Objective: To evaluate whether digital mobility and autonomic trajectories detect prediagnostic decline beyond normative aging.

Background: Identifying individuals in prodromal Parkinson disease who are nearing meaningful progression remains a major barrier to earlier intervention. Absolute digital thresholds are limited by between-person variability and aging effects. Large real-world datasets from existing digital Parkinson disease networks may enable study of personalized prediagnostic trajectories at scale.

Method: We analyzed retrospective smartphone- and smartwatch-derived data from a large consumer-facing Parkinson disease platform with more than 15,000 active users and deployment across 16+ clinical and pharmaceutical studies. Because historical phone- and watch-based data ingestion had been permissioned, participants who joined after diagnosis still contributed measurements from before diagnosis. Analyses included approximately 1,989 users with mobility data and a subgroup with both prediagnostic walking speed and heart rate variability (HRV) data. Mobility measures were derived from smartphone-based walking behavior near the body’s center of mass, and autonomic measures from smartwatch-derived HRV. We examined age effects on walking speed, within-person prediagnostic change, HRV decline tertiles, and whether HRV-defined groups showed corresponding differences in walking speed, step length, and double support.

Results: Walking speed was lower with older age across individuals, but within-person prediagnostic decline extended beyond this between-person age effect. Stratifying participants by HRV decline rate revealed distinct autonomic trajectory groups with similar age distributions, arguing against simple age-driven separation. Those same HRV-defined groups also showed graded differences in mobility trajectories across walking speed, step length, and double support. Concordant signals across independent devices therefore revealed detectable prediagnostic decline in both autonomic and mobility domains.

Conclusion: Personalized mobility and autonomic trajectories capture prediagnostic change that appears to extend beyond aging alone. These findings support digital longitudinal trajectories as a promising way to identify individuals nearing prodromal progression and provide rationale for prospective evaluation in prodromal cohorts and early-stage trial enrichment.

« Back to 2026 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/digital-prediagnostic-mobility-and-autonomic-trajectories-in-a-large-parkinson-disease-network/