Effect of gait retraining strategies on clinical and biomechanical outcomes in subjects with knee osteoarthritis: a systematic review with meta-analysis and GRADE recommendations.

Abstract

To assess the effects of gait retraining strategies on clinical and/or biomechanical variables in knee osteoarthritis. This study was registered in PROSPERO-CRD42023402778. Studies were searched up to June/2024. PEDro scale and GRADE approach were used to assess methodological quality and certainty of evidence. A meta-analysis was conducted using R-software. The included studies employed backward walking (5), toe-out (1), and freely strategies (2). The meta-analysis revealed a significant difference with low quality of evidence favoring gait retraining for pain (MD = -1.12; 95% CI: [-1.61; -0.62]; I² = 78%; <i>p</i> < 0.01) and with high quality for function (SMD = -0.80; 95% CI: [-1.12; -0.48]; I² = 0%; <i>p</i> = 0.42). A significant difference also favoring gait retraining was observed for first knee adduction moment (KAM) (MD = -0.20; 95% CI: [-0.38; -0.02]; I² = 84%; <i>p</i> < 0.01). The comparisons indicated similarity between groups for second KAM (MD = -0.16; 95% CI: [-0.47; 0.16]; I² = 79%; <i>p</i> = 0.03) and for knee flexion moment (MD = 0.07; 95% CI: [-0.07; 0.20]; I² = 76%; <i>p</i> = 0.02). Gait retraining improved pain, function and first KAM. Second KAM and knee flexion moment did not modify.