Decomposition of Heart Rate Variability Spectrum into a Power-Law Function and a Residual Spectrum.

Jane Kuo,Cheng-Deng Kuo

doi:10.3389/fcvm.2016.00016

Abstract

The power spectral density (PSD) of heart rate variability (HRV) contains a power-law relationship that can be obtained by plotting the logarithm of PSD against the logarithm of frequency. The PSD of HRV can be decomposed mathematically into a power-law function and a residual HRV (rHRV) spectrum. Almost all rHRV measures are significantly smaller than their corresponding HRV measures except the normalized high-frequency power (nrHFP). The power-law function can be characterized by the slope and Y-intercept of linear regression. Almost all HRV measures except the normalized low-frequency power have significant correlations with the Y-intercept, while almost all rHRV measures except the total power [residual total power (rTP)] do not. Though some rHRV measures still correlate significantly with the age of the subjects, the rTP, high-frequency power (rHFP), nrHFP, and low-/high-frequency power ratio (rLHR) do not. In conclusion, the clinical significances of rHRV measures might be different from those of traditional HRV measures. The Y-intercept might be a better HRV measure for clinical use because it is independent of almost all rHRV measures. The rTP, rHFP, nrHFP, and rLHR might be more suitable for the study of age-independent autonomic nervous modulation of the subjects.

Highlights

Heart rate variability (HRV) refers to the continuous oscillation of RR intervals (RRIs) around its mean value
The slope and Y-intercept can be obtained from linear regression analysis of log(PSD) versus log(Frq) to characterize the power spectrum of heart rate variability (HRV)
We found that the residual HRV (rHRV) measures are significantly smaller than their counterparts in the traditional HRV except the normalized high-frequency power (nrHFP), which is significantly greater than the nHFP in the traditional HRV

Summary

Introduction

Heart rate variability (HRV) refers to the continuous oscillation of RR intervals (RRIs) around its mean value. Power spectrum analysis of heart rate (HR) fluctuations provides a quantitative and noninvasive means to assess the sympathetic and vagal modulations of HR [1, 2]. The existence of the linear regression line indicates that the relationship between PSD and Frq in the power spectrum of HRV can be described by a power-law function. The power-law relationship of HRV has been used as a predictor of mortality in the elderly [9], Decomposition of HRV Spectrum and the analysis of the fractal characteristics of short-term RRI dynamics can yield more powerful prognostic information than the traditional HRV measures among patients with depressed left ventricular function after AMI [10], patients with Chaga’s disease [15], and pediatric patients with multiple organ failure [16]

Objectives

Methods

Results

Discussion

Conclusion