Abstract
BACKGROUND: Evidence on the effects of Pilates exercise in patients with knee osteoarthritis (KOA) remains limited. This meta-analysis aimed to evaluate its efficacy and safety. METHODS: We searched PubMed, Web of Science, EMBASE, Cochrane Library, and Google Scholar up to May 21, 2024, for randomized controlled trials (RCTs). Outcomes included Visual Analogue Scale (VAS), Western Ontario and McMaster Universities Osteoarthritis Index (WOMAC), knee range of motion (ROM), and adverse events. Risk of bias and evidence certainty were assessed; subgroup, sensitivity, and publication bias analyses were performed. RESULTS: Eight RCTs involving 322 participants were included. Risk of bias ranged from low to high, mainly due to lack of blinding. Compared to blank controls, Pilates significantly reduced WOMAC scores (SMD = -1.70; 95% CI: -3.14 to -0.25). One study comparing Pilates to health education reported significant reductions in VAS (MD = -1.74; 95% CI: -2.51 to -0.97) and WOMAC (SMD = -1.42; 95% CI: -2.39 to -0.45), but no significant ROM improvement. Compared to other exercises, Pilates showed no significant effects on VAS, WOMAC, or ROM. Funnel plots showed asymmetry for VAS and WOMAC; Egger's test indicated possible publication bias (p = 0.0039 and 0.0154, respectively). No adverse events were reported, but active monitoring was limited. GRADE rated evidence as "very low" for VAS and WOMAC, and "low" for ROM. CONCLUSION: Pilates may relieve pain and improve physical function, but has limited effects on ROM. Due to insufficient adverse event reporting, safety remains unclear. Given the low study quality, high heterogeneity, and possible bias, results should be interpreted cautiously. Future studies should use standardized protocols, report long-term adherence, and assess mental health benefits.