Sampling rate influences the regularity analysis of temporal domain measures of walking more than spatial domain measures

Background: The spatiotemporal dynamics of stepping can provide useful information about walking performance. Most often, the identification of gait motion is performed using 3-D cinematography. The sampling rate of motion capture systems may influence the accuracy of these measures albeit in varying degrees for measures within the spatial versus temporal domain. Research question: What are the effects of sampling frequency on common analysis methods of measures within the spatial and temporal domain? Methods: Specifically, mean, variability (i.e. standard deviation), and regularity (i.e. sample entropy) of step length (i.e. spatial domain) and step time (i.e. temporal domain) measures were assessed following ten minutes of preferred-speed treadmill walking in eleven young adults. Results: The spatiotemporal mean measures were not affected by changing sampling frequencies. Frequencies >= 120 Hz showed consistent results for spatial variability measures, while temporal variability increased due to decreased resolution in capturing variability when data was sampled at 120 Hz or less. In assessing regularity, poor temporal resolution at lower sampling rates led to "binning", limiting the variety of vector patterns. As a result, more vectors were classified as similar, leading to a signal appearing more periodic. For the spatial domain, sample entropy was not affected, indicating the greater sensitivity of step time to sampling rate compared to step length. Significance: Sampling rate influenced recognition of gait events. By reducing the sampling rate, the time intervals were increased and reduced the resolution leading to less accurate gait event detection in the temporal domain. The sampling rate of 120 Hz is the minimum sampling rate that should be used to calculate spatiotemporal data for variability and sample entropy.