Older persons decrease gait speed at fast and dual task walking but do not increase gait variability over a one year period

BACKGROUND AND AIM: Both gait speed and variability of gait measures are associated with falls [1, 2]. It has been shown that increased stride‐to‐stride variability in speed, step length and double support is independently associated with falls in a prospective study of older adults [3]. The aim of this study was to assess changes in gait speed and gait variability in older persons over a 1‐year period. METHODS: One hundred and sixty two healthy persons >70 years of age drawn from the National registry, mean age 78.6 ± 4.7 years, walked over an electronic gait mat (GaitRite) at two sessions one year apart. Subjects were instructed to walk at 3 different speeds (slow, preferred, fast) and at preferred speed while counting backwards from 50. One hundred and thirty three subjects completed both sessions. Dependent measures were walking speed and variability in Step Length (SD), Step Width (SD), Step Time (CV), and Single Support Time (CV). Paired t‐tests were used to assess change from baseline to 1 year follow‐up. Change in variability was assessed at a common gait reference speed for all participants. RESULTS: After one year, fast walking speed was significantly reduced (p 1.1 m/s). In the slow walking group, no significant changes in gait speed occurred in any of the walking conditions. The medium walkers increased slow walking speed while fast walkers decreased preferred, fast, and dual task walking speed. Variability measures for the entire group did not change from baseline to follow‐up (Fig 1 B & C) while for sub groups, fast walkers only increased CV Single Support Time. There were no changes for other measures of variability for any of the groups. CONCLUSIONS: The increased slow walking speed but decreased fast walking speed suggest differential effects between simpler and more complex walking tasks. The larger change in gait speed for the fastest walker suggests start of age‐related changes in previously fit persons. Lack of changes in variability may be explained by the analysis methods used (normalized speed) or that gait speed changes in demanding gait situations occur before changes in gait variability.