Clinical Validation of Pupil Response During Walking in Parkinson's Disease

Pupil response may be a useful biomarker in Parkinson's disease (PD) due to links with autonomic function and cognitive load. However, research has focused on static tasks, missing functional demands during real-world activities like walking.

We recruited 38 people with PD and 16 healthy controls who walked for 2 min under single- and dual-task conditions while wearing mobile eye-tracking glasses (Tobii Pro Glasses 2, 100 Hz). Pupil response outcomes (velocity, size, difference between eyes) were extracted alongside gait characteristics from inertial sensors. Known groups validity compared PD and controls; convergent/divergent validity examined relationships with cognitive, visual, clinical, and gait measures.

People with PD had significantly altered pupil constriction/dilation velocity (p = 0.01), a larger difference between their left and right pupils (p = 0.04), and a larger mean and minimum pupil size (p ≤ 0.01) compared to controls during walking. Pupil response correlated with cognitive function (JLO, CLOX1, TMTB), visual acuity, disease severity (MDS-UPDRS-III), and gait characteristics in both groups. No dual-task effects were observed.

Pupil response during walking demonstrates known groups and convergent validity, indicating potential as a clinical biomarker for PD. Following this initial study, more research is required to further validate pupil response in PD (e.g., analytical validation and testing within real-world ecologically valid environments).