Gait variability in community-dwelling older adults.

OBJECTIVES: To describe gait variability at usual and fast walking speeds in community-dwelling older adults and to describe the effects of increasing gait speed on gait variability.

DESIGN: Cross-sectional, descriptive study.

SETTING: The Cardiovascular Health Study at the University of Pittsburgh.

PARTICIPANTS: Ninety-five community-living older adults, 54 women and 41 men, age 65 and older (mean age +/- standard deviation 79.4 +/- 3.37).

MEASUREMENTS: Gait measured at participant's usual and fast walking speed collected using an instrumented walkway. Step-length and step-width variability were determined using the coefficient of variation.

RESULTS: Step-length variability was greatest in those who walked the slowest (r = -0.66, P < .001); step-width variability was smallest in those who walked the slowest (r -0.37, P < .001). Individuals who could not increase their walking speed (<0.10 m/second) on command had an increase in step-length variability and a decrease in step-width variability, whereas those who could increase their speed (>0.10 m/second) had an increase in step-width variability when walking at a faster speed.

CONCLUSIONS: Step-length and step-width variability have opposite associations with gait speed in older adults. Improvement in step-length and step-width variability with attempted acceleration might be a key factor to examine in future studies of disability risk and therapeutic interventions.