Differences in Marijuana Use Between White and American Indian Youth Living on or Near Reservations

Introduction. American Indian (AI) adolescent substance use rates for alcohol, marijuana, and other drugs are consistently higher than national rates.  For AI youth living on or near reservations, marijuana use rates are startlingly higher than national rates reported by Monitoring the Future.  For example, near daily use (defined as 20+ times per month) for 8^th grade reservation-based AI youth is nearly 7 times higher than the rate reported by MTF.  Very few studies have compared AI and non-AI youth who reside in the same areas - on or near reservations - and who are subject to many of the same socioeconomic risk factors. In this paper, we examine frequent marijuana use by individual and school ethnicity for AI and white youth who live on or near reservations.

Methods. This study uses data from 4,040 7^th – 12^th grade students attending 33 geographically diverse schools located on or near reservations between 2009 to 2012 (29% white; 71% AI).   Students were classified as frequent users of marijuana based on whether they reported using marijuana 10 times or more in the past month.  Multilevel logistic analysis was used to estimate 3 models, each of which builds on the previous model by adding in variables (control variables, perceived availability, and norms). The percentage of AI youth in the school was included at level 2 to measure contextual effects of ethnicity.

Results. When only control variables are included, students attending schools with a higher percentage of AI youth are more likely to be frequent marijuana users, no matter their individual ethnicity (OR_scheth=7.0).  Individual ethnicity is not significant in predicting frequent use. Adding perceived availability to the model (positively associated with frequent use) results in a decrease of OR_scheth to 5.6 while individual ethnicity is still non-significant. In the final model, parent and student injunctive norms are included. Both variables are strongly related to frequent use, with odds ratios of .41 and .67 respectively.  At the same time, the relationship between school ethnicity and frequent use becomes non-significant while the relationship between individual ethnicity and use becomes significant, with AI odds of frequent use being twice those of a comparable white youth.

Conclusions. These results suggest that the contextual effects observed in the first model may be due, in part, to parent and student injunctive norms.  Indeed, in examining whether norms vary by individual ethnicity and school ethnicity shows that for both AI and white youth, there is a strong relationship between the percentage of AI youth in the school and student injunctive norms.  This is not the case for parent norms, where only individual ethnicity is significantly associated with norms.