Accelerometry-derived activity fragmentation as a predictor of brain atrophy and disability progression in multiple sclerosis

BackgroundAccelerometry-derived activity fragmentation reflects how frequently active periods are broken by sedentary periods. It has been associated with frailty, mortality, and reduced gait speed in the general population; however, its prognostic significance in multiple sclerosis (MS) remains unknown. MethodsPeople with MS (PwMS) wore wrist tri-axial accelerometers at three-month intervals, underwent disability assessments biannually and MRI brain approximately annually. We evaluated whether active-to-sedentary transition probability (ASTP) and sedentary-to-active transition probability (SATP) during the 10 most active hours of each 24-hour period (M10) were associated with confirmed disability progression and brain substructure atrophy, by modeling between- and within-person effects. ResultsAmong 239 PwMS, 120 had confirmed Expanded Disability Status Scale(EDSS)-plus progression over a mean of 2.9{+/-}1.1 years. Within-person increase in ASTP slope during M10 (indicating increasing activity fragmentation throughout M10) was associated with higher risk of subsequent EDSS-plus progression (per 1 SD increase: HR 1.22 [95%CI 1.02,1.46],p=0.03) and lower deep gray matter and thalamic volumes over time (per 1 SD increase: -0.35 [95%CI - 0.58,-0.13],p=0.002, -0.42 [95%CI:-0.68,-0.15],p=0.002 respectively). There were no similar associations for SATP. ConclusionsWithin-person worsening in activity fragmentation was associated with higher risk of neurologic decline. Activity fragmentation may represent an early compensatory strategy for reduced physiologic reserve and serve as an early indicator of MS progression.