Database for Prevalence and Determinants of Frailty in the Elderly with Quantifying Functional Mobility

Abstract

Frailty is a common condition in older adults, characterized, among other things, by impairments in gait and movement patterns. The proposed FRAILPOL repository addresses the critical gap in geriatric research by offering a comprehensive, open-access, five body-worn inertial sensors (ankles, wrists, and back of sacrum) signals recorded during the Time Up and Go test of 668 participants, community-dwelling older adults. The gait data, as well as the stride-based spatio-temporal parameters along with demographic and health-related information, including cognitive health data, have been grouped according to established clinical criteria into three classes (robust, pre-frailty, and frailty). The technical verification includes classification by reporting results for both binary (robust, frailty) and multi-class (robust, pre-frailty, frailty) classification using classical machine learning models with acceptable accuracy.