2026 American Academy of Neurology Abstract Website

Objective:

To assess the relationships between research visit and real-world digital mobility outcome (DMO)-derived gait parameters and explore the correlations between DMOs and multiple sclerosis (MS)-related disability.

Background:

Over 75% of people with MS (pwMS) experience significant gait abnormalities, yet current gold-standard mobility assessments, such as timed walks, lack the sensitivity to detect subtle changes in gait mechanics. DMOs, derived from wearable sensors, may provide sensitive assessments of mobility impairment in clinical and real-world settings for pwMS.

Design/Methods:

Thirty adults with stable MS and Expanded Disability Status Scale (EDSS) score <6.5 were recruited from the Oregon Health & Science University MS clinic. Participants underwent a research visit assessing DMOs (i.e. 2 Minute Timed Walk [2MTW]) and patient reported outcomes (PROs) followed by seven days of real-world sensor measurements (worn on lumbar spine and bilateral ankles). Real-world DMO data was associated with research visit DMO data and patient reported outcomes (PROs) using Pearson correlations. The first 10 participants are presented here, and the rest of the cohort (n=30) will be included in the conference presentation.

Results:

The mean age of this cohort (n=10) was 55 years, 60% were female, and median EDSS score was 3.5 (2.5 - 6.5). Median EDSS of the entire cohort (n=30) was 5.0. Real-world increased average cadence was strongly and significantly correlated with longer 2MTW length during the research visit (r=0.69, p=0.04). Real-world 90^th percentile cadence during walking bouts lasting over 30 seconds was strongly and significantly correlated with 2MTW increased average gait speed (p=0.73, p=0.04), increased average cadence (p=0.88, p=0.004), decreased average gait cycle duration (p=-0.89, p=0.003), and decreased average step duration (r=-0.89, p=0.003). There were significant associations between real-world average speed and MS Impact Scale (0.65, p=0.4).

Conclusions:

Using sensors to measure real-world DMO could provide an accurate assessment of a participant’s average level of functioning across many ability levels.