The reliability of spatiotemporal gait data for young and older women during continuous overground walking.

OBJECTIVE

To examine the reliability and systematic bias in spatiotemporal gait parameters recorded in healthy women during repeated single and continuous overground walking trials.

DESIGN

Test-retest.

SETTING

University laboratory.

PARTICIPANTS

Young (n=13) and older adult (n=14) women volunteers.

INTERVENTIONS

Not applicable.

MAIN OUTCOME MEASURES

Spatiotemporal data were collected from an 8.1-m GAITRite mat during 10 trials of discrete single walks and 10 laps of a continuous circuit presented in random order over 2 separate test sessions. Paired t tests, intraclass correlation coefficients (ICCs), SE of measurement, and coefficients of variation (CV) were calculated.

RESULTS

The relative and absolute measures of reliability showed most spatiotemporal variables recorded during the single and continuous walking protocols were reliable. Step length, foot angle, and step and stance times were found to be the most reliable parameters, with ICCs ranging from 0.84 to 0.95, CVs from 2.06% to 4.02%, and SE of measurements of 1.59 to 2.04 cm for step length, 1.32 degrees to 1.71 degrees for foot angle, and 0.011 to 0.025 seconds for step and stance times. Reliability estimates were similar for the single and continuous trial conditions and between the young and older women. Although small mean differences in the gait parameters were found across the test sessions, many of these parameters showed systematic bias (P<or=.05). In the single trial condition, the majority (65%) of the gait parameters showed significant bias, whereas in the continuous condition only 19% of the parameters exhibited bias. For the young women, 54% of the parameters showed systematic bias (P<or=.05) in the single trial condition, whereas 77% of the parameters exhibited bias for the older women. In the continuous walking condition, 38% of the gait parameters showed systematic bias (P<or=.05) for the young women, whereas no systematic bias was found in the gait parameters of the older women.

CONCLUSIONS

This study shows that both the single and continuous walking protocols are reliable methods for the collection of gait data in young and older women. It also shows that a continuous overground walking protocol produces less bias in test-retest spatiotemporal gait data. Therefore, a continuous protocol may be a better method when attempting to monitor gait changes over time, especially for older women.