Characterizing accuracy, retention, and dose-response of an 18-session real-time gait biofeedback program following anterior cruciate ligament reconstruction.

BACKGROUND: Individuals following anterior cruciate ligament reconstruction (ACLR) commonly demonstrate aberrant walking biomechanics, characterized by a lesser first peak vertical ground reaction force (vGRF) during early stance. Gait biofeedback may address this deficit; however, the accuracy, retention, and dose-response characteristics of repeated training remain unclear. METHODS: Thirteen individuals (8 females and 5 males; age 22.4 ± 3.4 years; 32.6 ± 18.7 months post-ACLR) completed 18 gait biofeedback sessions over 6 weeks. Participants completed rehabilitation and exhibited involved limb first peak vGRF ≤1.12 body weight (BW) during baseline assessment of treadmill walking. Bilateral, step-by-step visual biofeedback targeted a first peak vGRF of 1.15 BW. Feedback was continuous during sessions 1-10 and was partially tapered thereafter. Outcomes included first peak vGRF target accuracy, defined as the percentage of steps within ±1% of the 1.15 BW target, first peak vGRF during retention, and the estimated training dose required to achieve target loading. FINDINGS: First peak vGRF target accuracy increased in both limbs from ∼13% during session 1 and plateaued by session 12 (20-25%). Retention testing of the involved limb showed significantly higher first peak vGRF than baseline beginning at session 3, with these improvements maintained through session 18. The median effective dose to achieve target loading was 5 sessions (95% confidence interval: 3-14 sessions). INTERPRETATION: Bilateral first peak vGRF target accuracy improved across training, while involved-limb improvements were retained before completing the 18-session protocol. These findings support the feasibility of shorter gait biofeedback regimens and provide clinically relevant dosing parameters to inform future clinical trials. CLINICALTRIALS: gov: NCT05848622.