Relationship between the executive function on walking motor imagery and actual walking ability in stroke patients: A...

Post-stroke walking impairment reflects both motor and higher-order cognitive deficits. We examined whether executive function mediates the association between walking-related motor imagery accuracy and actual walking performance and kinematics in patients with subacute hemiparetic stroke. Thirty inpatients underwent mental chronometry, involving timed imagined and actual maximum 10-m walks, and the discrepancy between the two was expressed as the Index of Walking Performance (IWP), with values near zero indicating high accuracy. Participants were categorized into overestimation and underestimation groups. Executive function was assessed with the Trail Making Test Part B (TMT-B). Walking outcomes included maximum 10-m walking time, the Timed Up and Go test (TUGT), and paretic peak knee flexion during stance and swing phases. Mediation analyses with TMT-B as the mediator were conducted both without covariates and after adjustment for age and position sense. Higher IWP values were associated with longer TMT-B times, slower maximum 10-m walking and TUGT, greater paretic stance-phase, and smaller swing-phase knee flexion. The overestimation group showed slower maximum 10-m walking and smaller swing-phase knee flexion than the underestimation group. Unadjusted models showed full mediation for the TUGT and stance-phase knee flexion, partial mediation for swing-phase knee flexion, and no mediation for the maximum 10-m walking time. After adjustment, these indirect effects were attenuated, and their confidence intervals included zero, whereas associations of IWP and TMT-B with the walking outcomes remained significant. Executive function may therefore contribute to the association between walking motor imagery accuracy and actual walking performance, although its mediating role remains preliminary.