The Effect of Walking Speed on Gait Variability in Healthy Young, Middle-aged and Elderly Individuals

Previous research with healthy young adults has suggested that the temporal structure of gait variability is not random but shows self-similarity that is dependent on speed. Specifically, the strength of the long-range correlation of stride intervals follows a quadratic relationship with the minimum values at the respective preferred walking speed (PWS). The purpose of this study was to investigate if this relationship is affected by increasing age. Ten healthy young, seven healthy, middle-aged and seven healthy, elderly adults completed five-minute walking trials at 80%, 90%, 100%, 110% and 120% of their PWS on a treadmill. We investigated the temporal structure of gait variability by using detrended fluctuation analysis. In addition, we computed the Coefficient of Variation (CV) to identify effects on amount of gait variability. Our results revealed a significant quadratic relationship between the temporal structure of gait variability and speed for all groups extending the previously reported existence of such a relationship in healthy young adults to older individuals. However, only significant negative linear relationships were found between amount of variability and speed providing support that this relationship is not quadratic but linear across individuals of different ages. In addition, we found that the examination of the temporal structure of gait variability is more sensitive in differentiating middle-age and younger individuals. If middle-age is where the aging process starts, then measures of the temporal structure of gait variability are essential as prognostic and diagnostic tools of aging.