Environmental Personal Exposure Clusters to Investigate Multiple Sclerosis and Amyotrophic Lateral Sclerosis Progression.

Reliable prognosis in Multiple Sclerosis (MS) and Amyotrophic Lateral Sclerosis (ALS) is hampered by data scarcity and variability. Beyond clinical variables, evidence suggests that environmental data can help capture disease trajectories. We investigated whether personal environmental measures can be organized into stable patterns that inform prognosis. In a multicenter cohort, 293 patients with MS or ALS were equipped with Atmotube air-quality sensors. We normalized volatile organic compound (VOC) time series and computed Dynamic Time Warping distances to capture temporal similarity. Hierarchical clustering yielded five daily exposure clusters, which were profiled using Atmotube variables (season, day type, humidity, temperature) and patient self-reports (work status, time outdoors), and evaluated by day-level differences between personal and fixed-station variables. These clusters can support interpolation of missing wearable intervals and generation of context-aware exposure estimates, thereby strengthening environmental inputs for prognostic modeling in MS and ALS.