Objective: To investigate whether passively acquired measures of gait and hand gestures are associated with health-related quality of life (HrQoL; PDQ-39) and health-state (EQ-5D-5L) in early Parkinson’s disease (PD).

Background: The severity of motor impairment among individuals with PD affects their perceived HrQoL. Smartphone- and smartwatch-based sensors appear to represent a reliable and valid means to estimate motor impairment in daily life.

Method: Data were collected from 316 individuals with early PD (<2y) participating in a phase II clinical trial (PASADENA, NCT03100149). Two-week averaged sensor data centered around the baseline and PDQ-39 scores were corrected for age and gender using regression analysis. Impaired patients were defined as those with PDQ-39 and EQ-5D-5L scores greater than 0 and 1, respectively. The associations between digital sensor features and HrQoL/health-state was tested using Spearman’s rank order correlations (PDQ-39 scores, [N=275]) and logistic regression (EQ-5D-5L [N=279], PDQ-39 item level, controlling for age and gender).

Results: Hand gesture data were negatively correlated to PDQ-39 communication, mobility, and PDQ-39 index scores (all rs(≥209) ≤.-14, p<.05) and distinguished non-impaired from impaired individuals across all EQ-5D-5L domains (e.g., lower median gesture power predicted difficulties with usual activities, β=-2.8, z=-.14, p =.005). Gait features correlated negatively and positively (depending on the sensor feature) with PDQ-39 mobility, communication, and social support scores (all rs[≥210] ≤|.14|, p<.05) and predicted endorsement of anxiety/depression and impairment in usual activities and self-care (EQ-5D-5L; e.g., higher step frequency variance predicted difficulties with usual activities, β=2.7, z=3.0, p=.003). Moreover, gesture and gait features were related to responses on PDQ-39 item level, e.g., individuals who reported difficulties with buttons/shoelaces showed less hand movements (β=-6.4, z=-3.6, p<.001) and those who reported difficulties getting around public places showed reduced turning speed (β=-3.6, z=-2.2, p<.001).

Conclusion: Passively acquired sensor data during PD patients’ daily lives were associated with self-reported HrQoL and health status. These findings may inform future developments of digital outcome scores that relate to meaningful, HrQoL-affecting impairment in motor function.

« Back to MDS Virtual Congress 2020

MDS Abstracts - https://www.mdsabstracts.org/abstract/passive-monitored-daily-motor-behavior-significantly-relates-to-quality-of-life-in-early-parkinsons-disease/