Estimation of baroreflex sensitivity by the bivariate phase rectified signal averaging method: a comparison with the...

A novel technique to assess spontaneous baroreflex sensitivity (BRS) by bivariate phase-rectified signal averaging (PRSA-BRS) has been recently proposed and its independent prognostic power demonstrated. This method, however, has never been compared with the phenyleprine test (Phe-BRS), commonly regarded as the reference method in clinical and research applications.

In 192 heart failure (HF) and 41 post-myocardial infarction (post-MI) patients we compared PRSA-BRS with Phe-BRS, assessing both association and agreement.

Phe-BRS and PRSA-BRS were (mean  ±  SD) 4.8  ±  5.0 (range:  -3.8,25.0) and 1.2  ±  1.5 (-2.1,6.9) ms mmHg^-1 in HF (p  <  0.0001), and 5.0  ±  3.8 (-1.2,12.5) and 0.8  ±  1.7 (-2.0,6.9) ms mmHg^-1 in post-MI patients (p  =  0.001). Moderate association was observed (r  =  0.53, p  <  0.0001 and r  =  0.43, p  =  0.004 in HF and post-MI, respectively). The vast majority (86% in HF and 90% in post-MI) of PRSA-BRS measurements were smaller than corresponding Phe-BRS values. The difference between PRSA-BRS and Phe-BRS was strongly dependent on the magnitude of BRS, with a trend towards more negative differences as BRS increased. Negative PRSA-BRS values were observed in 15% of HF and in 37% of post-MI patients, whereas negative Phe-BRS values were observed in 8% of HF and 5% of post-MI patients.

Although the association with Phe-BRS suggests that PRSA-BRS contains relevant information about cardiac autonomic control and reflects the strength of the baroreceptor-heart rate reflex, the marked disagreement between the two measurements indicates that PRSA-BRS measurements cannot be taken as estimates of BRS. Many factors may account for the observed lack of agreement: the different physiological conditions under which Phe-BRS and PRSA-BRS are measured, the inclusion of non-baroreflex mediated components of RR-intervals in PRSA-BRS and some computational aspects related to the normalization of PRSA-BRS values.