Comparison between spectral analysis and the phenylephrine method for the assessment of baroreflex sensitivity in chronic heart failure

Baroreflex sensitivity assessed by means of the phenylephrine test plays a prognostic role in patients with previous myocardial infarction, but the need for drug injection limits the use of this technique. Recently, several non-invasive methods based on spectral analysis of systolic arterial pressure and heart period have been proposed, but their agreement with the phenylephrine test has not been investigated in patients with heart failure. The two methods (phenylephrine test and spectral analysis) were compared in a group of 49 patients with chronic congestive heart failure both at rest and during controlled breathing. The linear correlation and the limits of agreement between the phenylephrine test slope and the alpha-index [alpha(c); corrected by the coherence function between the interbeat interval (RR interval) and systolic arterial pressure] were evaluated. Only 16 patients had a measurable alpha-index at rest in both the low-frequency (LF) and high-frequency (HF) bands; the alpha(c)-index allowed measurements in all patients. It correlated moderately with the phenylephrine test slope at rest (r = 0.71 and P < 0.001 in LF; r = 0.57 and P < 0.001 in HF) and during controlled breathing (r = 0.51 and P < 0.001 in LF; r = 0.63 and P < 0.001 in HF). Multivariate regression analysis showed that only alpha(c)LF during rest and alpha(c)HF during controlled breathing contributed significantly to baroreflex gain estimation. However, the agreement between methods was weak; the normalized limits of agreement and bias were - 162 to 243% (0.46 ms/mmHg) for alpha(c)LF and - 185 to 151% (-0.99 ms/mmHg) for alpha(c)HF. Thus the comparison between baroreflex sensitivity measurements obtained by the phenylephrine test and spectral analysis showed a moderate correlation between the two methods; however, despite the linear association, a consistent lack of agreement between the two techniques was found. Because both systematic and random factors contribute to the difference, these two techniques cannot be considered as alternatives for the assessment of heart failure.