The Effects of Resistance Exercise Training on Anxiety: A Meta-Analysis and Meta-Regression Analysis of Randomized Controlled Trials

Brett R. Gordon, Cillian P. McDowell, Mark Lyons, Matthew P. Herring

The salutary effects of resistance exercise training (RET) are well established, including increased strength and function; however, less is known regarding the effects of RET on mental health outcomes. Aerobic exercise has well-documented positive effects on anxiety, but a quantitative synthesis of RET effects on anxiety is needed. To estimate the population effect size for resistance exercise training (RET) effects on anxiety and to determine whether variables of logical, theoretical, and/or prior empirical relation to anxiety moderate the overall effect. Thirty-one effects were derived from 16 articles published before February 2017, located using Google Scholar, MEDLINE, PsycINFO, PubMed, and Web of Science. Trials involved 922 participants (mean age = 43 ± 21 years, 68% female/32% male) and included both randomization to RET (n = 486) or a non-active control condition (n = 436), and a validated anxiety outcome measured at baseline, mid-, and/or post-intervention. Hedges' d effect sizes were computed and random effects models were used for all analyses. Meta-regression quantified the extent to which participant and trial characteristics moderated the mean effect. RET significantly reduced anxiety symptoms (Δ = 0.31, 95% CI 0.17-0.44; z = 4.43; p < 0.001). Significant heterogeneity was not indicated (Q T(30) = 40.5, p > 0.09; I (2) = 28.3%, 95% CI 10.17-42.81); sampling error accounted for 77.7% of observed variance. Larger effects were found among healthy participants (Δ = 0.50, 95% CI 0.22-0.78) compared to participants with a physical or mental illness (Δ = 0.19, 95% CI 0.06-0.31, z = 2.16, p < 0.04). Effect sizes did not significantly vary according to sex (β = -0.31), age (β = -0.10), control condition (β = 0.08), program length (β = 0.07), session duration (β = 0.08), frequency (β = -0.10), intensity (β = -0.18), anxiety recall time frame (β = 0.21), or whether strength significantly improved (β = 0.19) (all p ≥ 0.06). RET significantly improves anxiety symptoms among both healthy participants and participants with a physical or mental illness. Improvements were not moderated by sex, or based on features of RET. Future trials should compare RET to other empirically-supported therapies for anxiety.