Proceedings of The Physiological Society

University of Cambridge (2004) J Physiol 555P, C28


Advanced method of analysis of the beat-to-beat cardiovascular variability for the investigation of the static and dynamic properties of the baroreflex function

G. Gulli, V. Cooper, V. Claydon and R. Hainsworth

Institute of Cardiovascular Research, University of Leeds, UK

In humans the baroreceptor reflex has been studied noninvasively from the relationship between the cardiac interval and the preceding value of arterial blood pressure. This can be studied as a response to an induced perturbation or as responses of 'spontaneously' occurring pressure changes. The methods provide a measure of the gain which is essentially a static value. Further information more on the dynamic characteristics of the reflex could be obtained from the blood pressure-pulse interval relationship by examining, in addition to the gain, the reflex delay. We also wished to examine those 'non-baroreflex' sequences in which changes in pulse interval preceded rather than followed the changes in pressure. We present a self made program (matlab), which performs the analysis of RR period (ECG) and systolic arterial pressure (SAP, Finapres) changes of time series recorded during 20 min supine and 60° head-up-tilt (HUT). RR-SAP relationship analyses were accepted only when there was a linear correlation (r2 > 0.85) between 4 consecutive heart beats. In addition to the reflex gain, we determined the delay in the response of the baroreflex. For each series the lag in the baroreflex response was determined identifying the beat from which the highest correlation between RR-SAP changes was found. Likewise, analysis of lag was performed to detect the delay of SAP changes due to RR changes. By filtering the time series, analysis was performed on the entire time series, and on the LF (0.04-0.15 Hz) and HF components (0.17-0.4 Hz.) separately. The experiments were carried out after approval from the local ethics committee.

The results obtained in 9 healthy volunteers, when the entire time series was analysed, showed that the optimal baroreflex response occurred with not constant beats of delay. In supine it usually occurred with zero beats of lag. In HUT it occurred with one or more beats of lag. SAP changes due to RR changes (non baroreflex) occurred mostly with two-three beats of delay and this remained constant also in HUT. In HUT there were more 'baroreflex' and less 'non baroreflex' sequences than in supine. When the LF and HF ranges were considered separately, similar results were obtained. Preliminary results on patients with postural-related syncope and poor tolerance in an orthostatic stress test were similar, though in supine there was a prevalence of

Where applicable, experiments conform with Society ethical requirements