Latent Class Analysis and Predictors of Marijuana Use among Reservation-based American Indian High School Students.

American Indian (AI) youth residing on reservations report higher rates of marijuana use compared to national youth. Latent class analysis (LCA) was used to identify unique types of marijuana use among 2,884 AI high school students surveyed from 26 schools across six indigenous geographic regions. Predictors of class membership were then assessed using social, cultural, and individual measures relevant to adolescent substance use. Classes and predictors were examined separately for males and females. Four-class models fit the data best for both male and female AI students. Classes differed by sex, as did predictors. Overall, social predictors related to family and peers and the individual predictor, using marijuana to cope, were the best predictors of class membership. Based on these results, prevention and intervention efforts should provide alternative coping methods for these adolescents who often live in difficult situations, and should focus on encouraging parents to effectively monitor their adolescent children and communicate clear sanctions against marijuana use.