The reproducibility of cardiac baroreceptor activity assessed non-invasively by spectral sequence techniques.

Baroreceptor sensitivity (BRS) is increasingly used as a prognostic indicator in cardiovascular disease. Traditionally it has been measured using invasive techniques with pharmacological manipulation of blood pressure (BP). With the advent of newer methods to measure pulse interval and beat-to-beat changes in BP it is now possible, using sophisticated mathematical modelling techniques, to calculate cardiac BRS non-invasively. However, there are virtually no data on the reproducibility of these newer techniques and what factors may affect the repeatability of these measurements. We studied 39 subjects, aged 22-82 years, with a supine systolic BP range 97-160 mmHg and a diastolic BP range 57-94 mmHg on two occasions between 1 week and 6 months apart. Cardiac BRS was measured by power spectral analysis using Fast Fourier Transformation (FFT), sequence analysis (using up, down and combined sequences) and from phase IV of the Valsalva manoeuvre. There was no significant difference between visits for any of the methods for measuring cardiac BRS. Mean BRS values were similar for FFT (16.7 +/- 11.2 ms/mmHg) and sequence analysis (15.8 +/- 11.4 ms/mmHg); however, results using phase IV of the Valsalva manoeuvre were significantly lower (8.1 +/- 2.9 ms/mmHg, p < 0.0001). The coefficient of variation for the five measures of cardiac BRS varied from 16.8% for Valsalva-derived values to 26.1% for 'down' sequence analysis. However, in ten subjects BRS could not be calculated from the Valsalva manoeuvre. None of the independent variables tested (including age, BP levels and time between testing) significantly influenced the degree of repeatability. In summary, there appears to be little difference between these non-invasive methods in their degree of reproducibility. These techniques would seem suitable for longitudinal studies of changes in cardiac BRS and overcome many of the problems associated with the invasive pharmacological methods.