Alireza Ardalany Moghadam

Background: Gait impairment in Parkinson's disease (PD) is multifactorial and may become more pronounced under perceptual–cognitive challenge. Virtual reality (VR) provides an altered locomotor context, yet its effects on both conventional magnitude‐based Center of pressure (CoP) metrics and higher‐order structural organization remain unclear. This study compared linear, nonlinear, and topological characteristics of stance‐phase CoP trajectories between individuals with PD and healthy older adults during normal and VR walking, with particular emphasis on whether VR elicited group‐specific differences in topological organization. Methods: Fifteen individuals with idiopathic PD and 17 age‐matched controls performed overground walking under normal gait (NG) and VR conditions. Stance‐phase CoP trajectories were analyzed using linear magnitude measures (path length, velocity, root mean square, ellipse area), sample entropy (SampEn), and persistent homology–derived persistence entropy (H0 and H1). Mixed‐factor repeated‐measures ANOVA models assessed Group and Task effects. Results: VR increased CoP path length ( p = 0.005) and SampEn ( p = 0.015) across groups, while H0 persistence entropy decreased ( p = 0.002). H1 persistence entropy showed a significant Group × Task interaction ( p = 0.025), with a between‐group difference during VR but not normal walking. No Group × Task interactions were found for conventional CoP measures. Stance duration was longer in PD ( p = 0.014). Conclusions: The primary analysis showed a group‐dependent H1 persistence‐entropy response across normal and VR walking. However, because zero entropy represented both zero‐ and one‐feature diagrams and robustness across alternative point‐cloud densities and persistence thresholds was not examined, this finding should be considered exploratory. Conventional CoP measures did not show diagnosis‐specific responses to VR. H1 persistence entropy may therefore provide complementary information about CoP trajectory organization, although its physiological and clinical significance requires further investigation.

Background: Ground reaction force (GRF) alterations are reported in Parkinson’s disease (PD), but most kinetic measures depend on walking speed. Since individuals with PD walk more slowly, apparent differences may reflect speed rather than disease‐specific biomechanics. Virtual reality (VR) may further modify gait kinetics beyond speed effects. Objective: To determine whether between‐group differences in GRF waveforms during overground walking persist after accounting for walking speed and whether VR induces speed‐independent kinetic modulation in individuals with PD. Methods: 14 adults with mild‐to‐moderate PD (Hoehn and Yahr stage II–III) and seventeen age‐matched controls performed overground walking under normal and immersive VR conditions. Three‐dimensional kinematics (100 Hz) and GRFs (1000 Hz) were recorded. Discrete spatiotemporal and kinetic variables were analyzed using mixed‐design repeated‐measures ANOVA. Time‐continuous GRF waveforms were examined using mixed‐model statistical parametric mapping (SPM) ANCOVA with walking speed entered as a covariate. Results: Individuals with PD walked significantly slower than controls in both conditions. VR increased stride length and modulated several discrete GRF variables, including anterior–posterior and mediolateral components. However, waveform‐level SPM analyses revealed no persistent group or group × task effects once walking speed was included as a covariate. In contrast, walking speed demonstrated widespread and robust associations with vertical and mediolateral GRF trajectories across multiple stance‐phase intervals. Conclusion: During comfortable overground walking in mild‐to‐moderate PD, many apparent kinetic differences are largely speed‐coupled rather than clearly speed‐independent. These findings underscore the necessity of speed‐aware modeling when interpreting gait kinetics and evaluating VR‐based paradigms in Parkinson’s disease.