Effectiveness of wearable and non-wearable biofeedback on knee biomechanics during functional tasks in anterior cruciate ligament injury prevention and rehabilitation: a systematic review and meta-analysis

Anterior cruciate ligament (ACL) injuries are common in sports and linked to modifiable biomechanical risk factors, including excessive knee valgus and low flexion angles. Biofeedback interventions (wearable and non-wearable) aim to correct these patterns, yet no comprehensive synthesis has evaluated their effectiveness across the injury prevention and rehabilitation continuum. This systematic review and meta-analysis examined the effects of biofeedback on knee biomechanics during functional tasks in ACL injury prevention and post-reconstruction rehabilitation. Following PRISMA guidelines, PubMed, Scopus, Web of Science, and Embase were searched up to June 2026. Eligible studies employed wearable or non-wearable biofeedback during functional tasks (landing, cutting, squatting, gait) and reported knee kinetic or kinematic outcomes in healthy individuals or those post-ACL reconstruction. Methodological quality was assessed with MINORS, and certainty of evidence with GRADE. Random-effects meta-analysis was performed on a subset of prevention studies. Twenty-four studies (15 prevention, 9 ACL reconstruction) involving 856 participants were included. In the meta-analysis of 6 prevention studies, biofeedback significantly reduced peak knee abduction angle (Hedges’ g = 0.52, 95% CI 0.06 to 0.98, p = 0.03; I² = 0%), with low certainty of evidence. No statistically significant pooled effects were found for peak knee flexion angle (Hedges’g = 1.72, 95% CI -0.09 to 3.54, p = 0.06; I² = 89.6%), peak knee abduction moment (Hedges’g = 0.17, 95% CI -0.19 to 0.53, p = 0.37; I² = 24.8%). The narrative synthesis of ACL reconstruction studies indicated that visual biofeedback can acutely increase knee flexion and alter vertical ground reaction force during gait, but does not consistently modify frontal-plane moments. Overall certainty of evidence ranged from low to very low, downgraded for risk of bias, imprecision, and inconsistency. Biofeedback interventions significantly reduced peak knee abduction angle in ACL injury prevention, but did not significantly alter peak knee flexion angle, hip flexion angle, knee abduction moment. The certainty of evidence was low to very low for all outcomes. Robust randomised trials with injury outcomes are required before clinical recommendations can be made. Level II.