Insights into gait disorders: Walking variability using phase plot analysis, Parkinson’s disease Patrick Esser a, *, Helen Dawes a,b,1 , Johnny Collett a,2 , Ken Howells a,3 a Movement Science Group, Oxford Brookes University, Gipsy Lane Campus, Headington, Oxford OX3 0BP, UK b Department of Clinical Neurology, University of Oxford, Oxford, UK 1. Introduction Parkinson’s disease (PD) is a progressive disorder of the central nervous system, presenting with movement impairments includ- ing resting tremor, short slow steps [1] and therefore a decreased CoM movement [2]. Furthermore an increase in variability of spatio-temporal parameters such as stride length and step time can be observed [3]. Spatio-temporal variability, such as stride-to- stride time and stride length has been explored using techniques such as fractal analysis, coefficient of variation analysis [4] or variability analysis based on complicated wavelet analysis [5]. These techniques utilize a single value for a spatio-temporal event that occurs during the gait cycle. Therefore multiple gait cycles are required to gather the data demanding a long walking distance and can therefore be seen as stressful or even impossible for participants, especially those with more severe gait disabilities. This limits the clinical application of the technique. Centre of mass (CoM) movement measured at a high sample frequency can provide a quick way of gathering larger amounts of data over relative few steps. Inspired by Poincare ´ analysis [6], we developed phase plot analysis using a single reference marker to describe vertical CoM movement. This simple approach requires only a small amount of steps utilising all individual data points of the vertical CoM excursion (measured at 100 Hz) which can then be employed in analysis over a short walking distance. This study explores the use of the phase plot method applied to human gait over 10 m in order to compare gait variability in people with PD compared to TDA. Furthermore this study explored the fidelity of a simple phase plot analysis based on Poincare ´ methods in order to detect step-to-step variability using theoretically generated sine waves. 2. Methods 2.1. Participants Fourteen people with consultant diagnosed PD, living in Oxfordshire, between the age of 30 and 75 were recruited under National Health Service ethical approval at Oxford Brookes University. A further ten typical developed adults (TDAs) of the Gait & Posture xxx (2013) xxx–xxx A R T I C L E I N F O Article history: Received 10 February 2012 Received in revised form 11 February 2013 Accepted 20 February 2013 Keywords: Variability Gait Parkinson’s disease Accelerometry A B S T R A C T Gait variability may have greater utility than spatio-temporal parameters and can, be an indication for risk of falling in people with Parkinson’s disease (PD). Current methods rely on prolonged data collection in order to obtain large datasets which may be demanding to obtain. We set out to explore a phase plot variability analysis to differentiate typically developed adults (TDAs) from PD obtained from two 10 m walks. Fourteen people with PD and good mobility (Rivermead Mobility Index 8) and ten aged matched TDA were recruited and walked over 10-m at self-selected walking speed. An inertial measurement unit was placed over the projected centre of mass (CoM) sampling at 100 Hz. Vertical CoM excursion was derived to determine modelled spatiotemporal data after which the phase plot analysis was applied producing a cloud of datapoints. SD A described the spread and SD B the width of the cloud with b the angular vector of the data points. The ratio (8) was defined as SD A : SD B . Cadence (p = .342) and stride length (p = .615) did not show a significance between TDA and PD. A difference was found for walking speed (p = .041). Furthermore a significant difference was found for b (p = .010), SD A (p = .004) other than SD B (p = .385) or ratio 8 (p = .830). Two sequential 10-m walks showed no difference in PD for cadence (p = .193), stride length (p = .683), walking speed (p = .684) and b (p = .194), SD A (p = .051), SD B (p = .145) or 8 (p = .226). The proposed phase plot analysis, performed on CoM motion could be used to reliably differentiate PD from TDA over a 10-m walk. ß 2013 Elsevier B.V. All rights reserved. * Corresponding author. Tel.: +44 01865 483 272. E-mail addresses: pesser@brookes.ac.uk (P. Esser), hdawes@brookes.ac.uk (H. Dawes), jcollett@brookes.ac.uk (J. Collett), kfhowells@brookes.ac.uk (K. Howells). 1 Tel.: +44 01865 483 293. 2 Tel: +44 01865 483 272. 3 Tel: +44 01865 483 256. G Model GAIPOS-3850; No. of Pages 5 Please cite this article in press as: Esser P, et al. Insights into gait disorders: Walking variability using phase plot analysis, Parkinson’s disease. Gait Posture (2013), http://dx.doi.org/http://dx.doi.org/10.1016/j.gaitpost.2013.02.016 Contents lists available at SciVerse SciVerse ScienceDirect Gait & Posture jo u rn al h om ep age: ww w.els evier.c o m/lo c ate/g aitp os t 0966-6362/$ – see front matter ß 2013 Elsevier B.V. All rights reserved. http://dx.doi.org/http://dx.doi.org/10.1016/j.gaitpost.2013.02.016