The Family Check-up in Public Middle Schools: The Effect of Selective Engagement On the Progression of Substance Use From Early Adolescence to Early Adulthood

INTRODUCTION

Substance use puts adolescents at risk of long-term health issues. Intensive programs enhancing parents’ practices can decrease substance use in their adolescents, but family-focused strategies must be adapted to the real world, where resources are limited. Strategies are needed that target youth and families most at risk, and that effectively engage them in cost effective interventions. The Family Check-Up (FCU) is a core component in an adaptive intervention strategy designed to reduce risk for substance use in middle school students. We hypothesized that effective engagement in the context of a randomized intervention trial would reduce substance use from early adolescence to adulthood.

METHOD

998 ethnically diverse, urban youths were recruited in Grade 6 in the U.S. Northwest (47.3% female; 42.3% Caucasian), and randomly assigned to intervention or control. The FCU involved an initial interview, an assessment session, and a feedback with the youth and parents. Self-reports of alcohol, tobacco, and marijuana use were collected at ages 12–17, and again at 19. Problematic substance use (symptoms of abuse or dependence) was measure at age 23. Gender, ethnicity, father absence, risk behavior (α = .95), deviant peers (α = .79), and family conflict (α = .81) were used as predictors of engagement and control variables.

ANALYSES

Complier Average Causal Effect (CACE) models were used in Mplus to test whether children of “engager” parents who were randomly assigned to the FCU used less substance than children of comparable “engager” parents in the control group, who did not have access to FCU. Engagers were at higher risk based on measures of deviant peers, father absence and family conflict. CACE provides a multivariate framework for examining the moderating role of engagement (23% in this study), on the effect of a randomized intervention. Latent growth of substance use was modeled (age 12–17), and regressions predicted problematic substance use at age 23.

RESULTS

Participants were efficiently assigned to engager and non-engager groups, with entropy values of .93–.96. Engagers from the control group (no access to treatment) had a steeper increase in alcohol (p<.01), tobacco (p<.001), and marijuana use (p<.01), and more problematic use of alcohol (p<.01), tobacco (p<.01), and marijuana (p<.10) in adulthood than engagers in the treatment group. Non-engagers from the treatment and control groups had low risk profiles and used less substance than engagers.

CONCLUSION

This study shows the efficacy of a family-centered, adaptive intervention implemented in public schools, which yielded an enduring decrease in substance use in high-risk youths. Many high-risk families were effectively engaged in the FCU, and lower risk families selected themselves out of the FCU.