Predicting Frailty Using Motor and Heart Rate Interconnection: Application of Convergent Cross Mapping to Identify Autonomic Dysfunction

Frailty is a common syndrome in older adults, marked by low physiological reserve, which can lead to an increased vulnerability to stressors. We utilized a validated upper-extremity function (UEF) test that involves 20-second rapid arm flexion to assess motor and cardiac performance. One hundred and seventy two older adults (≥65 years) were recruited and classified as non-frail, pre-frail, and frail using the Fried phenotype. For UEF, wearable motion sensors captured elbow angular velocity, and heart rate (HR) was continuously recorded using an ECG wearable recording system. The dynamic interconnection between angular displacement and HR was assessed using convergent cross-mapping (CCM). ANCOVA (adjusted for age, sex, and BMI) tested differences between the three frailty groups; effect sizes were reported. Across groups, HR increase differed significantly with smaller changes with frailty (p < 0.01; effect size = 0.15), HR recovery showed a trend toward group differences with smaller recoveries for frail individuals (p = 0.06; effect size = 0.19), and the HR–motor correlation decreased with frailty (p < 0.01; effect size = 0.33). This approach captures both cardiac and motor function within less than two minutes of a physical task while seated, to provide a unique tool for quick and objective assessment of frailty in a clinical setting.