Repeatability and variation of quantitative gait data in subgroups of patients with stroke.

We aimed to determine the repeatability and variation of quantitative gait data in patients with stroke and to compare the subgroups in terms of gait variability. Time-distance and kinematic characteristics of gait were evaluated in 90 inpatients (30 women) with hemiparesis (mean+/-S.D. age 57.7+/-12.5 years and time since stroke 5.99+/-6.46 months). Subgroups were based on "gender", "side of paresis", "lesion type", "motor recovery level", "sensory status", "time since stroke" and "walking velocity". Repeatability was adequate to excellent in all stroke subgroups (ICC range 0.48-0.98). Walking velocity was the most repeatable gait parameter after stroke. Variation in step length was significantly higher in women than in men (CV 16% versus 9%, p<0.05). Slow walkers (walking velocity <0.34 m/s) had a higher variation than fast walkers in step length (CV 12.5% versus 7.5%, p<0.01), single support time (CV 11.9% versus 6.3%, p<0.05), peak hip extensions in stance (CV 11.5% versus 3.7%, p<0.01) and knee flexion in swing (CV 11.8% versus 6.5%, p<0.05). In our stroke patients, their age, time since injury, lesion characteristics, impaired proprioception or level of motor recovery had no effect on gait variability. For better interpretation of quantitative gait data, clinicians should consider that variation in step length, single support time, peak hip extension in stance and knee flexion in swing differs according to walking velocity after stroke.