An examination of the effect of dual task on gait variability in Parkinson's disease and REM Sleep Behavior disorder

Technical Summary: An Examination of the Effect of Dual Task on Gait Variability in Parkinson's Disease and REM Sleep Behavior Disorder

Problem Statement
Gait variability is a recognized hallmark of Parkinson's disease (PD) and is linked to cognitive deficits and fall risk. While performing a dual task (walking while engaging in a cognitive task) typically increases gait variability, the extent of this effect in Rapid Eye Movement (REM) Sleep Behavior Disorder (RBD) remains unclear. RBD is a strong predictor of future synucleinopathies, including PD and Lewy body dementia (LBD), yet evidence regarding gait changes in RBD patients prior to phenoconversion is inconsistent. This study aimed to determine if dual-task gait variability distinguishes RBD and PD from healthy controls and if specific gait indices can predict clinical progression to Parkinsonism or dementia within an RBD cohort.

Methodology
The study utilized a prospective design involving three participant groups: 46 individuals with polysomnographically-confirmed RBD, 23 demographically matched controls (no RBD), and 14 individuals with PD. All participants were cognitively intact, aged 45 or older, and free of major psychiatric or neurological comorbidities.

• Gait Assessment: Participants underwent standardized gait analysis using a Zeno walkway. They completed five trials in random order: preferred speed walking, fast speed walking, dual-task walking (preferred speed while listing items in a category), tandem walking, and backward walking.
• Dual Task Cost: The primary metric for gait interference was "dual task cost," calculated as: (Normal Walking – Dual Task Walking) / Normal Walking * 100.
• Parameters Analyzed: The study evaluated the Enhanced Gait Variability Index (eGVI), Functional Ambulation Performance (FAP), velocity, step length, cadence, base of support (width), and double support time.
• Longitudinal Follow-up: Medical records were surveilled for three years post-enrollment. RBD participants were categorized as "converters" (those who developed PD or LBD) or "stable" (those who did not).
• Statistical Analysis: Group differences (Control, RBD, PD) were analyzed via ANCOVA with age and testing location as covariates. Longitudinal comparisons between RBD converters and non-converters utilized Welch's t-tests, with a secondary analysis controlling for age differences by excluding participants under 60 in the non-converter group.

Key Results

• PD vs. Controls/RBD: The PD group exhibited significantly greater eGVI (gait variability) under both normal and dual-task conditions compared to both RBD and control groups. PD participants also demonstrated significantly higher dual-task costs for FAP, cadence, base of support width, and double support time. Notably, while control and RBD groups did not significantly alter step width during dual tasks, the PD group widened their base of support by 26%.
• RBD vs. Controls: At baseline, there were no significant differences in gait variability or dual-task costs between the full RBD group and the control group. In fact, the RBD group showed lower dual-task costs for several parameters (FAP, step length, cadence, double support time) compared to controls, suggesting the RBD cohort may have been physiologically healthier or that study power was insufficient to detect subtle baseline differences.
• RBD Converters vs. Stable RBD: Among the RBD cohort, 6 participants converted to a neurodegenerative disorder (PD or LBD) over the three-year follow-up. Compared to stable RBD participants, these converters exhibited significantly greater dual-task costs for FAP, velocity, cadence, and double support time at baseline.
  • The gait profile of converters closely resembled that of the PD group, characterized by reduced velocity and cadence and increased double support time during dual tasks.
  • A distinct finding was that converters paradoxically adopted a narrower base of support during dual tasks, whereas stable RBD participants maintained a normal width adjustment.
  • Converters also showed a higher incidence of olfactory loss, urinary dysfunction, erectile dysfunction, and diabetes compared to stable RBD participants.

Significance and Claims
The paper concludes that while baseline gait variability does not necessarily distinguish RBD patients from healthy controls, increased gait variability during dual-task conditions may serve as a biomarker for imminent phenoconversion in RBD patients. Specifically, the study suggests that RBD patients who are about to convert to Parkinsonism or dementia exhibit a gait profile under dual-task stress that mirrors that of established PD patients.

The authors modestly claim that these kinematic changes (reductions in velocity/cadence and increases in double support time during dual tasks) could help identify RBD patients at risk of developing synucleinopathies before clinical motor symptoms appear. The study also validates the use of the Enhanced Gait Variability Index (eGVI) as a sensitive measure for detecting gait variability in PD.

Limitations Acknowledged
The authors note several limitations, including a relatively small control group compared to prior literature, the unexpected finding of higher gait variability in controls compared to the full RBD group (which may have obscured baseline differences), and the fact that all PD participants were tested in a different laboratory location than the initial RBD/control cohort, potentially introducing environmental variables. Despite these factors, the study posits that dual-task gait analysis offers a potential window into the pre-clinical progression of synucleinopathies.