Gait and balance features in Parkinson’s disease

Introduction: Parkinson’s disease (PD) is characterised by gait, balance, and cognitive impairments. These symptoms contribute to reduced mobility and increased risk of falls, substantially affecting the quality of life. Gait impairments are often reflected in increased step timing, greater gait variability, and postural instability (Ferreira et al., 2019). Wearable inertial measurement units (IMUs) provide unobtrusive and objective measures of walking and balance and may capture clinically meaningful features of disease severity. Cognitive impairment constitutes a contributing factor of gait and balance deterioration evidenced by findings that patients with the postural instability and gait difficulty (PIGD) phenotype are at increased risk of cognitive decline (Kelly et al., 2015). This study aims to evaluate the relationship between IMU-derived kinematic features and clinical scores of mobility and cognition in people with PD by analysing data collected by the Oxford Quantification in Parkinsonism (OxQUIP) study.Methods: We analysed data from the OxQUIP study (https://clinicaltrials.gov/study/NCT04139551) from 20 PD participants who completed a 2-minute walking task and a sway task where participants were instructed to stand as still as possible for 30 seconds whilst wearing six IMUs (2 feet, 2 wrist, sternum and lumbar spine). 128 gait and sway features were calculated and were used in this analysis. Patients clinical motor and cognition symptoms were assessed using the Movement Disorders Society Unified Parkinson’s Disease Rating Scale part III items 3.3-3.18 (MDS-UPDRS-III) and the Montreal Cognitive Assessment (MoCA), respectively. To examine the associations between gait and balance kinematics and clinical scores (MDS-UPDRS-III and MoCA), exploratory Spearman’s rank correlations were conducted.Results: All kinematic features from the walking task that were significantly associated with the clinical symptom severity (in terms of MDS-UPDRS-III) reflected gait variability. Variability measures quantify fluctuations in a movement parameter across repeated steps during the task, using the standard deviation. Balance features related to postural stability including jerk and path length of the L5 sensor were also significantly correlated with MDS-UPDRS-III scores. No walking or sway kinematic features were significantly correlated with MoCA.Conclusions: Variability measures of gait, referring to the standard deviation of double limb support, single limb support, swing, stance and arm swing velocity and postural stability measures, were associated with motor severity in PD but not global cognition. These findings suggest that variability-based gait and postural stability features may be particularly sensitive markers of motor impairment and links with previous work showing an association between gait variability and falls in PD (Sotirakis et al., 2024). We are currently working to combine these measures with other modalities, such as eye movements and speech, to aid development of sensitive biomarkers for disease severity and cognitive impairment in PD.