Abstract
BACKGROUND: Walking dysfunctions persist following poststroke rehabilitation. A major limitation of current rehabilitation efforts is the inability to identify modifiable deficits that, when improved, will result in the recovery of walking function. Previous studies have relied on cross-sectional analyses to identify deficits to target during walking rehabilitation; however, these studies did not account for the influence of a key covariate - maximum walking speed. OBJECTIVE: To determine the relationships between commonly studied poststroke variables and the long-distance walking function of individuals poststroke when controlling for maximum walking speed. METHODS: Correlation analyses of cross-sectional data from 57 individuals more than 6 months poststroke measured the relationships between standing balance, walking balance, balance self-efficacy, lower extremity motor function, and maximum walking speed versus long-distance walking function. For a subgroup of subjects who completed training, the relationship between changes in maximum walking speed versus changes in long-distance walking function was assessed. RESULTS: Each measurement of interest strongly correlated with long-distance walking function (rs from 0.448 to 0.900, all Ps ≤ .001); however, when controlling for maximum walking speed, none of the other measurements remained related to long-distance walking function. In contrast, when controlling for each of the other measurements, maximum walking speed remained highly related. Moreover, changes in maximum walking speed resulting from training were highly related to changes in long-distance walking function (r = .737, P ≤ .001). CONCLUSIONS: For individuals in the chronic phase of stroke recovery, improving maximum walking speed may be necessary to improve long-distance walking function.