Automatic prediction of falls via heart rate variability and data mining in hypertensive patients: The SHARE project experience

Accidental falls in elderly is a major problem. This paper presents the preliminary results of a retrospective study investigating association between Heart Rate Variability (HRV) measures and risk of falling, analyzing 168 clinical 24- hour ECG recording from hypertensive patients, 47 of them experienced at least one fall in the three months before/after the registration. Several HRV patterns, based on 68 linear and non-linear HRV measures, were analyzed in relation to falls using advanced statistical and data mining methods. The results demonstrated that there is a significant association between a depressed HRV and the risk of falling, suggesting that a depressed HRV could be a new independent risk factor for falls with an odds ratio of 5.12 (CI 95% 1.42-18.41; p<0.01)