Simulated gait training improves joint loading symmetry in unilateral transfemoral bone-anchored limb users

Bone-anchored limbs (BALs) are promising alternatives for addressing socket-related pathologies, but evidence suggests movement and joint loading asymmetries persist 12-months following transfemoral BAL implantation. Gait training targeting biomechanical symmetry is common in rehabilitation for people with transfemoral amputation. However, it is unknown if people with transfemoral BALs can achieve symmetrical gait due to absent anatomical structures. Our objective was to determine if gait symmetry is feasible in transfemoral BAL users and investigate the biomechanical effects of enforcing symmetry using optimal control and musculoskeletal modeling. Using OpenSim Moco, we simulated gait training interventions for 11 transfemoral BAL users by predicting a target gait pattern (assuming intact musculoskeletal systems) that were then tracked using subject-specific BAL models. Movement patterns after simulated training were evaluated to assess changes in symmetry using magnitude and phase difference metrics and Pearson correlation coefficients. Joint loading was compared to baseline values temporally (using statistical parametric mapping) and discretely (joint reaction force (JRF) impulses, stance time, and respective symmetries). Trained solutions were more in-phase and highly correlated with non-amputated kinematics and across limbs, demonstrating that the clinical objective was achievable. Amputated limb resultant hip JRF increased in loading response and terminal stance (p = 0.020 and p = 0.004, respectively), resulting in improved symmetry during terminal stance (p = 0.003) following simulated training. Joint loading impulses were also increased in the amputated limb (p < 0.001), with little change in the intact limb (p = 0.337), resulting in improved joint loading-symmetry (p = 0.001). These results demonstrate movement symmetry is achievable without deleterious joint loading effects in bilateral limbs.