Active feedback gait retraining alters foot pressure patterns and reduces knee adduction moment in an anterior cruciate ligament reconstructed population

Study Design Intervention study Background Anterior cruciate ligament (ACL) tear greatly accelerates osteoarthritis (OA) onset, and ACL reconstruction (ACLR) surgery has not been shown to prevent premature knee OA. It was recently shown that ACLR subjects with higher knee adduction moments (KAM) 2 years after ACLR had worse patient-reported outcomes at 8 years after surgery, and patients with higher relative contribution of KAM to the total loading at the knee joint had greater decreases in medial-to-lateral femoral cartilage thickness ratios over time. As such, interventions such as gait retraining to reduce the KAM in ACLR patients present an opportunity to reduce the risk of developing premature knee OA. Purpose The purpose of this study was to test the efficacy of a gait retraining method using active feedback to change the KAM by sensing lateral foot pressure and retraining to produce a medial weight transfer at the foot. Specifically, we hypothesized (1) there would be significant changes in foot pressure patterns (average heel pressure ratio) and joint loading (first peak KAM, KAM Impulse) with active feedback, (2) joint loading change is influenced by pressure threshold feedback setting (Level 1 vs. Level 2), and (3) joint loading is correlated with pressure distribution at the foot during walking. Methods Ten individuals with unilateral primary ACLR surgery (7M/3F; age: 37.7 ± 11.1 yrs; BMI: 25.1 ± 4.3 kg/m2, 42.8 ± 29.0 months post-ACLR) were tested at normal speed in three conditions: control, active feedback (“Level 1”), and active feedback at a higher setting (“Level 2”). For the active feedback trials, if pressure measured under the lateral side of the heel exceeded a subject-specific threshold, a vibration motor placed in contact with the skin at the ankle was activated. Subjects were instructed to walk in a manner to successfully keep the device from vibrating, by shifting weight medially at the foot, which determined Level 1 settings. Level 2 required a greater medial shift by using a lower lateral heel pressure threshold. For control trials the feedback system was inactive. The first peak KAM (KAM1) and KAM Impulse were determined using a motion capture system and force plate. Medial-tolateral (M/L) heel pressure ratio was determined from OpenGo insoles (Moticon). Differences in outcome variables were analyzed using repeated measures analysis of variance with post hoc paired Student’s tests. The relationship between KAM variables and heel pressure ratio was analyzed using linear regression analyses. P-values < 0.05 were considered significant. Results KAM1 was significantly reduced at Level 1 (-29.5%, P = 0.002) and Level 2 (-36.3%, P < 0.001) feedback settings as compared to control. The KAM Impulse was also significantly reduced at Level 1 (-49.4%, P = 0.003) and Level 2 (-57.3%, P = 0.001) feedback settings. The reductions at the Level 2 setting were significantly greater than at the Level 1 setting (KAM1: P = 0.017; KAM Impulse: P = 0.020). The average M/L pressure ratio was significantly increased at both Level 1 (70.9%, P < 0.001) and Level 2 (64.7%, P < 0.001) feedback settings, with no difference between settings (P = 0.21). The reductions in KAM1 and KAM Impulse increased with greater pressure transfer from the lateral to medial heel. Specifically, reductions in KAM1 (Figure 1a) and KAM Impulse (Figure 1b)were significantly correlated with the changes in M/L pressure ratios when considering both Level 1 and Level 2 pressure setting data. However, the patient response to M/L pressure change was variable, with individual subject R2 ranging from 0.01 to 0.95 for KAM1 and R2 ranging from 0.10 to 0.95 for KAM Impulse. Conclusion ACLR subjects responded to an active feedback gait retraining that produces a medial weight-bearing shift at the foot in a manner that reduces the first peak KAM and KAM Impulse, with greater changes at the Level 2 feedback setting. The results that reductions were associated with increases in the M/L foot pressure ratio suggest a potential dose response for this intervention. The variations in patient dose response suggest that future applications of this method should be further customized to the patients. Finally, considering the long term risk of developing premature medial knee OA in this ACLR population, reducing KAM may be beneficial and these results suggests the need for future research to determine the long term effect of the active feedback intervention on joint loading as well as clinical outcomes.