Variability of gait patterns during unconstrained walking assessed by satellite positioning (GPS)

Terrier, Philippe ; Schutz, Yves

In: European Journal of Applied Physiology, 2003, vol. 90, no. 5-6, p. 554-561

Add to personal list
    Summary
    It is established that the ratio between step length (SL) and step frequency (SF) is constant over a large range of walking speed. However, few data are available about the spontaneous variability of this ratio during unconstrained outdoor walking, in particular over a sufficient number of steps. The purpose of the present study was to assess the inter- and intra-subject variability of spatio-temporal gait characteristics [SL, SF and walk ratio (WR=SL/SF)] while walking at different freely selected speeds. Twelve healthy subjects walked three times along a 100-m athletic track at: (1) a slower than preferred speed, (2) preferred speed and (3) a faster than preferred speed. Two professional GPS receivers providing 3D positions assessed the walking speed and SF with high precision (less than 0.5% error). Intra-subject variability was calculated as the variation among eight consecutive 5-s samples. WR was found to be constant at preferred and fast speeds [0.41 (0.04)m·s and 0.41 (0.05)m·s respectively] but was higher at slow speeds [0.44 (0.05)m·s]. In other words, between slow and preferred speed, the speed increase was mediated more by a change in SF than SL. The intra-subject variability of WR was low under preferred [CV, coefficient of variation = 1.9 (0.6)%] and fast [CV=1.8 (0.5)%] speed conditions, but higher under low speed condition [CV=4.1 (1.5)%]. On the other hand, the inter-subject variability of WR was 11%, 10% and 12% at slow, preferred and fast walking speeds respectively. It is concluded that the GPS method is able to capture basic gait parameters over a short period of time (5s). A specific gait pattern for slow walking was observed. Furthermore, it seems that the walking patterns in free-living conditions exhibit low intra-individual variability, but that there is substantial variability between subjects