LAUSR

BACKGROUND Alcohol use, and the association between alcohol-related variables and drinking outcomes, seem to be context dependent. We employed Latent Class Analysis, a person-centered approach, to identify distinct subpopulations based on contexts of alcohol use. We also examined if the resulting classes differ in a set of alcohol-related variables that hold promise as potential targets of interventions. METHOD Argentinean college students N = 1083; 64 % women; M age = 19.73 ± 3.95) completed an online survey that assessed alcohol outcomes and related variables [motives for substance use, protective behavioral strategies (PBS) and normative perceptions of alcohol use]. RESULTS Latent Class Analysis identified a 4-classes model for drinking contexts. These classes showed differential patterns of alcohol use (both in terms of frequency and quantity) and also differed in alcohol-related variables. Alcohol use was more prevalent in contexts characterized by the presence of peers, particularly the closest friend, small same-sex or coed groups. Close to 50 % of the sample reported drinking with family members. Students with a high probability of engaging in pregaming and nightclub drinking tend to use more PBS to control how much alcohol is consumed than those who do not drink in these contexts. All the classes significantly differed in their social reasons for drinking and in the number of alcohol-related negative consequences. CONCLUSIONS Our findings revealed subpopulations of college students that are heterogeneous regarding contexts of alcohol use, patterns of use and in a number of relevant variables. These distinctive subpopulations require different targeted interventions.