The Relationship Between Variability of the Center of Mass and Base of Support Is Altered during Gait in Persons with Multiple Sclerosis
Objectives: The purpose of this study was to measure variability of movement at the trunk relative to variability of movement at the feet during walking to determine if persons with multiple sclerosis (PwMS) exhibit altered relationships between the trunk and foot segments during walking.
Methods: Forty PwMS (age 21-57) and forty age-matched HC walked on a treadmill for 3-minutes at self-selected pace wearing tri-axial accelerometers on the sternum and on the right ankle. Acceleration in the frontal and sagittal planes of motion were obtained. Linear and nonlinear variability measures were calculated from the acceleration time series. Linear measures: root mean square (RMS), range, and standard deviation (SD). Nonlinear measures: approximate entropy (ApEn), sample entropy (SaEn), and Lyapunov exponents (LyE). The ratio of trunk to foot acceleration variability was calculated for each variability measure. T-tests examined differences in this ratio between groups.
Results: Gait speed was similar between groups. In the frontal plane, PwMS have a significantly lower trunk to foot acceleration ratio compared to HC for RMS, range, SD, and LyE, but a significantly higher trunk to foot acceleration ratio compared to HC for ApEn and SaEn (p<0.05 for all). In the sagittal plane, PwMS have a significantly lower trunk to foot acceleration ratio compared to HC for RMS and LyE only (p<0.05).
Conclusions: The present study shows that relationships between trunk and foot motion are altered in PwMS compared to HC, particularly in the frontal plane. During gait, frontal plane motion is largely controlled by active sensorimotor mechanisms. Since PwMS have slowed neural conduction velocities, their relationship between trunk and feet is altered and likely affects postural control during gait and could be related to fall risk. There are fewer group differences in the sagittal plane, likely due to relatively passive motion control in this plane being unaltered in PwMS. Examining the relationship between motion at the trunk and the feet during gait using wireless sensors could be further developed for fall risk assessment in clinical and real-world settings.