Comparison of gait variability and stability indices

Indices of variability and stability are essential in evaluating the motions of the human gait. Owing to the inconsistent results of earlier studies, the relationships among popular indices, i.e., maximum Lyapunov exponent (λ), MeanSD, and long-range correlation (α), are inconclusive. In this study, we examined the correlations among these indices using the velocity of the center of body mass for 10 participants walking on a treadmill. In terms of the velocity along the gravitational direction, MeanSD exhibited a negative correlation with the short-term λ _S and the long-term λ _L when the walking speed was 4.0 km/h. Although a positive correlation between MeanSD and λ is expected based on the meanings of these indices, the present and earlier studies collectively suggest that there is no positive correlation between MeanSD and λ. In the same direction, a positive correlation was found between λ _S and α when the walking speed was 3.5 km/h. MeanSD and α tended to exhibit a negative correlation for some conditions, which is reasonable because MeanSD and α indicate the level of variability and the long-term consistency of walking, respectively.