Do Brief Alcohol Interventions Also Reduce Tobacco Use Among Youth: A Systematic Review and Meta-Analysis

Introduction: Despite an array of alcohol and tobacco use prevention efforts, youth worldwide still initiate and use these substances. Brief interventions (BIs), i.e., interventions delivered in less than five hours of total contact time, have been used in many treatment and prevention settings and may be one resource-efficient tool in prevention efforts. Given the association and similarities between alcohol and tobacco use, a generalization of intervention effects may occur for these two risky health behaviors.

Objectives:We conducted a systematic search for studies of BIs that target alcohol consumption and report alcohol and tobacco use outcomes among adolescents and young adults. Our goal was to assess whether such interventions reduce alcohol and tobacco use for participants among interventions that targeted alcohol consumption and either (1) directly targeted tobacco use or (2) did not target tobacco use but measured it as a secondary outcome.

Methods:Our search involved multiple electronic databases and included grey literature sources (1980-2012). There were no geographic or language restrictions. Eligible studies were coded for key variables and analyses were conducted using random effects inverse-variance weighted three-level models.

Results: We identified 18 eligible randomized or controlled quasi-experimental studies representing 5949 participants. Studies were conducted in Brazil, England, Switzerland, and the United States. Tobacco use was directly targeted in 12 of the studies and was measured as a secondary outcome in 6. Most intervention participants were male (61%) and 43% were white. Eleven studies enrolled adolescents and seven enrolled young adults. BIs were associated with significant reductions in alcohol consumption outcomes only: relative to control groups, intervention participants significantly reduced their alcohol consumption (ḡ = 0.11, 95% CI [0.04, 0.17]) but not their tobacco use (ḡ= 0.07, 95% CI [-0.01, 0.16]). Whether the intervention directly addressed tobacco or simply measured it as a secondary outcome was not a significant moderator affecting tobacco use outcomes. Post-hoc exploratory analysis illustrated there may be differences between younger and older participants.  

Conclusions: Although brief interventions are applicable in a wide variety of settings, they may not produce secondary effects beyond the primary targeted outcomes for risk behaviors such as alcohol and tobacco use. Future research should further explore the mechanisms by which they are effective, considering different pathways for diverse risk behaviors and populations.