RESUMO
Atrial fibrillation (AF) associates with disability and frailty. Aim of this study was to evaluate in older AF patients, using artificial intelligence (AI), the relations between geriatric tools and daily standing and resting periods. We enrolled thirty-one > 65 years patients undergoing electrical cardioversion of AF (age: 79 ± 6 years; women: 41.9%; CHA2DS2-VASc: 3.7 ± 1.2; MMSE: 27.7 ± 2.7; GDS: 3.0 ± 2.8). The data of the first day following the procedure were analyzed using machine-learning techniques in a specifically designed cloud platform. Standing, activity, time (582 ± 139 min) was directly associated with MMSE and inversely with GDS. Sleep length was 472 ± 230 min. Light sleep, the longer resting phase, was inversely related to GDS. The Chest Effort Index, a measure of obstructive sleep apnea, grew with GDS. In conclusion, AI devices can be routinely used in improving older subjects' evaluation. A correlation exists between standing time, MMSE, and depressive symptoms. GDS associates to length and quality of sleep.