Ankle exoskeleton assistance increases task-relevant variability without altering center of mass control during walking

Background Ankle exoskeletons alter joint kinematics during walking, yet their effects on whole-body coordination remain unclear. This study investigated how ankle exoskeleton assistance influences the structure of motor variability and center-of-mass movement during steady-state walking. Methods Twenty healthy adults walked on a treadmill at 1.1 m/s without the exoskeleton (noExo) and with active exoskeleton assistance (Exo). Whole-body kinematics were recorded. Variability was analyzed using the Uncontrolled Manifold (UCM) approach. Joint angle variability was decomposed into components that do not affect center-of-mass position (UCM || ) and components that do (UCM ⊥ ). Their ratio (UCM ratio ), reflecting the synergy stabilizing the center-of-mass, was calculated. Center-of-mass position was analyzed separately in three dimensions. Time-continuous differences across the gait cycle were evaluated using statistical parametric mapping. Results UCM || and UCM ratio showed no significant differences between conditions. UCM ⊥ was higher with exoskeleton assistance over large portions of the gait cycle (0–80%, p = 0.001; 91–100%, p = 0.022). Center-of-mass movement in the mediolateral and anteroposterior directions did not differ between conditions, while small differences were observed in the vertical direction (higher in Exo at 23–31% and 75–84%, p = 0.037; lower at 48–54%, p = 0.042). Conclusion Ankle exoskeleton assistance increased joint-level variability affecting center-of-mass movement, but overall center-of-mass control was preserved, with only small changes in vertical center-of-mass displacement. Healthy adults maintained whole-body coordination despite altered mechanical conditions introduced by the exoskeleton. These findings are relevant for clinical exoskeleton use, where assistive devices should support walking without compromising center-of-mass control and balance.