ABSTRACT
Since conventional screening tools for assessing frailty phenotypes are resource intensive and unsuitable for routine application, efforts are underway to simplify and shorten the frailty screening protocol by using sensor-based technologies. This study explores whether machine learning combined with frailty modeling could determine the least sensor-derived features required to identify physical frailty and three key frailty phenotypes (slowness, weakness, and exhaustion). Older participants (n = 102, age = 76.54 ± 7.72 years) were fitted with five wearable sensors and completed a five times sit-to-stand test. Seventeen sensor-derived features were extracted and used for optimal feature selection based on a machine learning technique combined with frailty modeling. Mean of hip angular velocity range (indicator of slowness), mean of vertical power range (indicator of weakness), and coefficient of variation of vertical power range (indicator of exhaustion) were selected as the optimal features. A frailty model with the three optimal features had an area under the curve of 85.20%, a sensitivity of 82.70%, and a specificity of 71.09%. This study suggests that the three sensor-derived features could be used as digital biomarkers of physical frailty and phenotypes of slowness, weakness, and exhaustion. Our findings could facilitate future design of low-cost sensor-based technologies for remote physical frailty assessments via telemedicine.
Subject(s)
Frailty , Aged , Aged, 80 and over , Biomarkers , Frail Elderly , Frailty/diagnosis , Geriatric Assessment , Humans , Machine Learning , PhenotypeABSTRACT
BACKGROUND: Traditional physical frailty (PF) screening tools are resource intensive and unsuitable for remote assessment. In this study, we used five times sit-to-stand test (5×STS) with wearable sensors to determine PF and three key frailty phenotypes (slowness, weakness, and exhaustion) objectively. MATERIALS AND METHODS: Older adults (n = 102, age: 76.54 ± 7.72 y, 72% women) performed 5×STS while wearing sensors attached to the trunk and bilateral thigh and shank. Duration of 5×STS was recorded using a stopwatch. Seventeen sensor-derived variables were analyzed to determine the ability of 5×STS to distinguish PF, slowness, weakness, and exhaustion. Binary logistic regression was used, and its area under curve was calculated. RESULTS: A strong correlation was observed between sensor-based and manually-recorded 5xSTS durations (r = 0.93, P < 0.0001). Sensor-derived variables indicators of slowness (5×STS duration, hip angular velocity range, and knee angular velocity range), weakness (hip power range and knee power range), and exhaustion (coefficient of variation (CV) of hip angular velocity range, CV of vertical velocity range, and CV of vertical power range) were different between the robust group and prefrail/frail group (P < 0.05) with medium to large effect sizes (Cohen's d = 0.50-1.09). The results suggested that sensor-derived variables enable identifying PF, slowness, weakness, and exhaustion with an area under curve of 0.861, 0.865, 0.720, and 0.723, respectively. CONCLUSIONS: Our study suggests that sensor-based 5×STS can provide digital biomarkers of PF, slowness, weakness, and exhaustion. The simplicity, ease of administration in front of a camera, and safety of 5xSTS may facilitate a remote assessment of PF, slowness, weakness, and exhaustion via telemedicine.