Step width variability, but not step length variability or step time variability, discriminates gait of healthy young and older adults during treadmill locomotion

The variability of spatial and temporal step kinematics have separately been shown to prospectively predict falls by older adults. However, the published literature has not addressed the relative importance of the information related to locomotion control contributed by variability of spatial variables, temporal variables, or both. We conducted a post hoc analysis to determine the extent to which the variability of spatial and temporal step kinematics are independent descriptors of locomotion control in healthy young and older adults. A second purpose of the analysis was to establish the extent to which treadmill walking mimics overground walking using a benchmark for step kinematics. Eighteen young adults and 12 older adults walked at a self-selected velocity on an instrumented treadmill from which step length variability, step width variability and step time variability were computed. Stepwise discriminant analysis correctly classified group membership of 70 percent of the 30 subjects (i.e., young or older adult) using only step width variability (p=0.037, Wilkʹs LAMDA=0.854). Step width variability was 70±57 percent larger than step length variability (p<0.001). This relationship was similar to that of the benchmark established for overground locomotion. The results suggest that for healthy younger and older adults step width variability is a more meaningful descriptor of locomotion control than step length variability and step time variability. Given the potential clinical impact, further systematic study and improvement of the methods and technology for acquiring step kinematic data are warranted.