Center-of-Mass Work Patterns Reveal a Dissociation Between Gait Organization and Limb-level Mechanical Function in Post-stroke Walking
Hosseini-Yazdi, S.-S.; Fitzsimons, K.; Bertram, J. E.
Show abstract
Walking speed is widely used to assess gait recovery following stroke, yet it provides limited insight into how walking performance is mechanically organized. This study examined how center-of-mass (COM) work organization and propulsion-support coupling vary across walking speeds in individuals with post-stroke hemiparesis to distinguish recovery of gait organization from recovery of limb-level mechanical function. Eleven individuals with post-stroke hemiparesis performed treadmill walking across speeds ranging from 0.2 to 0.7 m/s while ground-reaction forces were recorded. Limb-specific COM power and work were computed using an individual-limbs framework, and interlimb asymmetry in net and positive work, along with the propulsion-support ratio (PSR), were quantified. A qualitative transition in gait organization was observed: at lower walking speeds, COM power exhibited a simplified two-phase pattern, whereas at higher walking speeds (approximately [≥]0.5 m/s), a structured four-phase COM power pattern emerged, including identifiable push-off and preload phases. Despite this recovery of gait organization, interlimb work asymmetry remained elevated and paretic PSR remained reduced across all speeds, indicating persistent limb-level mechanical deficits. These findings demonstrate that increases in walking speed and the emergence of typical COM power structure reflect recovery of gait organization rather than restoration of underlying limb-level mechanical capacity. Consequently, walking speed alone is insufficient to characterize gait recovery after stroke, and biomechanically informed measures of COM work organization and propulsion-support coupling provide complementary insight by distinguishing organizational recovery from limb-level mechanical recovery.
Matching journals
The top 7 journals account for 50% of the predicted probability mass.