Agreement between BTS G-walk and GaitLab in spatiotemporal and pelvic angle measurements in active older adults

TL;DR

This study compares the accuracy of two gait analysis tools in older adults, finding that one is reliable for spatial measurements but less so for timing and pelvic angles.

Contribution

The study provides a large-sample validation of the BTS G-Walk system's performance in measuring gait and pelvic angles in active older adults.

Findings

BTS G-Walk accurately measures spatial gait parameters like cadence and speed in older adults.

Temporal parameters and pelvic rotation show poor agreement with GaitLab, likely due to lower sampling rate and sensor placement.

Pelvic tilt and obliquity show moderate agreement, but rotation has low agreement.

Abstract

Aging impacts gait, a vital health indicator in older adults, reducing speed and increasing double support time, linked to falls and disability. The World Health Organization advises 150 min/week of moderate or 75 min/week of vigorous exercise, plus strength training, to sustain neuromuscular integrity and locomotor capacity associated with healthy gait patterns in older adults. While 3D motion capture is the gold standard for gait analysis, its cost and complexity limit use, boosting interest in portable inertial sensors like the BTS G-Walk. Yet, their accuracy in active older adults for spatiotemporal parameters and pelvic angles is underexplored, prompting this comparison with BTS GaitLab.

Fifty-nine active older adults (aged 65–87, mean 71.2 ± 5.5 years; eight men, 51 women) were assessed using BTS GaitLab (200 Hz cameras, 400 Hz force plates, Helen Hayes protocol, 18 markers) and BTS G-Walk (100 Hz, at L5) during four 10-meter walks at comfortable speed. Spatiotemporal parameters (cadence, speed, step length, stance/swing/support times) and pelvic angles (tilt, obliquity, rotation) were compared. Agreement used intraclass correlation coefficient (ICC), differences used paired t-tests, and variability used coefficient of variation (CV), with p < 0.05.

Spatial parameters showed strong agreement: cadence (ICC = 0.98), speed (ICC = 0.96), step/stride length (ICC = 0.90–0.92), with no significant differences for cadence and speed (p > 0.05). Temporal parameters had weak agreement (e.g., stance time ICC = −0.18, swing time ICC = −0.54) and significant differences (p < 0.001). Pelvic angles had moderate agreement for tilt and obliquity (ICC = 0.48–0.78, p < 0.007), but low for rotation (ICC < 0.38, p > 0.038). Variability was high for double support time (CV 15.3–20.9%) and pelvic angles (CV > 27%). The sample (59) exceeds typical studies (7–23).

BTS G-Walk accurately measures spatial gait parameters in active older adults, ideal for quick clinical assessments where 3D systems are unavailable. However, its weak temporal parameter and pelvic rotation performance, possibly due to 100 Hz sampling, L5 placement, and algorithms, limits detailed analysis like stability. With a large sample and pelvic angle focus, this study supports G-Walk as a complementary tool in gerontology, while noting needs for improved temporal and angular precision.

Figures