Gait speed assessment in confined spaces: Development of a novel automated 4‐m static‐start test to measure dynamic‐start gait speed

Abstract

OBJECTIVES: To promote standardization and feasible measurements of gait speed across the field, we developed a device that used light detection and ranging (LiDAR) technology to measure gait speed from a standing-start testing procedure conducted on a 4-m total walkway. We compared this automated (LiDAR-based) standing-start 4-m gait speed test (AS-4MG) with automated and manual (stopwatch-based) dynamic-start 4-m gait speed tests (AD-4MG and MD-4MG, respectively) on between-method agreement, measurement repeatability, and predictive validity with functional outcomes.

METHODS: A sample of 48 community-dwelling adults (mean [SD], 69 [9] years) participated in this method comparison study. Participants completed a survey and a physical function assessment, from which self-reported stair difficulty, handgrip strength, sit-to-stand performance, and gait speed (AS-, AD-, and MD-4MG) were measured.

RESULTS: Mean AS-4MG, AD-4MG, and MD-4MG were 1.12, 1.13, and 1.14 m/s, respectively, with a strong correlation (r = 0.94) and no systematic bias observed between AS-4MG and the referent AD-4MG. In the analyses of repeated gait measurements, coefficients of repeatability were <0.20 m/s for the automated tests (0.16 and 0.18 m/s for AS-4MG and AD-4MG, respectively) but not for MD-4MG (0.22 m/s). Correlations between gait speed and functional measures ranged between 0.44 and 0.68 (Ps < 0.01). Correlations for the automated tests were comparable, while AS-4MG correlations tended to exceed MD-4MG correlations.

CONCLUSIONS: In community-dwelling adults, the AS-4MG test was unbiased when compared with the AD-4MG, encompassing a 9-m walkway, with both tests showing equivalence of measurement repeatability and predictive validity. The AS-4MG potentially allows gait measurements in confined spaces (e.g., doctor's office) where the gait speed test was previously unfeasible. Geriatr Gerontol Int 2025; 25: 449-453.