Asymmetric acceleration/deceleration dynamics in heart rate variability

The heart rate variability (HRV) is an important physiological signal used either to assess the risk of cardiac death or to model the cardiovascular regulatory dynamics. Asymmetries in HRV data have been observed using 2D Poincare plots, which have been linked to a non-equilibrium operation of the cardiac autonomic system. This work further explores the presence of asymmetries but in the serial correlations of the dynamics of HRV data. To this end, detrended fluctuation analysis (DFA) was used to estimate the Hurst exponent both when the heart rate is accelerating and when it is decelerating. The analysis is conducted using data collected from subjects under normal sinus rhythm (NSR), congestive heart failure (CHF) and atrial fibrillation (AF) . For the NSR cases, it was found that correlations are stronger (p<0.05) when the heart rate is accelerating than when it is decelerating over different scales in the range 20-40 beats. In contrast, the opposite behavior was detected for the CHF and AF patients. Possible links between asymmetric correlations in the dynamics and the mechanisms controlling the operation of the heart rate are discussed, as well as their implications for modeling the cardiovascular regulatory dynamics.