[Purpose] The aim of this study was to determine if gait index predicts the efficacy of
weight-support treadmill training (BWSTT) in hemiplegic stroke patients. [Participants and
Methods] In total, 21 patients who had sustained a hemiplegic stroke, on an average 71
days prior, and could walk independently on level ground were included in the study. BWSTT
was performed under 20% of bodyweight unloading at the maximum speed possible for each
participant to a perceived level of fatigue of 15 on the 20-point Borg scale. The
immediate effects of BWSTT were evaluated as the change in the following variables,
calculated from 5 level ground gait cycles; walking speed and rate, root mean square,
coefficient of variability, auto-correlation coefficient, and single leg stance time
ratio. All indices were calculated from the triaxial accelerometer attached to the waist
of the participant. Linear regression was used to identify predictive variables of BWSTT
effectiveness. [Results] Only single leg stance time ratio on level ground was extracted
as a predictor of BWSTT effectiveness. [Conclusion] Single leg stance time ratio was a
predictive factor of improved gait symmetry after BWSTT and therefore, could be used as a
factor to select patients who might benefit from BWSTT as a component of stroke
rehabilitation.