This systematic review and meta-analysis assessed the isolated effects of physical activity interventions on depressive and anxiety symptoms across the perinatal period. The primary objective was to determine whether standalone physical activity programs delivered during pregnancy (antenatal) or after birth (postpartum) reduce symptoms of depression or anxiety compared with usual care or no intervention.
The review was registered in PROSPERO (CRD42022301284) and conducted in accordance with the Cochrane Handbook and PRISMA guidance. Eligible studies were randomized controlled trials (RCTs) that evaluated standalone physical activity interventions of any type, frequency, intensity, or duration. Trials were eligible only if the comparison group received usual care or no intervention without additional active components.
The authors distinguished between prevention-oriented trials (populations unselected for elevated symptoms) and symptomatic/treatment-oriented trials (participants with elevated depressive or anxiety symptoms at baseline), and synthesized these groups separately. Standard meta-analytic methods were applied, and heterogeneity was reported using I² statistics. Certainty of evidence for pooled outcomes was assessed and reported (e.g., low, very low).
From an initial pool of 10,938 records identified, 17 randomized trials met the inclusion criteria and were included in the quantitative synthesis where appropriate. The included trials varied in intervention types, timing (antenatal or postpartum), frequency, duration, and participant baseline symptom status; these variations contributed to clinical and methodological heterogeneity across pooled estimates.
In prevention-oriented trials pooled for antenatal depressive symptoms, physical activity interventions were associated with lower depressive symptom scores compared with usual care or no intervention. The pooled standardized mean difference (SMD) was -0.52 (95% CI -0.75 to -0.30) across seven trials, with moderate heterogeneity reported (I² = 46%). The certainty of evidence for this pooled estimate was judged to be low.
For prevention-oriented trials examining postpartum depressive symptoms, pooled results also favored physical activity, with an SMD of -0.33 (95% CI -0.57 to -0.09) across seven trials. However, heterogeneity was substantial (I² = 86%), and the certainty of evidence for the postpartum pooled estimate was rated very low. The high heterogeneity reduces confidence in the consistency and generalizability of the observed effect.
When antenatal anxiety outcomes were pooled from prevention-oriented trials, physical activity did not demonstrate a clear benefit compared with no intervention. The pooled SMD was -0.03 (95% CI -0.31 to 0.25) across three trials, with negligible heterogeneity (I² = 0%). The certainty of evidence for antenatal anxiety outcomes was assessed as very low, indicating insufficient evidence to draw firm conclusions.
A separate exploratory analysis focused on symptomatic or treatment-oriented trials. In this subgroup, physical activity interventions were associated with lower postpartum depressive symptom scores compared with no intervention. The pooled SMD was -0.38 (95% CI -0.72 to -0.04) across three trials with low heterogeneity (I² = 28%), but the certainty of evidence for this finding was very low. The authors framed this synthesis as exploratory given the limited number of trials and evidence quality.
Exploratory secondary analyses examined intervention characteristics. These analyses suggested that structured physical activity programs and those delivered approximately three times per week for about 60 minutes per session might be associated with more favorable effects on depressive symptoms. However, these observations were derived from subgroup or exploratory syntheses and the overall certainty of evidence for such program-level inferences was low or very low.
Across pooled outcomes the certainty of evidence ranged from low to very low. Key limitations included clinical and methodological heterogeneity between trials (varying intervention types, timing, delivery, participant characteristics) and limited numbers of trials for some outcomes. Where heterogeneity was substantial (for example, postpartum depressive symptoms in prevention trials, I² = 86%), the precision and generalizability of pooled estimates were reduced. For several outcomes—most notably antenatal anxiety—the available evidence was described as insufficient and inconclusive.
The meta-analysis indicates that isolated physical activity interventions during the antenatal and postpartum periods may be associated with reductions in depressive symptom scores, particularly in prevention-oriented samples. An exploratory synthesis of symptomatic/treatment-oriented trials also suggested lower postpartum depressive symptoms with physical activity. However, these findings are tempered by low to very low certainty of evidence and substantial clinical and methodological heterogeneity across included RCTs.
Evidence for an effect on antenatal anxiety symptoms was not supported by pooled data and remains inconclusive. The authors note that program characteristics such as structured formats and sessions around three times weekly for ~60 minutes showed potential benefits in exploratory analyses, but the low certainty precludes definitive recommendations.
Clinicians and researchers should interpret these pooled results cautiously. Further high-quality, well-powered RCTs with consistent intervention definitions and standardized outcome measurement are needed to clarify the magnitude, timing, and clinical relevance of benefits of physical activity for perinatal depressive and anxiety symptoms.