TL;DR
This study leverages wearable sensors to analyze gait variability during a 6-minute walk test (6MWT) and successfully differentiates between healthy individuals and those with lumbar spinal stenosis (LSS) or knee osteoarthritis (KOA). Notably, the middle portion of the 6MWT emerged as a critical window for detecting subtle differences in stride length variability, suggesting it can help phenotype the underlying cause of gait impairment.
What Problem Does This Paper Address?
Mobility limitations are a growing concern, particularly in aging populations. While gait analysis can identify individuals with mobility disorders, pinpointing the specific underlying pathology remains a challenge. Lumbar spinal stenosis (LSS), a neurogenic condition, and knee osteoarthritis (KOA), a structural joint disease, are two common causes of mobility impairment, yet they manifest differently. Existing gait analysis methods often rely on lab-based equipment and may not capture the dynamic changes that occur during a walk. This paper aims to determine if gait parameters, specifically variability, derived from wearable sensors during a 6MWT can effectively phenotype these distinct orthopedic gait impairments.
Methods
The researchers employed wearable inertial measurement units (IMUs) attached to both feet of participants. These IMUs captured motion data during a 6-minute walk test (6MWT). Eleven gait parameters were extracted, categorized into four domains: rhythm, pace, asymmetry, and variability. A key methodological innovation was the minute-by-minute analysis of these parameters throughout the 6MWT, allowing for the investigation of dynamic changes over time. This temporal analysis was crucial for identifying subtle differences that might be missed when considering the entire walk as a single event. Statistical analysis involved one-way ANOVA for overall group comparisons and two-way ANOVA for minute-by-minute comparisons, with a significance level of p < 0.05.
Datasets
The study utilized data from 30 participants: 10 with Lumbar Spinal Stenosis (LSS), 10 with Knee Osteoarthritis (KOA), and 10 matched healthy controls (HC). The full inclusion/exclusion criteria were detailed in a previous publication. No new benchmark datasets were introduced.
Key Insights and Ideas
- Gait Variability as a Phenotyping Tool: The study strongly suggests that gait variability, particularly stride length variability, can serve as a sensitive biomarker to distinguish between neurological gait impairments (like LSS) and structural joint diseases (like KOA).
- The 6MWT's Middle Ground: The middle portion (specifically the 3rd and 4th minutes) of the 6MWT proved to be the most informative for detecting differences in stride length variability between the LSS and KOA groups. This highlights the importance of analyzing dynamic changes within a test, rather than just overall averages.
- Neurological vs. Structural Impairment: Increased stride length variability in LSS, a neurogenic condition, is linked to a loss of gait rhythmicity, which is a known consequence of neurological control issues. This contrasts with KOA, where the impairment stems from structural joint problems.
Key Results
- Overall Differentiation: During the entire 6MWT, gait parameters across all four domains (rhythm, pace, asymmetry, and variability) successfully distinguished between healthy controls and both disease groups (LSS and KOA).
- Disease Group Similarities (Overall): When considering the entire 6MWT, no statistically significant differences were found between the LSS and KOA groups for most parameters. However, a trend towards higher stride length variability was observed in the LSS group (p = 0.057).
- Minute-by-Minute Differences: The minute-by-minute analysis revealed a significant difference in stride length variability between LSS and KOA groups during the middle portion of the 6MWT. Specifically, the LSS group showed significantly higher stride length variability in the 3rd minute (p < 0.05) and a trend towards higher variability in the 4th minute (p = 0.06).
Limitations and Future Work
- Small Sample Size: The study acknowledges its small sample size, typical for early investigations, and suggests that larger studies are needed to confirm the findings.
- 6MWT Duration: The authors note that a six-minute walk might not be sufficiently long to elicit significant gait changes in all groups, particularly for detecting subtle differences.
- Disease-Specific Sensitivity: The most sensitive portion of a gait test for phenotyping may be disease-specific, requiring further investigation.
- Future Directions: The findings suggest that gait variability, particularly stride length variability, is a promising biomarker for phenotyping gait impairments. Future research should focus on validating these findings in larger cohorts and exploring other dynamic gait tests to further refine the ability to distinguish between different causes of mobility limitations.
Summary updated