Abstract
OBJECTIVE: Gait symmetry is an important assessment indicator in patients with stroke and is classified into temporal and kinematic types. This study aimed to clarify the clinical assessments and ability characteristics of subtypes by gait symmetry through clustering analysis using symmetry indices. DESIGN: Retrospective, cross-sectional, observational study. PARTICIPANTS: Fifty-nine patients with chronic stroke who could walk independently without aids during measurement, including those who habitually used an assistive device. METHODS: Gait was analysed with a 3-dimensional motion analysis system and force plates. Temporal symmetry was assessed using the swing-time ratio; kinematic symmetry was assessed by the normalized cross-correlation of left-right knee angle waveforms; k-means clustering used the 2 indices; and clinical measures and comfortable gait speed were compared across subtypes. RESULTS: Four subtypes were identified. Subtypes with high kinematic but low temporal symmetry had moderate motor paralysis, moderate gait speed, and high paretic-side propulsion, whereas subtypes with high temporal but low kinematic symmetry showed moderate motor paralysis, higher muscle tone, moderate gait speed, and lower paretic-side propulsion. CONCLUSION: This study provides new insights into gait symmetry in stroke and may help lay the groundwork for future studies on gait classification and rehabilitation.