Journal of NeuroEngineering and Rehabilitation (May 2019)
Development and validation of Comprehensive Gait Assessment using InerTial Sensor score (C-GAITS score) derived from acceleration and angular velocity data at heel and lower trunk among community-dwelling older adults
Abstract
Abstract Background Although some gait parameters from inertial sensors have been shown to be associated with important clinical issues, because of controversial results, it remains uncertain which parameters for which axes are clinically valuable. Following the idea that a comprehensive score obtained by summing various gait parameters would sensitively reflect declines in gait performance, we developed a scoring method for community-dwelling older adults, the Comprehensive Gait Assessment using InerTial Sensor score (C-GAITS score). The aim of this study was to examine the internal consistency and the construct validity of this method. Methods In this cross-sectional study, the gait performance of 378 community-dwelling older people (mean age = 71.7 ± 4.2 years, 210 women) was assessed using inertial sensors attached to the heel and lower trunk. Participants walked along a 15-m walkway, and accelerations, angular velocity, and walking time were measured. From these data, walking speed, mean stride time, coefficients of variation of stride time and swing time, and autocorrelation coefficients and harmonic ratios of acceleration in vertical, mediolateral, and anteroposterior directions at the lower trunk were calculated. Scoring was performed based on quartile by gender (i.e., scored from 0 to 3) for each of the 10 gait parameters. The C-GAITS score was the sum of these scores (range: 0–30). Lower extremity strength, balance function, fall history, and fear of falling were also assessed. Results An exploratory factor analysis revealed that the C-GAITS score yielded four distinct factors explaining 57.1% of the variance. The Cronbach’s alpha coefficient was 0.77. A single linear regression analysis showed a significant relationship between total C-GAITS score and walking speed (adjusted R 2 = 0.28). Results from bivariate comparisons using unpaired t-tests showed that the score was significantly related to age (p = 0.002), lower extremity strength (p = 0.007), balance function (p < 0.001), fall history (p = 0.04), and fear of falling (p < 0.001). Conclusions Good internal consistency and appropriate construct validity of the C-GAITS score were confirmed among community-dwelling older adults. The score might be useful in clinical settings because of ease of use and interpretation and capability of capturing functional decline.
Keywords