[An analysis of heart rate variability. Spectrum analysis methods and complex signal demodulation]

The analysis of heart rate variability provides valuable information in the noninvasive study of neurovegetative activity and its modifications induced by drugs, physiological maneuvers or disease processes, and in the evaluation of prognosis and risk stratification in different cardiovascular diseases. A review is made of the different methods used to study heart rate variability, and an account is given of the information provided by spectral methods and the new procedures based on the complex demodulation of the time series composed of consecutive cardiac cycles. The limitations of the spectral methods are deal with, particularly in defining the time-dependent changes in variability and their relation to clinical events. Likewise, a description is given of the ability of complex demodulation to define the time course of the oscillations into which the analyzed time series is decomposed. Complex demodulation based on the fast Fourier transform and its inverse is able to establish the instantaneous amplitude and frequency of each of the oscillations contained in the time series, separated by specific filters in the previously selected frequency bands (high: 0.15-0.40 Hz, low: 0.04-0.15 Hz, or very low: < 0.04 Hz), and from which the original signal may be reconstructed. The evaluation of the different methodological approaches, and the analysis of the causal relations between the variability modifications and clinical events will further extend the clinical relevance of the study of heart rate variability.