Methodology of spontaneous baroreflex relationship assessed by surrogate data analysis

We tested the hypothesis that the spontaneous beat-by-beat interactions of systolic blood pressure (SBP) and R-R interval reflected true baroreflex events rather than chance interactions. Original data sets of 1,024 heartbeats obtained in seated rest from six healthy subjects [R-R interval = 953 +/- 94 (+/- SE) ms] were compared with isospectral [generated by a windowed (inverse) Fourier transform with phase randomization] and isodistribution (data points randomly shuffled) surrogate data sets. The isospectral data set was used to test for random phase relationships, and the isodistribution data set was used for effects of white noise between SBP and R-R interval. Spontaneous baroreflex sequences were defined as three or more beats in which SBP and the R-R interval of the same (lag 0), next (lag 1), or next following (lag 2) beat changed in the same direction. The total number of baroreflex sequences in the original data was significantly greater than the surrogates (P < 0.001). In the original data, there were significantly (P < 0.001) more lag 0 than lag 1 or lag 2 baroreflex sequences. Therefore, these results indicated that spontaneous baroreflex sequences represented physiological rather than chance interactions and that baroreflex responses can occur within the same beat.