Interpreting oscillations of muscle sympathetic nerve activity and heart rate variability

Computer analysis of spontaneous cardiovascular beat-by-beat variability has gained wide credibility as a means of inferring disturbances of autonomic cardiovascular regulation in a variety of cardiovascular conditions, including hypertension, myocardial infarction and heart failure. Recent applications of spectral analysis to muscle sympathetic nerve activity (MSNA) offer a new approach to a better understanding of the relationship between cardiovascular oscillations and autonomic regulation. However, areas of uncertainty and unresolved debates remain, mostly concerning different methodologies and interpretative models that we will consider in this article. Perusal of all available literature suggests that average sympathetic nerve activity and its oscillatory components, although correlated to some extent, are likely to provide different types of information. In addition, the specific experimental context is of paramount importance, as the rules that seem to govern the relationship between average and oscillatory properties of MSNA appear to be different in usual conditions and in conditions of extremes of activation or disease. In general, dynamic experiments, such as with graded tilt or with vasoactive drugs, are more suited to investigations of the complexity of autonomic regulation than are static comparisons. In addition, because the information is spread across variables and is affected by a potentially large error, it appears that several different techniques should be perceived as complementary rather than as mutually exclusive. Available evidence suggests that low-frequency and high-frequency oscillations in peripheral signals of variability might have a predominantly central, rather than a peripheral, origin and that this applies in particular to low-frequency oscillations. A crucial point in the assessment of the meaning of spectral components relates to consideration of the varying level of very-low-frequency noise, and the mathematical manipulation of derived indices, particularly using a normalization procedure. This appears easier to obtain with auto-regressive than with fast Fourier techniques. With this approach, discrepant interpretations seem to be resolved, provided adequate care is taken in separating direct physiological data from derived meaning, which relates to hidden information and neural codes; in the case of sympathetic discharge, the latter display greater complexity than simple average spike activity per unit time. Accordingly we believe, in conclusion, that the judicious use of spectral methodology, in addition to other techniques, might provide unprecedented, useful insights into autonomic cardiovascular regulation, in both physiopathological and clinical circumstances.