Family interventions and their effect on adolescent alcohol use in general populations; a meta-analysis of randomized controlled trials.

AIMS In order to quantify the effectiveness of family interventions in reducing adolescent drinking, we conducted a meta-analysis of randomized controlled trials. METHODS We searched the Cochrane Database of Systematic Reviews, ERIC (Educational Research Information Center), Medline and PsycInfo for studies published between 1995 and September 2006. Summary estimates (OR and Cohen's d) were derived from the difference in changed alcohol consumption between family intervention and control group. Random effect models were used to estimate the overall effect and heterogeneity among studies. Eighteen papers describing nine independent trials were eligible for inclusion in this meta-analysis. RESULTS The overall effect of family interventions in reducing alcohol initiation (OR: 0.71; 95% CI: 0.54, 0.94) and frequency of alcohol use (d: -0.25; 95% CI: -0.37, -0.12) show the success of these programs. There was heterogeneity between studies reporting on alcohol initiation (p-heterogeneity: <0.001; I2: 78.6%). Yet, the most successful interventions continued to be effective in reducing alcohol initiation even at 48 months follow-up (pooled estimate (OR): 0.53; 95% CI: 0.38, 0.75). CONCLUSION The results from this meta-analysis suggest that the overall effect of family interventions on adolescent alcohol use is small, yet consistent and effective even at 48 months.