Real-world Digital Biomarkers Index Circadian Rest-activity Rhythms in Persons Taking Dopaminergic Medications for Parkinson’s Disease
Jun Ha Chang1, Danish Bhatti2, Ergun Uc3, Jennifer Merickel1, Matthew Rizzo1
1University of Nebraska Medical Center, 2University of Central Florida College of Medicine, 3University of Iowa
Objective:

Quantify circadian rest-activity rhythms in persons taking dopaminergic medications for Parkinson’s Disease (PD), using real-world (RW) outputs from personal devices (actigraphy) as actionable clinical data for PD clinical management.

Background:

Circadian rhythms are physical, mental, and behavioral changes over 24-hour cycles. Clinical assessments and self-report suggest that rest-activity rhythms correlate with PD symptom severity. Actigraphy indexes continuous RW activity and has potential to objectively index circadian PD symptom severity and effect of dopaminergic treatment.

Design/Methods:

Twenty-seven patients with PD on dopaminergic medications (mean age=67.8 years, 19 men, median H&Y=2) wore wrist actigraphy for four weeks (N=636 days). Dopaminergic medication use as indexed by Levodopa Equivalent Dose [LED] was classified as High:>400mg/day (N=9) and Low:≤400mg/day (N=18). Actigraphy captured circadian rest-activity rhythms including intra-day variability (IV, low better), inter-day stability (IS, high better), and relative amplitude (RA, high better). We performed multivariate linear regressions with MDS-UPDRS section scores (outcome) and the circadian rest-activity rhythms (predictor), stratified by dopaminergic medication use, adjusted for age, gender, working status, and MDS-UPDRS Part III score.

Results:

Worse non-motor symptoms (MDS-UPDRS Part I) in Low LED PD participants correlated with more irregular rhythms on all 3 actigraphy measures (IV: b=19.9, p=0.013; IS:  b=-49.4, p=0.001; RA: b=-31.9, p=0.015). Correlations of MDS-UPDRS Part 1 and its other sections with circadian rest-activity rhythms across all patients or across high/low LED strata were non-significant.

Conclusions:

Findings show that non-motor symptom severity in PD is associated with disrupted circadian rest-activity cycles in patients who use lower doses of dopaminergic medications, suggesting higher dosing might improve non-motor symptoms and restore normal circadian rest-activity cycles. Findings on circadian rest-activity cycles using actigraphy offer potential new digital biomarkers for PD, which can optimize dopaminergic drug dosing in PD patients and be used in objectively evaluating RW drug efficacy in clinical trials.

10.1212/WNL.0000000000204550