The moderating effects of group cognitive-behavioral therapy for depression among substance users

Abstract

This study examined the prospective longitudinal relationship between changes in depressive symptoms on alcohol and/or drug (i.e., substance) use among addiction participants in treatment, and whether group cognitive-behavioral therapy for depression (GCBT-D) moderated the relationship. Using a quasi-experimental intent-to-treat design, 299 residential addiction treatment clients with depressive symptoms (Beck Depression Inventory-II, BDI-II scores > 17; Beck, Steer, & Brown, 1996) were assigned to either usual care (n = 159) or usual care plus a 16-session GCBT-D intervention (n = 140). Two follow-up interviews were conducted, one 3 months after the baseline interview corresponding to the end of the intervention, and then one 3 months later. Parallel-process growth modeling was used to examine changes in depressive symptoms and the associated changes in abstinence and negative consequences from substance use over time. Treatment group was included as a moderator of the association. Participants in the GCBT-D condition showed a greater increase in abstinence and greater decreases in depressive symptoms and negative consequences over time. There were significant interaction effects, such that the associations between depressive symptoms, negative consequences, and abstinence changes were larger in the usual-care condition than in the GCBT-D condition. The results suggest that the intervention may be effective by attenuating the association between depressive symptoms and substance use outcomes. These findings contribute to the emerging literature on the prospective longitudinal associations between depressive symptoms and substance use changes by being the first to examine them among a sample receiving GCBT-D in an addiction treatment setting.

Trial registration: ClinicalTrials.gov NCT01191788.

2013 APA, all rights reserved

Figures

Figure 1

Parallel process moderation model (residual variances not shown in figure, but were estimated in the model). Circles indicate latent variables and squares indicate observed variables. PDA = percent days abstinent; BDI-II = BDI-II inventory scores; Treatment = dummy coded comparison of Usual Care Control and GCBT-D groups.

Figure 2

Association between the change in Beck Depression Inventory – II scores and the change in percent days abstinent from baseline to 6-months following the baseline assessment by treatment groups. Dashed lines indicate 95% confidence interval.