 12 months. In contrast, 11% of the sample reported no significant change over the 12-month period. Four minority subpopulations were also identified. In sum, 82% of the sample responded to intervention, but 84% of the sample reported intervention decay over the subsequent 12 months. Female gender, being an, upperclassmen, beginning drinking later in life, not engaging in drinking games, and lower norms predicted a greater likelihood of responding to intervention. CONCLUSION Individual-level interventions are successful at effecting change in most college students, but these effects tend to decay to baseline behaviors by 12 months. These results suggest intervention efforts need to find ways to engage freshmen men and those who play drinking games. (PsycINFO Database Record