Sensor-based Characterization of Daily Walking: A New Paradigm in Frailty Assessment
Abstract Background Frailty is an increasingly recognized geriatric syndrome resulting in decline in reserve across multiple physiological systems. Impaired physical function is a prime indicator of frailty.Objective The goal of this study was to develop an algorithm that discriminates between frailty groups (non-frail, pre-frail, and frail) based on gait performance parameters derived from unsupervised daily physical activity (DPA).Methods DPA was acquired for 48 hours from older adults (≥65 years) using a tri-axial accelerometer motion-sensor. Purposeful continuous bouts of walking (≥60s) without pauses were identified from acceleration data. These were then used to extract qualitative measures (gait variability, gait asymmetry, and gait irregularity) and quantitative measures (total continuous walking duration and maximum number of continuous steps) to characterize gait performance. Association between frailty and gait performance parameters was assessed using multinomial logistic models with frailty as the dependent variable, and gait performance parameters along with demographic parameters as independent variables.Results 126 older adults (44 non-frail, 60 pre-frail, and 22 frail, based on the Fried index) were recruited. Step- and stride-times, frequency domain gait variability, and purposeful walking quantitative measures were significantly different between non-frail and pre-frail as well as non-frail and frail groups ( p <0.05). Using the logistic model pre-frail/frail group (vs. non-frail) was identified with 76.8% sensitivity and 80% specificity.Discussion Everyday walking characteristics were found to be associated with frailty. Along with quantitative measures of physical activity, qualitative measures are critical elements representing the early stages of frailty. In-home gait assessment offers an opportunity to screen for and monitor frailty.