This study aimed to explore and characterize the relationships between anterior-posterior GRFs and clinically relevant gait indicators, including those based on CoM and CoP.
Seventy-eight individuals with hemiparesis following stroke (age: 64.8 ± 12.2 years) performed at least three paretic stance trials (except for two trials in one individual) at their comfortable speed. The location of the CoM and paretic CoP in the horizontal plane was computed using three-dimensional motion capture and two force plates during the stance phase. Values were positive when CoP and CoM are located anteriorly or medially to the heel or other reference point, or when the CoP was located anteriorly or medially to the CoM. Clinically meaningful gait metrics, including anteroposterior and mediolateral positions at maximum, minimum, and key events such as foot-contact and foot-off, were computed. After excluding metrics with a correlation coefficient over 0.8, 10 unique metrics were retained. The relationships between these metrics and peak or mean braking, propulsive, and late braking forces were analyzed using Spearman correlation coefficients.
Seven individuals who did not generate propulsive force were excluded from the late braking force analysis. Peak and mean braking forces showed the highest correlation with ‘maximal anterior distance of CoP relative to CoM’ (rs = -0.75, -0.77). Peak and mean propulsive forces had the strongest correlation with ‘maximal posterior distance of CoP relative to CoM’ (rs = -0.91, -0.91). Additionally, the ‘mediolateral position of CoP at foot-off’, the 'change in lateral distance between CoP and CoM', and ‘loading duration on a hindfoot’ were also significant. Peak and mean late braking forces strongly correlated with ‘CoP and CoM move closer following maximal posterior shift of the CoP’ (rs = 0.79, 0.78) and extended ‘anterior position of CoP at foot-off’ (rs = -0.58, -0.60), which were identified as new metrics.
Anterior and posterior distances between CoM and CoP were key metrics related to braking and propulsive forces. These metrics were also associated with lateral control parameters, including regional impairments in hemiparesis. Late braking forces were characterized by CoP and CoM moving closer, or a forward CoP shift during late stance, suggesting clinical features such as backward CoM displacement or toe dragging.
Clarifying the relationship between multiple new simplified gait metrics reflecting gait strategies and GRF control could provide useful insights for gait rehabilitation.
ground reaction force
center of mass
