Teasing apart a multiple component approach to adolescent alcohol prevention: what worked in Project Northland?

This paper presents the results of a post hoc component analysis designed to tease apart the effects of different intervention strategies used in Project Northland, a group-randomized, community-wide, multi-level intervention trial originally conducted in the 1990's to prevent and reduce alcohol use among a cohort of mainly White students in rural Minnesota. This study focuses on Phase I, when students were in 6th-8th grade. The intervention during this phase included five components: classroom curricula, peer leadership, youth-driven/led extra-curricular activities, parent involvement programs, and community activism. Student exposure to/participation in these components was followed over time using reliable process measures. These measures were used as time-varying covariates in growth curve analyses to estimate the effects of the intervention components over time. Multi-item scales from annually-administered student surveys were used to measure relevant outcome variables, like alcohol use. The impact of the components appears to have been differential. The strongest effects were documented for the planners of extra-curricular activities and parent program components. The classroom curricula proved moderately effective, but no effects were associated with differential levels of community activism. The interactions tested here did not provide support for synergistic effects between selected intervention components. Care must be taken when selecting and combining intervention strategies meant to reduce adolescent alcohol use.