To evaluate the effectiveness of various digital health technology (DHT) measures in differentiating early stage Parkinson’s disease (PD) from healthy controls and monitoring short-term disease progression.
Approach:
Study Design: A multicenter, observational study examining tremor, bradykinesia, and axial symptom measures across healthy volunteers and newly diagnosed PD patients, with follow-up assessments.
Participants: Involved 45 age-matched healthy volunteers and 54 newly diagnosed PD patients, with 40 participants followed after 12 months.
Data Collection: Participants wore sensors to collect data on motor functions, and clinical assessments were conducted using the MDS-UPDRS.
Key Findings:
Some functional measures effectively differentiate healthy controls from newly diagnosed PD patients but show limited sensitivity to early progression.
Other measures are insensitive to initial diagnosis yet can capture longitudinal changes.
Models trained on disease-stage specific features demonstrated improved performance for progression detection.
Interpretation:
The study emphasizes the need for digital biomarker design and feature selection to align with disease stage and clinical objectives to enhance monitoring and diagnosis.
Limitations:
The cohort was predominantly white and well-educated, which may limit generalizability.
Exclusion of participants who initiated symptomatic therapy during the study may affect the findings.
Conclusion:
Adaptive, symptom- and side-specific DHT measures may improve sensitivity in trial population selection and short-term progression monitoring.
by Matthew D. Czech, Samantha Sawicki, Cindy Zadikoff, Chengcheng Liu, Weining Robieson, Ying Liu, Weihua Shi, Jie Shen, Michelle Crouthamel, Maria S. Quinton, Josh Cosman, E. Ray Dorsey, Jamie L. Adams, Naomi Nevler
