Day-to-day variability of cardiac autonomic regulation parameters in normal subjects.

We examined the reproducibility of day-to-day variability in cardiovascular autonomic nervous function parameters (classical clinical tests and domain analysis of heart rate variability) in four healthy men during a period of 1 working week. The results did not show any significant difference in any of the parameters over the five repeated measurements. The maximum-minimum difference as percentage of the mean was under 15% for expiration to inspiration (E/I) ratio, Valsalva ratio, tachycardia ratio, 30/15 ratio, acceleration index and brake index; about 45% for baroreflex sensitivity for systolic and diastolic blood pressure and for root mean square difference (RMSSD) of successive R-R intervals; about 65-85% for low and high frequency bands, total power and medium to high frequency ratio; and about 125% for medium frequency band. The intraclass correlation coefficient (ICC) values showed that the agreement for classical autonomic parameters (except for brake index) was good. ICC for RMSSD, baroreflex sensitivity for systolic blood pressure and the spectral estimates of heart rate variation were less good. Coefficient of variation (CV) was 4% for E/I ratio, 2% for 30/15 ratio, 6% for Valsalva, 3% for tachycardia ratio, 4% for acceleration index and 5% for brake index. CV for baroreflex sensitivity and for RMSSD was about 20%. It is concluded that the variation in baroreflex sensitivity is clearly larger than in the classical autonomic nervous function parameters. One-minute fixed pace breathing period seems to be too short to allow reproducible measurement of RMSSD and the spectral parameters of heart rate variation. Learning effect could be excluded.