Validating phase relations between cardiac and breathing cycles during sleep.

Institute for Pathophysiology, Friedrich Schiller University.
IEEE Engineering in Medicine and Biology Magazine (Impact Factor: 2.73). 01/2001; 20(2):101-6. DOI: 10.1109/51.917730
Source: PubMed

ABSTRACT The objective of this article is to investigate phase relations between heart beat and breathing cycles and their dependence on sleep states. Furthermore, the appearance of phase synchronizations between these signals is proved statistically by analyzing the phase relations between breathing and heart beat periods. The phase synchronizations between these signals are searched for by testing appropriate parameters of surrogate data with similar power spectra but randomly shuffled phase relations.

  • Source
  • Source
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Inspiratory fall of systolic blood pressure (IFSBP) is used as an index to assess the respiratory influences on non-linear dynamics of heart rate variability. In an in-vivo situation it is particularly difficult to isolate individual effects of heart rate, vascular tone, pleural pressure variation, and ventricular interdependence. A computer model of the cardiopulmonary system was adapted to this problem, relating mechanisms such as baroreflex regulation of heart rate in response to respiratory oscillations. The model provided time-course simulations of hemodynamics by numerically integrating 28 nonlinear, time-varying differential equations. Two approaches for baroreflex regulation were tested, including a simple 1st-order relationship between R-R interval (RRI) and SBP and the autoregressive moving average (ARMA) model. Experimental data were obtained retrospectively from 22 patients with chronic airway obstruction before and during breathing through an external resistance. Magnitude and phase relations between arterial pressure and pleural pressure were evaluated. The computer model provided good fits to arterial pressure waveforms: correlation coefficients (r) ranging from 0.71 to 0.96 (mean±SD: 0.87±0.06) with a simple 1st-order RRI-SBP model. It was observed that the ARMA model did not further improve the goodness of fit. However, no dominant parameter was found for phase relations.