Effect sizes of randomized-controlled studies of cognitive behavioral therapy for anxiety disorders over the past 30 years

Abstract

Background

Cognitive-behavioral therapy (CBT) is an effective treatment for the range of anxiety disorders as demonstrated in many randomized controlled trials (RCTs) conducted over the past 30 years.

Objective

To examine the change of CBT effect sizes over time, we tested whether publication year was a significant moderator.

Data sources

We pooled studies from three previously published meta-analyses and searched three electronic databases (PubMed, PsychINFO, Web of Science) to identify any additional studies published from February 1, 2022 to January 24, 2025. The final analysis included 49 studies comprising a total of 3645 participants.

Study selection

We selected RCTs comparing CBT for anxiety disorders with psychological or pill control conditions.

Data extraction

Two independent raters used predefined data fields, including study quality indicators.

Data synthesis

The mean effect size of all RCTs comparing CBT vs. controls revealed a Hedges' g of 0.51, 95 % CI [0.40, 0.62], with significant differences in effect sizes between the diagnostic groups. Uncontrolled pre-post effect size calculations revealed a large effect for CBT, Hedges' g = 1.18, 95 % CI [1.01, 1.34], and a medium effect for the control conditions, Hedges' g = 0.59, 95 % CI [0.47, 0.70]. A linear meta-regression of publication year on effect sizes (Hedges g) showed no significant change in effect sizes as compared to the control conditions over the span of the last 30 years (B = −0.008, SE = 0.006, t(47) = −1.18, p = 0.24).

Conclusion

Although CBT is an effective treatment for anxiety disorders, the effect sizes did not increase over the last 3 decades. This calls for studies on the processes of treatment change to improve the efficacy of CBT.