Substance Use Prevention for Urban American Indian Youth: A Efficacy Trial of the Culturally Adapted Living in 2 Worlds Program.

Abstract

This article describes a small efficacy trial of the Living in 2 Worlds (L2W) substance use prevention curriculum, a culturally adapted version of keepin' it REAL (kiR) redesigned for urban American Indian (AI) middle school students. Focused on strengthening resiliency and AI cultural engagement, L2W teaches drug resistance skills, decision making, and culturally grounded prevention messages. Using cluster random assignment, the research team randomized three urban middle schools with enrichment classes for AI students. AI teachers of these classes delivered the L2W curriculum in two schools; the remaining school implemented kiR, unadapted, and became the comparison group. AI students (N=107) completed a pretest questionnaire before they received the manualized curriculum lessons, and a posttest (85% completion) 1 month after the final lesson. We assessed the adapted L2W intervention, compared to kiR, with paired t tests, baseline adjusted general linear models, and effect size estimates (Cohen's d). Differences between the L2W and kiR groups reached statistically significant thresholds for four outcomes. Youth receiving L2W, compared to kiR, reported less growth in cigarette use from pretest to posttest, less frequent use of the Leave drug resistance strategy, and less loss of connections to AI spirituality and cultural traditions. For other substance use behaviors and antecedents, the direction of the non-significant effects in small sample tests was toward more positive outcomes in L2W and small to medium effect sizes. Results suggest that evidence-based substance use prevention programs that are culturally adapted for urban AI adolescents, like L2W, can be a foundation for prevention approaches to help delay initiation and slow increases in substance use. In addition to study limitations, we discuss implementation challenges in delivering school-based interventions for urban AI populations.