Substance use during adolescence poses serious developmental risks, particularly among underserved populations such as American Indian (AI) youth. This study uses multilevel modeling to examine the relationship between school-level factors and the likelihood of AI youth substance use, accounting for student-level variation. Self-report surveys were collected from 8th, 10th, and 12th-grade students during the 2004 spring semester. Analyses focused on a subsample of 2,188 self-identified AI youth, ages 11 to 19. Results showed that 60% of youth reported ever drinking alcohol, 50.7% ever using marijuana, and 22.9% ever using illicit drugs. Results from the multilevel model revealed that school clustering explained some of the variation in the likelihood of alcohol and drug use for AI students; however, school-level factors (urban/rural, school size, school SES, teacher-student ratio) had little to no effect on substance use. Academic failure was associated with higher odds for lifetime and 30-day alcohol and marijuana use, and lifetime illicit drug use. Low school commitment was associated with higher odds for lifetime and 30-day alcohol and illicit drug use, and 30-day marijuana use. Additionally, students' sex, age, depressive symptoms, and drug accessibility were also significantly associated with increased risk of substance use. These findings suggest that while school structural characteristics exert limited direct influence on AI youth substance use, student-level experiences within schools, including academic failure, low commitment, and depressive symptoms, are critical targets for prevention, highlighting the value of multilevel approaches for understanding substance use etiology in AI communities.
Herrera et al. (Mon,) studied this question.