Reproducibility of Heart Rate Variability Revealed by Repeated Measurements during and after Hemodialysis.

BACKGROUND/AIMS: Trajectory of heart rate variability (HRV) represents a noninvasive real-time measure of autonomous nervous system (ANS) and carries the capability of providing new insights into the hemodynamic compensation reserve during hemodialysis (HD). However, studies on HRV reproducibility during HD are scarce and did not refer to different reading periods. In this observational study, we aimed to establish the best suited and most reliable and reproducible HRV index in routine HD treatments including different reading rates.

METHODS

HRV was characterized by standardized mathematical variation expressions of R/R' intervals: SD of all R/R' intervals (ms), square root of the root mean square of the sum of all differences between adjacent R/R' intervals (ms), percentage of consecutive R/R' intervals that differ by >50 ms (%), low-frequency spectral analysis HRV (LF, expressing sympathetic activity), and high-frequency HRV (HF, expressing parasympathetic activity). To compare robustness of these HRV indices during HD procedures, we compared HRV indices means between different HD sessions and controlled for association with clinical parameters.

RESULTS

In 72 HD treatments of 34 patients, we detected the highest reproducibility (89%) of HRV measures when analyzing the low-frequency to high-frequency (LF/HF) ratio in long-term (3 h) readings. Long-term LF/HF was able to discriminate -between patients with and without heart failure NYHA classes ≥3 (p = 0.009) and type 2 diabetes (p = 0.023). We were unable to study relationships between ANS and intradialytic complications because they did not appear in our cohort. Short-term readings of HRV indices did not show any significance of pattern change during HD.

CONCLUSION

In summary, our data provide evidence for high robustness of long-term LF/HF in analyzing HRV in HD patients using future automated monitoring systems. For short-term analysis, mathematical real-time analysis must evolve.