Continuous wavelet transform based processing for estimating the power spectrum content of heart rate variability during hemodiafiltration

Power spectral analysis of heart rate variability (HRV) offers a possibility to non-invasively quantify autonomic cardiac regulation. The technique splits the spectrum of the HRV into high frequency (HF), low frequency (LF) and very-low frequency (VLF) bands, and relates their power content to parasympathetic outflow, sympathetic outflow, and hormonal factors as well as thermoregulation, respectively. In this work, we applied this technique on HRV data from 20 end-stage renal disease (ESRD) patients to estimate power distribution in these three bands as well as their changes respect to time during hemodiafiltration (HDF) therapy; as an indirect way to assess the modulatory activity over the cardiac system. HRV data was processed using the continuous wavelet transform (CWT), power was estimated using 5-min as well as 10-min epochs for comparison purposes, and the results were normalized to unity before linear regression analysis. For 5-min epochs, the LF power increased (p = 0.035), while there were no significant changes in the HF power (p = 0.2). In contrast, for 10-min epochs, both the LF (p = 0.002) and HF (p = 0.043) power decreased, while the VLF power increased (p<<0.05) and had larger values than the other two spectral powers. There were no statistically significant changes in the LF/HF ratio for any of the two epochs evaluated (5-min: p = 0.56, 10-min: p = 0.44). Results based on 5-min epochs are in agreement with previous studies, while those based on 10-min epochs suggest that modulatory mechanisms within the VLF band override the sympathovagal balance and play a key role in managing the cardiovascular response during hemodiafiltration.