Digital Biomarkers of Mobility in Parkinson's Disease During Daily Living

Abstract

Background: Identifying digital biomarkers of mobility is important for clinical trials in Parkinson's disease (PD).

Objective: To determine which digital outcome measures of mobility discriminate mobility in people with PD from healthy control (HC) subjects over a week of continuous monitoring.

Methods: We recruited 29 people with PD, and 27 age-matched HC subjects. Subjects were asked to wear three inertial sensors (Opal by APDM) attached to both feet and to the lumbar region, and a subset of subjects also wore two wrist sensors, for a week of continuous monitoring. We derived 43 digital outcome measures of mobility grouped into five domains. An Area Under Curve (AUC) was calculated for each digital outcome measures of mobility, and logistic regression employing a 'best subsets selection strategy' was used to find combinations of measures that discriminated mobility in PD from HC.

Results: Duration of recordings was 66±14 hours in the PD and 59±16 hours in the HC. Out of a total of 43 digital outcome measures of mobility, we found six digital outcome measures of mobility with AUC > 0.80. Turn angle (AUC = 0.89, 95% CI: 0.79-0.97) and swing time variability (AUC = 0.87, 95% CI: 0.75-0.96) were the most discriminative individual measures. Turning measures were most consistently selected via the best subsets strategy to discriminate people with PD from HC, followed by gait variability measures.

Conclusion: Clinical studies and clinical practice with digital biomarkers of daily life mobility in PD should include turning and variability measures.

Keywords: Parkinson’s disease; biomarkers; continuous monitoring; digital outcome measures of mobility; inertial sensors.

Conflict of interest statement

CONFLICT OF INTEREST

Drs. McNames, El Gohary, and Horak have a significant financial interest in APDM, a company that may have a commercial interest in the results of this research and technology. Dr. Horak also consults with Biogen, Neuropore, Sanofi, and Takeda. This potential conflict has been reviewed and managed by OHSU.

Figures

Fig. 1.

A representative example of a profile of the inertial sensor data from feet and lumbar over a day, and the zoomed-in version of identified gait bouts and turns.

Fig. 2.

Receiver Operating Characteristic (ROC) curve showing the comparison of a typical activity measure (number of strides/hour) with the performance of the top digital outcome measures of mobility (AUC> 0.8) that differentiated PD from HC.

Fig. 3.

A) 5-fold cross-validated AUC versus Bayesian Information Criterion (BIC) plot for various combinations of digital outcome measures of mobility. B) ROC curve for the top models with 2 to 4 digital outcome measures of mobility (see Table 2).