Wavelet-derived indices of cardiac vagal and vasomotor oscillations as a new tool to better understand autonomic imbalance in post-COVID-19 dysautonomia
Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background
When aiming to quantify autonomic imbalance, conventional heart rate variability (HRV) analysis generally reduces each posture to a single spectral value derived from interbeat intervals (IBI) alone. This analysis based only in IBI obscures transient events during blood pressure changes. We propose two wavelet-derived indices that quantify cardiac vagal and vasomotor sympathetic oscillations separately and over time. Since autonomic imbalance is directly related to baroreflex sensitivity (BRS), we evaluated them against an estimate of baroreflex sensitivity in post-COVID-19 patients.
Methods
In this study, 121 participants (32 healthy controls and 89 patients with persistent post-COVID-19 autonomic symptoms) underwent an active standing test with continuous finger blood pressure. The continuous wavelet transform with generalized Morse wavelets yielded time-varying high-frequency power of IBI and low-frequency power of systolic blood pressure (SBP). Two new indices were proposed: Index I combined the magnitude and variability of HF IBI power (aimed to quantify parasympathetic activity), and the index J combined LF SBP power with the rate of its transient elevations (bursts), which aims to quantify sympathetic vasomotor activity; J/I was their ratio. Indices were computed for supine and orthostatic periods and compared with conventional HRV and BRS measurements.
Results
Index I correlated strongly with HF power of IBI and was lower in all patient subgroups than in controls, Index J had only a weak relation with LF power of IBI. On standing, J increased most in the subgroups with the worst baroreflex function. Orthostatic J/I was higher than in controls amongst patients, while LF/HF showed no differences. J/I was inversely correlated with BRS, most strongly when standing, and more strongly than LF/HF also. Adding J to I raised the cross-validated variance explained in orthostatic BRS substantially. In one example patient, a window during symptoms showed vasomotor oscillations that were hidden in the posture average.
Conclusions
Wavelet-derived indices from routine beat-to-beat recordings separate cardiac vagal modulation from sympathetic vasomotor oscillations and their transients. J/I correlates with BRS impairment more closely than LF/HF, providing a potential alternative estimate of autonomic balance, and its time-resolved visual output allows analysis of clinically relevant windows within a posture. The new proposed variables allowed an evaluation of autonomic imbalance that could be applied to different pathologies.