Exercise-based interventions for post-stroke depression: an overview of systematic reviews and evidence mapping

Exercise-based interventions are increasingly used to address post-stroke depression, but existing systematic reviews differ in population definitions, intervention types, outcome measures, and methodological rigor. This overview synthesized the clinical evidence, appraised review quality and reporting, reassessed certainty, and examined primary-study overlap.

Systematic reviews with or without conventional pairwise meta-analysis were eligible when they evaluated exercise-based interventions for post-stroke depression or depressive symptoms after stroke. Network meta-analyses and reviews without separately extractable exercise evidence were excluded. Methodological quality, reporting completeness, and risk of bias were assessed using AMSTAR 2, PRISMA 2020, and ROBIS, respectively. Certainty was reassessed using GRADE. Primary-study overlap was quantified using the corrected covered area and explored through sensitivity analyses.

Six systematic reviews were included. Two were rated low and four critically low using AMSTAR 2; ROBIS identified one review at low risk, three at unclear risk, and two at high risk. The main overlap analysis included 67 trial occurrences across 50 confirmed unique trials and indicated moderate overlap (CCA 6.80%). Most reviews reported favorable effects of exercise on depressive symptoms, although effect magnitude, heterogeneity, and certainty varied. Evidence for secondary clinical outcomes and safety was limited.

Exercise may reduce depressive symptoms after stroke, but it remains unclear whether this finding applies specifically to patients with formally diagnosed post-stroke depression. The optimal exercise type and dose and whether benefits persist over time are also uncertain. Confidence in the evidence is limited by methodological weaknesses, heterogeneity, incomplete safety reporting, and primary-study overlap.

https://www.crd.york.ac.uk/PROSPERO/view/CRD420261448674, CRD420261448674.