Day-by-day variability of spontaneous baroreflex sensitivity measurements: implications for their reliability in clinical and research applications

Objective To assess day-by-day variability of spontaneous baroreflex sensitivity (BRS) measurements, providing implications for their reliability in clinical and research studies. Methods Forty-four healthy volunteers and 57 patients with previous myocardial infarction (MI) underwent an 8+8 min (spontaneous+paced breathing) recording of ECG and noninvasive arterial pressure on two consecutive days. BRS was computed according to the sequence method (BRS-seq), the original and modified transfer function method (BRS-TF and BRS-TF _mod) and the alpha method (BRS-αLF and BRS-αHF). Absolute and relative reliability were assessed by the 95% limits of random variation (LoV) and by the intraclass correlation coefficient (ICC), respectively. The sample size needed to detect a clinically relevant change was also estimated. Results In healthy volunteers during spontaneous breathing, BRS-seq, BRS-TF, BRS-αLF and BRS-αHF could not be measured in 18, 3, 2 and 2% of recordings, respectively. By definition, BRS-TF _mod could always be measured. The 95% LoV indicated that individual day-by-day changes may range from -50% to +101% for BRS-TF _mod (best case) and from -58% to +135% for BRS-αHF (worst case). The ICC ranged from 0.70 (BRS-seq) to 0.76 (BRS-TF _mod). The sample size varied from 56 (BRS-TF _mod) to 80 (BRS-seq). In MI patients, measurability was lower whereas reliability indexes were similar. Results during paced breathing were similar. Conclusion Day-by-day variability should be taken into account when using spontaneous BRS measurements to detect treatment effects in individual patients. The observed substantial to good relative reliability, as assessed by the ICC, indicates that spontaneous BRS measurements are suitable to detect differences between individuals, which is a prerequisite for proper diagnosis and prognosis.