Power spectral analysis of heart rate variability and baroreflex gain.

Experimental open-loop studies [1] indicate that when a sinusoidal pressure function is forced onto the arterial baroreceptors, a sinusoidal oscillation is reflexly transferred on to R-R interval and arterial pressure, with a rapid fall of gain with increasing oscillatory frequencies. Sleight et al. [2] addressed this issue in man with an elegant approach employing frequency domain analysis. In the paper 'Physioloy and pathophysiology of heart rate and blood pressure variability in humans: is power spectral analysis largely an index of baroreflex gain?' [2], these authors showed that a slow (either 0.1 or 0.2 Hz) sinusoidal neck suction, likely to stimulate carotid sinus baroreceptors, was capable of inducing a synchronous component in the spectrum of the R-R interval or systolic arterial pressure signals. The amplitude of the induced spectral component was greater at 0.1 Hz and related to the phenylephrine baroreflex slope of the three subjects (one healthy volunteer and two patients with congestive heart failure) examined during controlled respiration. Sleight et al. did not explicitly address the open loop (i.e. stimulus-eresponse) design of their study, and stated that 'LF power (and hence LF/HF ratio) is determined by and requires the baroreflex in order to generate the sympathetic oscillation at about 0.1 Hz'. This is an important point which merits some comments. There is little question that, also in closed-loop conditions, a reduced arterial pressure acting on baroreceptive reflexogenic areas can induce a sympathetic excitation, leading to a consistent increase in the LF power (in normalized units), or in the LF/HF ratio of R-R variability. This was first observed in man with tilting manoeuvres or in conscious dogs with pharmacologically induced hypotensions [3]. Moreover, chronic sino-aortic denervation in dogs abolished this reflex increase in the LF component. What we disagree with is an exclusiveness attributed to any single neural circuit in generating a given cardiovascular rhythm. In fact, the following evidence cannot be reconciled with the oversimplification inherent in the statement by Sleight et al.: 555