Category: Parkinson’s Disease: Clinical Trials
Objective: To identify digital phenotypes from continuous wearable monitoring in a prospective Parkinson’s disease cohort and characterize their clinical profiles.
Background: Advanced Parkinson’s disease lacks standardized quantitative criteria for early identification. Wearable-derived digital phenotypes may reveal distinct clinical profiles and provide a foundation for future risk stratification.
Method: Fifty patients were prospectively enrolled. Baseline clinical variables were harmonized with the PPMI dataset. Participants underwent a comprehensive clinical assessment evaluating motor severity, disease stage, general cognition, functional independence in activities of daily living, and non-motor domains including sleep disturbances, mood symptoms, autonomic dysfunction, quality of life, freezing of gait, and overall non-motor symptom burden. Continuous monitoring for approximately one week was performed using wrist- and waist-worn sensors capturing motor activity, fluctuations, gait, heart rate, and sleep. Unsupervised clustering of wearable metrics identified digital phenotypes, which were compared with clinical measures.
Results: Three different phenotypes emerged from digital wearable monitoring, differing in daily activity, OFF time, sleep duration, and heart rate variability. Kruskal-Wallis tests revealed significant differences across phenotypes in motor severity (MDS-UPDRS, p = 0.0085), disease stage (Hoehn & Yahr, p = 0.0061), cognitive function (MOCA, p = 0.0224), sex distribution (p = 0.0058), and olfactory function (UPSIT, p = 0.0343). No significant differences were observed for age or other clinical measures. These findings suggest that wearable-derived digital phenotypes capture clinically meaningful differences in motor, non-motor, and autonomic function, independent of demographic factors.
Conclusion: Continuous wearable monitoring can identify distinct digital phenotypes in Parkinson’s disease, reflecting motor, non-motor, and autonomic profiles. These phenotypes provide a foundation for future AI-based risk stratification and potential digital biomarkers for disease progression.
To cite this abstract in AMA style:
A. Saenz, I. Gabilondo, S. Seijo, A. Ochoa, U. Zalabarria, I. Cuenca, B. Tijero, T. Fernández, M. Ruiz, M. Acera, J. Gomez, R. Del Pino. Identification of Phenotypes in a Prospective Parkinson’s Disease Cohort Using Digital Wearable Monitoring [abstract]. Mov Disord. 2026; 41 (suppl 1). https://www.mdsabstracts.org/abstract/identification-of-phenotypes-in-a-prospective-parkinsons-disease-cohort-using-digital-wearable-monitoring/. Accessed October 1, 2026.« Back to 2026 International Congress
MDS Abstracts - https://www.mdsabstracts.org/abstract/identification-of-phenotypes-in-a-prospective-parkinsons-disease-cohort-using-digital-wearable-monitoring/