This meta-analysis investigates the efficacy of exercise as a treatment for DSM-IV diagnosed anxiety disorders.

We searched PubMED and PsycINFO for randomized, controlled trials comparing the anxiolytic effects of aerobic exercise to other treatment conditions for DSM-IV defined anxiety disorders. Seven trials were included in the final analysis, totaling 407 subjects. The control conditions included non-aerobic exercise, waitlist/placebo, cognitive-behavioral therapy, psychoeducation and meditation. A fixed-effects model was used to calculate the standardized mean difference of change in anxiety rating scale scores of aerobic exercise compared to control conditions. Subgroup analyses were performed to examine the effects of (1) comparison condition; (2) whether comparison condition controlled for time spent exercising and (3) diagnostic indication.

Aerobic exercise demonstrated no significant effect for the treatment of anxiety disorders (SMD�=�0.02 (95%CI: −�0.20–0.24), z�=�0.2, p�=�0.85). There was significant heterogeneity between trials (χ² test for heterogeneity�=�22.7, df�=�6, p�=�0.001). The reported effect size of aerobic exercise was highly influenced by the type of control condition. Trials utilizing waitlist/placebo controls and trials that did not control for exercise time reported large effects of aerobic exercise while other trials report no effect of aerobic exercise.

Current evidence does not support the use of aerobic exercise as an effective treatment for anxiety disorders as compared to the control conditions. This remains true when controlling for length of exercise sessions and type of anxiety disorder. Future studies evaluating the efficacy of aerobic exercise should employ larger sample sizes and utilize comparison interventions that control for exercise time.