Patterns of alcohol use and response to digital brief interventions in college students: a secondary analysis of a...

Drinking among college students in India is rising, influenced by social and cultural shifts. This study aimed to classify drinking behaviors among underage students using Latent Class Analysis (LCA) and assess the effectiveness of digital screening and brief intervention (DSBI) and brief advice (DSBA) across identified classes.

This was a secondary analysis of a two-stage cluster randomized trial that screened 693 college students (age 18-22 yrs) with the Alcohol Use Disorders Identification Test (AUDIT). LCA was employed to identify latent classes in a main sample (N = 693, AUDIT ≥1) and a subsample of enrolled participants (N = 548, AUDIT scores 8-19). Model fit was assessed using AIC, BIC, and log-likelihood values, with a 3-class solution preferred for its interpretability and fit. A General Linear Model (GLM) analyzed the effectiveness of digital interventions over 3 and 6 months.

Three latent classes emerged from the main sample: "recreational drinkers" (minimal issues), "hidden problem drinkers (limited family concerns)," and "problem drinkers" (higher family concerns). In addition to the "hidden problem drinkers", the subsample comprised "emerging" and "observable" problem drinkers. Significant reductions in AUDIT scores were observed across all classes from baseline to 3 months, which stabilized by 6 months. A significant interaction effect was found between time and latent class [F (4) =14.60, p<.0001], "observable problem drinkers" outperforming the two other classes.

Underage drinking represents three latent classes. Digital interventions can effectively reduce alcohol use across all classes of underage college problem drinkers.