Comparing spatiotemporal gait parameters between overground walking and self-paced treadmill walking in persons after stroke

Purpose Since self-paced treadmills enable more natural gait patterns compared to fixed-speed treadmills we examined the use of a self-paced treadmill as a alternative for overground gait analysis in persons after stroke. Material and methods Twenty-five persons after stroke (10 males/15 females; 53 +/- 12.05 years; 40.72 +/- 42.94 months post-stroke) walked at self-selected speed overground (GAITRite, CIR Systems) and on a self-paced treadmill (GRAIL, Motek) in randomized order. Spatiotemporal parameters, variability and symmetry measures were compared using paired-sample t-tests or Wilcoxon Signed Rank tests. Concurrent validity was assessed using intraclass correlation coefficients and Bland-Altman plots. A regression model determined the contribution of the walking velocity to the changes in spatiotemporal parameters. Results The velocity on the treadmill was significant lower compared to overground (p < 0.001). This difference predicted the significant changes in other spatiotemporal parameters to varying degrees (27.7%-83.8%). Bland-Altman plots showed large percentage of bias and limits of agreement. Variability and symmetry measures were similar between conditions. Conclusions When considering gait analysis in persons after stroke a self-paced treadmill may be a valuable alternative for overground analysis. Although a slower walking velocity, and accompanying changes in other spatiotemporal parameters, should be taken into account compared to overground walking.