Reproducibility of three different methods of measuring baroreflex sensitivity in normal subjects

1. Baroreflex sensitivity is a useful tool for investigating cardiovascular reflexes in a number of clinical settings. Several different methods of measuring baroreflex sensitivity are available. In order to determine a clinically useful non-invasive method of measuring baroreflex sensitivity we compared two methods (spectral analysis and the Valsalva manoeuvre) with regard to reproducibility, agreement with a standard invasive method (phenylephrine infusion) and failure rate. 2. Twenty-six healthy subjects aged 22 to 63 years attended on three separate occasions for measurement of baroreflex sensitivity using the different methods. The effect of a recent head-up tilt on baroreflex sensitivity was measured. 3. Reproducibility was best for the low-frequency component of the spectral method [coefficient of variation 25.0% (range 3.5-2.4%)] and worst for the Valsalva method [coefficient of variation 29.3% (range 13.8-93.1%)]. Both non-invasive methods overestimated values compared with the phenylephrine method [bias of low-frequency component of the spectral method, 1.17 (0.38-3.6); bias of the Valsalva method, 1.13 (0.19-6.7)]. The high-frequency component of the spectral method did not agree with the phenylephrine method. 4. The spectral analysis method had the fewest failures (seven subjects with a failure on at least one occasion), and the phenylephrine method the most (16 subjects with a failure on at least one occasion). A short head-up tilt did not affect the subsequent non-invasive measurement of baroreflex sensitivity. 5. It was concluded that the low-frequency component of the spectral method was the most clinically useful non-invasive measurement of baroreflex sensitivity.