Insights into gait disorders: Walking variability using phase plot analysis, Huntington's disease

•We investigated gait variability in HD using traditional and non-linear methods.•Phase plot analysis revealed subtle changes in premanifest HD walking.•Quantified changes to the phase plots were associated with disease severity.•Phase plots data can be obtained during standard clinical walking tests.

Huntington's disease (HD) is a progressive inherited neurodegenerative disorder. Identifying sensitive methodologies to quantitatively measure early motor changes have been difficult to develop. This exploratory observational study investigated gait variability and symmetry in HD using phase plot analysis. We measured the walking of 22 controls and 35 HD gene carriers (7 premanifest (PreHD)), 16 early/mid (HD1) and 12 late stage (HD2) in Oxford and Cardiff, UK. The unified Huntington's disease rating scale-total motor scores (UHDRS-TMS) and disease burden scores (DBS) were used to quantify disease severity. Data was collected during a clinical walk test (8.8 or 10 m) using an inertial measurement unit attached to the trunk. The 6 middle strides were used to calculate gait variability determined by spatiotemporal parameters (co-efficient of variation (CoV)) and phase plot analysis. Phase plots considered the variability in consecutive wave forms from vertical movement and were quantified by SDA (spatiotemporal variability), SDB (temporal variability), ratio ∀ (ratio SDA:SDB) and Δangleβ (symmetry). Step time CoV was greater in manifest HD (p < 0.01, both manifest groups) than controls, as was stride length CoV for HD2 (p < 0.01). No differences were found in spatiotemporal variability between PreHD and controls (p > 0.05). Phase plot analysis identified differences between manifest HD and controls for SDB, Ratio ∀ and Δangle (all p < 0.01, both manifest groups). Furthermore Ratio ∀ was smaller in PreHD compared with controls (p < 0.01). Ratio ∀ also produced the strongest correlation with UHDRS-TMS (r = −0.61, p < 0.01) and was correlated with DBS (r = −0.42, p = 0.02). Phase plot analysis may be a sensitive method of detecting gait changes in HD and can be performed quickly during clinical walking tests.