Exercise for perinatal depressive symptoms: A systematic review and meta-analysis of randomized controlled trials in perinatal health services

Background: Exercise improves perinatal depressive (PD) symptoms, but reports call for more robust evidence. This systematic review and meta-analysis aimed at synthesizing evidence exclusively from randomized controlled trials (RCTs) examining the effects of exercise on PD symptoms in women recruited through perinatal health services. Methods: Nine e-databases and fifteen systematic reviews were searched for relevant RCTs. Exercise-specific tools extracted/coded data. A meta-analysis using a random effects model (Standardized Mean Difference [SMD]) investigated the effects of exercise on PD scores post-intervention. Results: From 285 records, 14 RCTs (2.025 participants) were considered eligible including two RCTs with clinically diagnosed PD women. Exercise showed a statistically significant, small, overall antidepressant effect (SMD = -0.21, 95% CI = -0.31, -0.11, p = 0.0001) with low/non-significant heterogeneity (Q = 17.82, I-2 = 16%, p = 0.27). Only the fail-safe criterion recorded marginally significant publication bias, but trim-fill analysis added no study. Sensitivity analyses increased the overall effect in RCTs showing lower risk of bias or delivering =150 min/week moderate intensity aerobic exercise. Subgroup analyses revealed significant antidepressant effects for exercise across various settings, delivery formats, depressive symptoms severities and outcome measures used. Heterogeneity was low/non-significant in all analyses (I-2 = 50%). Hedges' g corrections did not influence the results. Limitations: Study limitations include the small number of available trials and clinically diagnosed PD samples and the variety of exercise modalities. Conclusions: Exercise improved PD symptoms, especially in RCTs with lower risk of bias or with >= 150 min/day moderate intensity aerobic exercise interventions. Findings are clinically useful but more RCTs for clinically diagnosed PD women are needed for firmer conclusions.