Walking stride interval variability in patients with diabetic Charcot foot: A pilot study

Abstract

Introduction: The variability of human walking has been studied for about twenty years. It appears not to be random but long-term autocorrelated, i.e. on time scales corresponding to several hundreds of gait cycles. It has been shown that the central nervous system (CNS) plays an important role in this variability. Indeed, people with CNS alterations have a different autocorrelation pattern than healthy individuals. In this pilot study, we address the question of the peripheral nervous system (PNS) influence on this long-term variability through the assessment of gait in patients with diabetic Charcot foot. Methods: The population included 4 patients with diabetic Charcot foot and a control group of 6 healthy individuals. Ages were matched in both groups. Patients walked on a treadmill and the accelerations of their right ankle during the walk were recorded using a homemade inertial sensor. The acceleration versus time was used to calculate the durations of successive gait cycles. These stride intervals time series were analyzed by calculating several parameters related to variability: coefficient of variation (CV), Hurst exponent and sample entropy. Results: Gait speed and average stride interval were significantly lower and higher in the Charcot group respectively. The coefficient of variation was significantly higher in the Charcot group. Hurst exponent and sample entropy were also greater in the Charcot group, though non significantly. Conclusion: Patients with diabetic Charcot foot have a slower walk. Their long-term stride interval variability has a larger amplitude (larger CV) and predictibility (larger Hurst exponent) than healthy individuals. It is tempting to relate these features to an increased risk of fall.