Study links marijuana use and distress with college binge drinking

This abstract-based summary describes an observational machine-learning study asking which factors predict binge drinking among 31,120 college students in the 2021–2022 Healthy Minds Study. Researchers used random forests, partial-dependence plots, and mixed-effects logistic regression to examine marijuana use alongside social, psychological, and demographic characteristics. The study did not test a cannabis intervention or establish that marijuana use causes binge drinking.

Marijuana use, older age, lower religiosity, depression, and a weaker sense of belonging were among the strongest predictors of binge drinking. Emotional distress and diminished belonging were associated with higher predicted risk, while loneliness showed the opposite pattern; younger age and higher religiosity were associated with lower predicted risk. The abstract reports associations and prediction patterns, not treatment effects or evidence that changing any factor would prevent binge drinking. The most important limitation is that this abstract cannot establish causation, and it does not provide effect sizes or enough detail to judge the practical importance of each predictor.