Journal of Contextual Behavioral Science (JCBS)
Volume 28, April 2023, Pages 149-168
Authors
Shuanghu Fang, Dongyan Ding
Abstract
The extant meta-analyses that specifically focus on the differences between ACT and tCBT were published more than ten years ago. They had some methodological shortcomings (i.e., for the same outcome variable, only one primary effect size was included) and were not broad enough to consider factors such as moderating variables, follow-up changes, and dropout rates. Thus, it could not draw accurate and specific conclusions. Learning about the possible differences between tCBT and ACT can identify the more efficient intervention for individuals based on their characteristics and the moderators that may impact the intervention's efficacy and the dropout rate. In this way, treatment effectiveness can be maximized with limited resources. It can also facilitate a better comprehension of the two therapies and promote their further evolution or integration toward process-based therapy. This study aims to provide an updated three-level meta-analysis of comparison between the efficacy of ACT and tCBT. The three-level meta-analysis is able to effectively deal with the dependence between effect sizes by extracting all relevant effect sizes from each study, and thus can retain all the information and obtain the maximum statistical power.
The current study included 38 studies with 2536 individuals. It was found that tCBT outperformed ACT on anxiety with a small effect both in the short term [g = 0.22, 95%CI (0.01, 0.23), p < 0.01; Q (df = 40) = 49.59, p > 0.05; Level 2 = 0, Level 3 = 26.2%] and long term [g = 0.16, 95%CI (0.03, 0.29), p < 0.05; Q (df = 29) = 25.96, p > 0.05; Level 2 = 0,
Level 3 = 30.97%]. However, ACT outperformed tCBT on mindfulness in the short term with near medium effect [g = 0.48 (95%CI: 0.11 to 0.85; < 0.05); Q (df = 5) = 10.28, p > 0.05; = 54.34%].
Population characteristics could be a moderating variable for the differences between ACT and tCBT. In the long term, for improving depression, tCBT may be more effective for individuals with physical symptoms. In the short term, for improving psychological flexibility, ACT may be more effective for no diagnoses reported participants or participants with behavioral health conditions, while tCBT may be more effective for individuals with physical symptoms. Besides, in the long term, for increasing psychological flexibility, ACT may outperform tCBT for no diagnoses reported participants. However, the moderating effect should be interpreted with caution due to the small number of RCTs included and the limited statistical power. Recommendations for future research, practical implications, and limitations are discussed.