Effect of compensation procedures for velocity on repeatability and variability of gait parameters in normal subjects

Objective: To evaluate the effects of a mathematical procedure that adjusts for gait velocity on the variability seen in gait analysis. Design: Evaluation before and after compensation. Setting: Gait laboratory, Fukui University Hospital, Japan. Subjects: Fourteen normal volunteers. Outcome measures: A computerized gait analysis system with two forceplates and a light source spot measuring device was used. Gait measurement in each subject was performed on three different days. Results: The gait parameters measured on three different days were significantly different, particularly step length, stride length, velocity, the components of floor-reaction forces and hip motion. In these parameters, intraclass correlation coefficient, ICC (1,1) was not high (range 0.05-0.71). However, there was a high correlation between these parameters and velocity and high repeatability was obtained following compensation for velocity (ICC (1,1), range 0.73-0.97). In contrast, compensation of parameters whose measurement was not significantly different before compensation did not improve ICC (1,1). Variability of all parameters was acceptable, however CV (an index of variability) improved significantly after compensation compared with that before compensation in six of 15 parameters. Conclusions: Our findings suggest that low repeatability in gait parameters should be considered even when the gait of normal subjects is measured on different days. A high repeatability and more acceptable variability were obtained when the data were compensated for velocity