Application of HRV-CD for estimation of life expectancy in various clinical disorders

Abstract

Background Low heart rate variability (HRV) was found in various medical conditions including heart failure and acute myocardial infarction. Decreased HRV in these conditions predicted poor prognosis. Methods HRV was estimated in 133 unselected inpatients with relevant clinical bedside conditions by non-linear analysis derived from chaos theory, which calculates the correlation dimension (CD) of the cardiac electrophysiologic system (HRV-CD). Results Mean HRV-CD in the entire group was 3.75 ± 0.45. Heart failure, coronary artery disease, cardiac arrhythmia, low serum potassium, renal dysfunction, and diabetes mellitus were significantly associated with reduced HRV-CD compared to their counterparts [3.6 vs. 3.9 ( P < .001), 3.65 vs. 3.87 ( P = .005), 3.58 vs. 3.8 ( P = .01), 3.38 vs. 3.81 ( P = .02), 3.59 vs. 3.8 ( P = .04), and 3.66 vs. 3.82 ( P = .04), respectively]. Stepwise logistic regression showed heart failure to be the condition most significantly associated with low HRV-CD (odds ratio 4.2, 95% confidence interval 1.90–9.28, P < .001). In the entire group, decreased HRV-CD (≤ 3.75 vs. > 3.75) was associated with lower survival ( P = .01). Mortality of diabetic patients with HRV-CD ≤ 3.75 exceeded the mortality in patients with HRV-CD > 3.75 ( P = .02). Heart failure, renal dysfunction or age over 70 combined with HRV-CD ≤ 3.75 also appeared to be associated with augmented mortality. Conclusions Diminished HRV-CD is associated with heart failure, coronary artery disease, cardiac arrhythmia, renal dysfunction, diabetes mellitus and low serum potassium. Among the latter, heart failure is most significantly associated with decreased HRV-CD. Decreased HRV-CD values, especially in diabetics, are also associated with lower survival.