Related Experiment Video
Updated: Jun 25, 2026
Evaluation of Patients' Posture and Gait Profile After Lumbar Fusion Surgery by Video Rasterstereography and Treadmill Gait Analysis
Published on: March 23, 2019
Chronic low back pain is associated with remotely collected gait quality and quantity parameters from a wrist-worn
Daniel Ky Zheng1, Xijun Sun1, Tongyu Ma1
1Department of Rehabilitation Sciences, The Hong Kong Polytechnic University, Hong Kong SAR, China.
Background Context:
Although laboratory research has revealed altered gait in some individuals with chronic low back pain (CLBP), large-scale studies using wearable sensors to characterize gait patterns of individuals with CLBP in real-world settings are lacking.
Purpose:
To identify key gait parameters from wrist-worn accelerometers that distinguish individuals with CLBP from those without in everyday environments.
Study Design:
Cross-sectional study.
Patient Sample:
A total of 8,093 participants (979 with CLBP and 7,114 asymptomatic controls) from the UK Biobank who wore accelerometers for 5 to 7 days and provided CLBP data.
Outcome Measures:
CLBP status was self-reported. Twelve gait parameters related to gait quantity and quality were extracted from wrist-worn accelerometers' data.
Methods:
To minimize confounding, the propensity score matching was applied to create 2 balanced comparison groups. Independent t-tests and logistic regression analyses were conducted to identify key gait parameters that best discriminated CLBP status. Effect sizes were expressed as odds ratios (OR), and discriminative performance was assessed using the area under the receiver operating characteristic curve (AUC).
Results:
After matching, 1,956 participants were included in the final analysis: 978 with CLBP (mean age 61.4±7.7 years, 52.2% female) and 978 controls (61.4±7.6 years, 53.3% female). Compared to controls, those with CLBP displayed significantly lower daily gait quantity (fewer steps per day, shorter longest walking duration, slower maximal walking speed, and lower cadence) and inferior gait quality (higher step-time variability and lower step regularity) (p<.05). Among all gait parameters, step-time variability was the most discriminative parameter, even after adjusting for sociodemographic, psychological, and sleep-related covariates. However, its effect size (OR = 1.18-1.19) and discriminative performance (AUC = 0.52-0.57) were limited.
Conclusions:
Wrist-derived gait parameters, particularly step-time variability, reflect functional differences in individuals with CLBP. Wrist-worn accelerometers show promise for remote gait monitoring, although findings are limited by the cross-sectional design, self-reported pain status, and modest effect size and discriminative ability. Future rigorously designed prospective studies should clarify causal relationships and clinical utility.
