Complexity of human gait pattern at different ages assessed using multiscale entropy: From development to decline

Multiscale entropy (MSE) has been applied in biomechanics to evaluate gait stability during human gait and was found to be a promising method for evaluating fall risk in elderly and/or pathologic subjects. The hypothesis of this work is that gait complexity is a relevant parameter of gait development during life, decreasing from immature to mature gait and then increasing again during old age. In order to verify this hypothesis, MSE was applied on trunk acceleration data collected during gait of subjects of different ages: toddlers at the onset of walking, pre-scholar and scholar children, adolescents, young adults, adults and elderlies. MSE was estimated by calculating sample entropy (SEN) on raw unfiltered data of L5 acceleration along the three axes, using values of τ ranging from 1 to 6. In addition, other performance parameters (cadence, stride time variability and harmonic ratio) were evaluated. The results followed the hypothesized trend when MSE was applied on the vertical and/or anteroposterior axis of trunk acceleration: an age effect was found and adult SEN values were significantly different from children ones. From young adults to elderlies a slight increase in SEN values was shown although not statistically significant. While performance gait parameters showed adolescent gait similar to the one of adults, SEN highlighted that their gait maturation is not complete yet. In conclusion, present results suggest that the complexity of gait, evaluated on the sagittal plane, can be used as a characterizing parameter of the maturation of gait control.