Gait Profile Score in able-bodied and post-stroke individuals adjusted for the effect of gait speed

Background: The Gait Profile Score (GPS) measures the quality of an individual's walking by calculating the difference between the kinematic pattern and the average walking pattern of healthy individuals. Research questions: The purposes of this study were to quantify the effect of speed on the GPS and to determine whether the prediction of gait patterns at a specific speed would make the GPS outcome insensitive to gait speed in the evaluation of post-stroke individuals. Methods: The GPS was calculated for able-bodied individuals walking at different speeds and for the comparison of post-stroke individuals with able-bodied individuals using the original experimental data (standard GPS) and the predicted gait patterns at a given speed (GPS velocity, GPS(v)). We employed standard gait analysis for data collection of the subjects. Sixteen participants with a stroke history were recruited for the post-stroke group, and 15 age-matched, able-bodied participants formed the control group. Results: Gait speed significantly affects the GPS and the method to predict the gait patterns at any speed is able to mitigate the effects of gait speed on the GPS. Overall, the gap between the GPS and GPS(v) values across the post-stroke individuals was small (0.5 degrees on average, range from 0.0 degrees to 1.4 degrees) and not statistically significant. However, there was a significant negative linear relationship in the absolute difference between the GPS and GPS(v) values for the participants of the post-stroke group with gait speed, indicating that a larger difference between the speeds of the post-stroke participant and the reference dataset resulted in a larger difference between the GPS and GPS(v). Significance: The modified version of the GPS, the GPS(v), is effective in reducing the impact of gait speed on GPS; however, the observed difference between the two methods was only around 1 degrees for the slowest individuals in comparison to the reference dataset.