Murdoch University Research Repository

Welcome to the Murdoch University Research Repository

The Murdoch University Research Repository is an open access digital collection of research
created by Murdoch University staff, researchers and postgraduate students.

Learn more

P2‐226: Continuous monitoring of gait: What can it tell us about dementia?

McArdle, R., Galna, B.ORCID: 0000-0002-5890-1894, Thomas, A. and Rochester, L. (2018) P2‐226: Continuous monitoring of gait: What can it tell us about dementia? Alzheimer's and Dementia, 14 (7S_Part_14). P754-P755.

Link to Published Version: https://doi.org/10.1016/j.jalz.2018.06.913
*Subscription may be required

Abstract

Background
Gait impairment has been associated with cognitive impairment and dementia, and may reflect the conditions-specific neurodegeneration. Lab-based gait analysis is useful in predicting and aiding dementia diagnosis. However, gait laboratories are expensive and require specialist knowledge. Advances in technology have allowed body-worn sensors to provide similar information regarding gait impairment and walking activity in an individual's home and everyday environment. Continuous monitoring of gait impairment and activity is important for identification and monitoring of different types of dementia. This study aimed to assess both gait impairment and walking activity differences across dementia subtypes and controls.

Methods
87 participants were recruited across three groups; 33 Alzheimer's Disease (AD;(mean±sd) Age: 78±7; MMSE: 23±4), 27 Lewy Body Dementia (LBD; Age: 78±7; MMSE: 24±3) and 27 controls (Age:74±9; MMSE: 29±1). Dementia subtypes ranged from mild cognitive impairment to moderate dementia. A tri-axial accelometer (Axivity AX3) recorded data pertaining to gait and walking activity over 7 days. One way ANOVAs and non-parametric equivalents assessed group differences.

Results
Preliminary results report significant gait impairments in both disease groups compared to controls for pace (step velocity, step length; (p ≤ .05)) and asymmetry (step, stance and swing time asymmetry (p ≤ .05); Figure 1). LBD alone were more variable in gait (step, stance, swing time and step velocity variability; (p ≤ .05)) compared to both AD and controls. People with LBD were also more variable than AD for step velocity (p = .021) and step length variability (p= .041). LBD spent less time walking compared to controls and AD (p ≤ .05; Figure 2). AD took fewer average steps per day compared to controls (p=.019). LBD and AD were less variable in their walking bouts, and LBD performed a greater proportion of shorter walking bouts (higher alpha) compared to controls.

Conclusions
Body-worn sensors provide an individualised representation of gait with information about discrete characteristics of gait impairment and changes in walking activity across all stages of cognitive impairment. This is important for both diagnostic and interventional purposes. Future research should aim to establish effects of environmental context on gait in free-living conditions to refine clinical interpretation.

Item Type: Journal Article
Publisher: Wiley
URI: http://researchrepository.murdoch.edu.au/id/eprint/62768
Item Control Page Item Control Page