Identifying and predicting fast versus slow Parkinson's disease motor progressors using clinical and digital data

Area of Science:

• Neurology
• Digital Health
• Biomarkers

Background:

• Parkinson's disease (PD) progression is heterogeneous, complicating clinical trials.
• Objective, high-frequency digital health measures offer potential for individual patient stratification.
• Current methods lack sufficient precision for identifying fast vs. slow motor progressors.

Purpose of the Study:

• To identify and predict fast versus slow motor progressors in Parkinson's disease.
• To integrate longitudinal clinical data with smartphone-based motor testing.
• To assess the utility of digital health tools for personalized PD management.

Main Methods:

• Subgroup analysis of the 96-week Exenatide-PD3 trial.
• Motor progression assessed via MDS-UPDRS Part III OFF scores and smartphone testing.
• Data-driven subtyping to identify progressor groups; mixed-effects models for longitudinal analysis.
• Baseline clinical and digital features used to predict 96-week motor progression.

Main Results:

• Data-driven clustering identified 26.5% of participants as fast motor progressors.
• Fast progressors exhibited significantly higher baseline and 96-week motor impairment.
• Prediction models combining smartphone features with clinical scores outperformed clinical models alone (AUC=0.78 vs. 0.53).
• Smartphone assessments demonstrated high user acceptability (>96%) and predictive value (AUC=0.80).

Conclusions:

• One in four early-to-mid stage PD patients were identified as fast motor progressors.
• Integrating smartphone data with clinical scores enhances baseline motor progression prediction.
• This approach facilitates individual-level PD stratification, addressing heterogeneity and improving trial design.