Trial Outcomes & Findings for The Neural Mechanisms of Split-belt Treadmill Adaptation in People With Multiple Sclerosis (NCT NCT05878873)

A total of 55 participants consented to be screened for eligibility, and 51 participants (31 people with multiple sclerosis and 20 healthy controls) were enrolled.

Rows are split between people with multiple sclerosis and healthy controls.

Cortical activation is measured using functional near-infrared spectroscopy (fNIRS) during split-belt treadmill walking. Hemodynamic responses are modeled using a general linear model (GLM) applied to the oxyhemoglobin (HbO) signal. The model includes regressors for distinct phases of walking, with the primary contrast comparing early adaptation (strides 6-30 after split-belt onset) to a baseline walking period. The outcome is defined as the difference in this HbO beta weight contrast with TENS ON compared to TENS OFF. Activation is averaged across all fNIRS channels to provide a whole-brain estimate of cortical activity. A larger value indicates a greater increase in activation from baseline walking to early adaptation. This was measured on both training session 1 and training session 2 to account for the crossover design (i.e. participants are receiving TENS on different days).

Adaptation savings is defined as the difference in early adaptation performance between training session 1 (Day 1) and training session 2 (Day 28) during split-belt treadmill walking. Early adaptation is quantified using relative step length asymmetry (SLA), calculated from strides 6 to 30 following split-belt onset. SLA is computed from three-dimensional motion capture and force data as the difference between step lengths of the legs, normalized to total stride length: SLA = (Step Length\_fast - Step Length\_slow) / (Step Length\_fast + Step Length\_slow). This yields a unitless measure of asymmetry. The outcome measure is the difference in SLA between visits (training session 2 - training session 1). Larger values reflect faster adaptation at training session 2, consistent with retention of prior learning.

Step length asymmetry during early adaptation, representing the rate of adaptation. Early adaptation is quantified using relative step length asymmetry (SLA), calculated from strides 6 to 30 following split-belt onset. SLA is computed from three-dimensional motion capture and force data as the difference between step lengths of the legs, normalized to total stride length: SLA = (Step Length\_fast - Step Length\_slow) / (Step Length\_fast + Step Length\_slow). This yields a unitless measure of asymmetry. The outcome measure is the difference in early adaptation SLA during TENS ON compared to TENS OFF. Values closer to zero reflect faster adaptation.This analysis was performed only on data from each participant's first visit to avoid known effects of increased adaptation rate (learning) during subsequent exposures.