Is behavioral activation an effective and acceptable treatment for co-occurring depression and substance use disorders? A meta-analysis of randomized controlled trials

Background Depression often co-occurs with substance use problems and is associated with poor treatment outcomes. While the efficacy of behavioral activation (BA) has been tested in clinical trials with substance users, outcomes have not yet been quantitatively synthesized.

Methods The study team performed a random effects meta-analysis of the randomized clinical trial evidence base. We compared outcomes for individual or group BA against passive or active controls. We also compared attendance and dropout rates. The meta-analysis used a grading of recommendations assessment, development, and evaluation (GRADE) approach to assess the quality of each meta-analytic comparison.

Results We included five trials in the meta-analysis (N = 195). The analysis found no significant differences between BA and controls with regard to depression (Post-treatment: k = 5; N = 195; SMD: 0.19, CI –0.10 to 0.49; p = 0.20; GRADE = Low; Follow-up: k = 5; N = 195; SMD: –0.10, CI –0.51 to −0.30; p = 0.62; GRADE = Low) or substance use (post-treatment: k = 4; N = 151; SMD: 0.14, CI –0.33 to −0.6; p = 0.57, GRADE = Low; Follow-up: k = 4; N = 151; SMD: 0.17, CI –0.34 to 0.69; p = 0.51, GRADE = Low) and there was little evidence of publication bias. The average session attendance rate for BA was 72%. An average dropout rate of 35% was reported for both BA and comparator conditions.

Conclusion BA does not emerge as a differentially efficacious treatment for comorbid depression and substance use disorders, although it does appear to be an acceptable treatment option. Our confidence in the results are limited by the number and quality of the original studies and the possibility of the effect of small study bias. We make suggestions for improving the methodological quality and direction of future BA trials.