Abstract
Mobility impairment is common in multiple sclerosis, especially in older adults with multiple sclerosis (OAMS). Grey matter (GM) changes are well documented in MS, and GM atrophy is common in older adults. The relationship between GM changes and mobility disability in OAMS is unknown. We sought to identify GM patterns associated with gait speed in OAMS and healthy older controls, using structural covariance network analysis. OAMS (n = 102; 64.8 ± 4.4 years) and healthy controls (n = 106; 68.2 ± 7.3 years) underwent brain MRI and gait assessments; structural covariance networks were constructed to elucidate brain regions with significant associations between GM volume and 25-foot walk gait speed. We used voxel-wise linear regression analyses to elucidate per-network subregions with significant correlations with gait speed. Voxel-wise moderation analysis tested for group differences in these associations. Across the entire cohort, the following networks demonstrated significant gait speed associations: bilateral hippocampus, bilateral caudate/pallidum/putamen, bilateral thalamus/putamen, right middle temporal gyrus and multiple cerebellar regions. There were no significant group-by-network interaction effects. In summary, structural network analysis reveals unique brain patterns of gait speed in older adults, but these patterns are common amongst healthy older adults and OAMS and highlight the importance of cerebellar and subcortical networks in supporting gait speed.