Heart rate and blood pressure variability in normal subjects compared with data from beat-to-beat models developed from de Boer's model of the cardiovascular system

The objective of this study was to assess the ability of de Boer's model of the cardiovascular system to reproduce the heart rate and blood pressure variability observed in a range of normal subjects, and to make modifications to improve its performance. ECG, blood pressure and chest wall movement were recorded from 12 normal human subjects during controlled breathing. For each beat, systolic pressure, diastolic pressure, arterial time constant and RR interval were extracted. RR interval and systolic pressure spectral power in low and high frequency bands and the baroreflex sensitivity index, alpha, were then determined. For each subject, mean values were input to the model and the beat-to-beat output compared with the actual data for that subject. Finally, the effects of reducing the influence of baroreflex on peripheral vascular resistance and of providing separate sympathetic and vagal baroreflex sensitivities were assessed. Simulations resulted in data which were qualitatively similar to those of each subject's recording. With the modifications, the log ratio of simulated to real data improved from 7.2 to 1.5 (p = 0.003) for low frequency RR, from 0.27 to 0.55 (p = 0.011) for high frequency RR and from 8.5 to 0.9 (p = 0.003) for low frequency systolic pressure. We conclude that de Boer's model reproduces many of the characteristics of heart rate and blood pressure variability, and our modifications to baroreflex sensitivities and the feedback effect on peripheral resistance resulted in significant improvements.