Differentiating idiopathic Parkinson's disease from vascular parkinsonism and essential tremor plus with spatiotemporal variables of gait

Abstract

Diagnostic accuracy to differentiate Parkinson’s Disease (PD) from other neurological disorders by movement disorder specialists is >85% for longer duration (>5years) and medication-responsive people with PD [1]. However, this accuracy remains suboptimal (<55%) in the early stages of disease, i.e. when symptoms may be very subtle [1]. Therefore, it’s of paramount importance to identify diagnostic markers which could help in timely and accurate diagnosis and in clinical management [2]. Gait may be a potential marker since its impairments appear in the very early disease, may precede the onset of overt motor signs and evolve more rapidly than other motor features [2,3]. The purpose of this work is to assess whether gait variables, known to correctly classify PD from healthy subjects [4], are capable to differentiate PD from other Parkinsonism in subjects with an onset of symptoms from less than 3 years. 17 drug-naïve individuals with PD (mean age 70.0±7.4 yrs), 9 with vascular parkinsonism (VP, mean age 76.6±5.5 yrs) and 9 with essential tremor plus (ET, mean age 65.9±8.1 yrs) were enrolled. Eight spatiotemporal variables of gait [4] were acquired using a GAITRite® electronic walkway under comfortable walking. To assess the diagnostic performance of each variable in discriminating PD from VP and ET, the area under the receiver operating characteristic curve (AUC) and its 95% Confidence Interval (95% CI) were assessed. Participants with PD were not different from VP and ET regarding sex (p=0.86 and p=0.13, respectively) and age (0.07 and 0.06, respectively). Of the gait variables considered, gait speed, step time and step time asymmetry showed a moderate ability in differentiating PD from the other movement disorders (Table 1). In addition, double support time and swing time CV were able to differentiate individuals with PD from those with ET. The gait model, obtained combining the 8 gait variables, showed the highest ability in discriminating PD from VP (AUC=0.90) and ET (AUC=0.86), with a diagnostic accuracy of 88.5% and 80.8%, respectively. This study found that spatiotemporal gait variables, assessed with a sensorized mat, can be used to differentiate PD from VAS and ET. In particular, the model of gait obtained combining several variables had a good diagnostic accuracy in the first ages of disease. This provides a foundation for future research to investigate the use of gait assessment as a clinical tool to aid diagnostic accuracy of PD subtypes, allowing the correct treatment and care to be applied [5].