Objective: To evaluate longitudinal prospective progression from gait and mobility data from individuals who were later diagnosed with PD.

Background: Diagnosis of clinical Parkinson’s disease (PD) is based on the existence of distinct motor symptoms, including gait deficits. Subtle motor alterations likely precede clinical diagnosis by many years but are often undetected. Wearable sensors can objectively monitor and quantify movement alterations, potentially identifying pre-clinical PD.

Method: Data was collected between 2012-2023 through the Genetics in Parkinson’s project. Non-diagnosed participants were included if they had a family history of PD or were non-manifesting carriers of genetic mutations associated with PD. Participants walked for 1-minute while wearing inertial measurement units on bilateral wrists and lower back (OPAL, APDM/Clario). All participants had at least two assessment points. Twenty-nine mobility features were extracted from the acceleration and gyroscope data. The slope of  change between visits was calculated for each feature and compared between participants who later developed clinical PD (converters) and age-matched undiagnosed (non-converters).

Results: The study included 11 convertors (age 59.4±8.0 years at first visit; 45% male) and non-convertors (56.4±10.2 years; 64% male). At baseline, groups did not differ significantly on MDS-UPDRS II (p = 0.113), MDS-UPDRS III (p = 0.214), MoCA (p = 0.178), or gait speed (p = 0.956). Mean follow-up was 3.5±1.3 years; similar between groups (p = 0.270). The final assessment for converters occurred 1.1±1.5 years prior to diagnosis. Over time, converters showed steeper changes in stride variability (p = 0.010), arm-swing asymmetry (p = 0.004), and arm-swing amplitude p = 0.015) compared with non-converters. No significant differences were found in cadence, jerk, or gait speed (p > 0.727). Changes in MDS-UPDRS III and MoCA were similar between groups (p > 0.171), but MDS-UPDRS II increased more in converters (p = 0.031).

Conclusion: Subtle gait and arm-swing alterations were observed years before PD diagnosis in individuals who later converted, compared with non-converters. These results suggest that wearable sensors can quantify early motor changes preceding clinical diagnosis, providing a basis for prospectively validating digital biomarkers of prodromal PD.

« Back to 2026 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/wearable-sensors-identify-mobility-features-associated-with-parkinsons-disease-years-before-clinical-diagnosis/