Multiscale Analysis of Heart Rate Variability in Non-Stationary Environments

Jianbo Gao,Brian M. Gurbaxani,Jing Hu,Keri J. Heilman,Vincent A. Emanuele,Greg F. Lewis,Maria Davila,Elizabeth R. Unger,Jin-Mann S. Lin

DOI: https://doi.org/10.3389/fphys.2013.00119

IF: 4

2013-01-01

Frontiers in Physiology

Abstract:Heart rate variability (HRV) is highly non-stationary, even if no perturbing influences can be identified during the recording of the data. The non-stationarity becomes more profound when HRV data are measured in intrinsically non-stationary environments, such as social stress. In general, HRV data measured in such situations are more difficult to analyze than those measured in constant environments. In this paper, we analyze HRV data measured during a social stress test using two multiscale approaches, the adaptive fractal analysis (AFA) and scale-dependent Lyapunov exponent (SDLE), for the purpose of uncovering differences in HRV between chronic fatigue syndrome (CFS) patients and their matched-controls. CFS is a debilitating, heterogeneous illness with no known biomarker. HRV has shown some promise recently as a non-invasive measure of subtle physiological disturbances and trauma that are otherwise difficult to assess. If the HRV in persons with CFS are significantly different from their healthy controls, then certain cardiac irregularities may constitute good candidate biomarkers for CFS. Our multiscale analyses show that there are notable differences in HRV between CFS and their matched controls before a social stress test, but these differences seem to diminish during the test. These analyses illustrate that the two employed multiscale approaches could be useful for the analysis of HRV measured in various environments, both stationary and non-stationary.