Abstract
Subtle gait and cognitive dysfunction are common in Parkinson's disease (PD), even before most evident clinical manifestations. Such alterations can be assumed as hypothetical phenotypical and prognostic/progression markers. To compare spatiotemporal gait parameters in PD patients with three cognitive status: cognitively intact (PD-noCI), with subjective cognitive impairment (PD-SCI) and with mild cognitive impairment (PD-MCI) in order to detect subclinical gait differences. One hundred PD patients were consecutively enrolled and divided in three groups based on both the first item od MDS-UPDRS part I and an extensive neuropsychological evaluation: 41 PD-noCI, 15 PD-SCI and 44 PD-MCI. They were evaluated with gait analysis acquired in three different conditions (normal gait, motor and cognitive dual task). Spatiotemporal variables were extracted. A univariate statistical analysis (parametric ANOVA test or non-parametric Kruskal-Wallis test, as appropriate) with post-hoc analysis was carried out in order to evaluate the significant differences among the groups. In normal gait task, the three groups showed several differences, all due to the comparison between PD-MCI and PD-noCI, as disclosed by post-hoc analysis. In dual task conditions, mostly in the cognitive dual task, the three groups showed increased gait alterations that, at post-hoc analysis, mirrored the magnitude of cognitive dysfunction (PD-noCI < PD-SCI < PD-MCI). Peculiar prodromal gait patterns-especially those highlighted by cognitive dual task-could be considered possible markers to objectify self-reported symptoms-based construct, like SCI, and to early intercept subjects with different clinical evolutions and prognoses, even representing an innovative clustering/phenotyping tool for PD subtypes.