Sympathovagal Balance: A Reappraisal
To the Editor:
The article by D.L. Eckberg1 will surely promote a florid discussion. Its structure is based on a number of arguments about which our disagreement is substantial.
The term “sympathovagal balance,” belonging to traditional physiology, was introduced by us in the study of heart rate variability (HRV) when we wrote in 1983 that upright posture “is expected to shift the sympathovagal balance toward a sympathetic predominance.”2 Two years later, Pomeranz et al3 wrote that “autonomic control of the heart in response to postural movements strikes a balance between the activities of the parasympathetic and sympathetic nervous system.” After the emphasis provided by Pagani et al,4 the term became widely used.
A hypothesis is a highly democratic entity, and its acceptance cannot be imposed. In our mind, like a horizontal beam pivoted at its center, “sympathovagal balance” refers to a reciprocal functional relationship,5 implying that when 1 of the 2 components of the autonomic outflow is excited, the other is inhibited, according to a central push-pull pattern of organization.
We shall now analyze the major issues of disagreement in the sequence in which they appear in Eckberg’s article.1
Beginning from Figure 1,1 reproduced from our own work,5 Eckberg writes: “My integration of the two [low-frequency] spectral powers in this figure suggests that sympathetic spectral power is ≈15% greater than vagal spectral power”1 and, subsequently, “sympathetic contributions to 0.1 Hz spectral power are only marginally greater than vagal contributions.”
It is hard to understand how this comparison was made. In that experiment, the impulse activities of 2 distinct nerve filaments, unlikely to contain an equal number of active units, were simultaneously recorded with different amplifications, providing 2 variability signals with different variance. In the frequency domain, considering in the Figure the much greater scale of the sympathetic power spectrum density (16×103 versus 8×102) and the larger area of the sympathetic low-frequency (LF) component, it follows that the sympathetic LF power is at least 1 order of magnitude greater than vagal LF power. Most important, however, is that this comparison of absolute power is not only erroneous but deprived of physiological meaning and, in short, is not a brilliant overture for a “critical appraisal” article.
The Figure was intended by us to show that LF and high-frequency (HF) components are simultaneously present in both sympathetic and vagal discharges, which suggests that both LF and HF components of RR variability are likely to have a mixed origin and that “a rhythm, being a flexible and dynamic property of neural networks, should not necessarily be restricted to one specific neural pathway to carry a functional significance.”5
Regarding the subsequent issue raised by Dr Eckberg, the work by Introna et al6 is also misquoted. Spinal anesthesia, when the spread of spinal block reached the highest thoracic segments (above T3), induced a remarkable abatement of total power of HRV and, surprisingly, of both LF and HF, indicating that both components arise from a complex interaction of sympathetic and vagal mechanisms.
The next issue is of paramount importance. The article by Pagani et al7 reporting, as the main finding, a very tight correlation (P<10−6) between LF normalized units (nu) of RR and LFnu of muscle sympathetic nerve activity (MSNA) variability was discounted by Eckberg1 largely because individual regressions were not reported. We had introduced a nonparametric statistical analysis in a previous article8 in which individual regressions were included. In the article by Pagani et al,7 this was omitted “according to accepted practice” as stated by Koh et al9 and applied to their own article (see Figure 3 in Reference 9). Saul et al10 calculated group and individual regressions, presenting only part of the data “for simplicity.”
The Table⇓ reports the individual regressions, according to Theil (P is the probability that regression does not exist), computed from some of the data collected by Pagani et al.7 In A, regressions are calculated using LF absolute power of RR variability and bursts per minute of MSNA; as in Saul et al,10 very little correlation is present. In B, regressions are computed using LFnu for both RR and MSNA variability (data reported in the right inferior panel of Figure 6 in Reference 7); a significant correlation is present in 7 of 8 subjects, whereas a clear trend is observed in the remaining subject. Moreover, in B, all β-coefficients are positive and of similar magnitude, supporting the average correlation P<10−6. In addition, coherence analysis was used to ascertain, individually, the statistical link between oscillations of MSNA, RR, systolic arterial pressure (SAP), and respiration (Figure 7 in Reference 7).
In respect to other differences between the studies by Pagani et al7 and Saul et al,10 the following should be stressed: (1) Saul et al10 normalized their data in the time domain, whereas Pagani et al7 normalized their data in the frequency domain; (2) as in other studies,4 5 8 we verified that the spontaneous respiratory frequency was clearly separated from LF (see Figure 4 in Reference 7) (metronome breathing does not occur in normal life and, indeed, when the controlled frequency is close to the spontaneous cycle, it substantially increases the HF power, shifting the balance4 ); (3) fast Fourier transforms (FFTs) and autoregressive algorithms do not provide identical results, because only with the latter approach is it possible to perform a spectral decomposition, usually resulting in an LF component greater than that obtained with FFT (see Figure 5 in Reference 11).
The example of the “diving reflex” is not well taken because spectral analysis of HRV seems quite well suited to explore this peculiar interaction between vagal and sympathetic excitatory components. In our hands, cold stimulation of the face or water immersion most often induced a shift of the balance toward vagal predominance, in spite of the emotional arousal that, when prevailing, produces an opposite effect.
Concerning the reciprocal relationship of sympathetic and vagal outflows, we had demonstrated that stimulation of cardiac sympathetic afferents reflexly induces, respectively, an excitation or an inhibition of impulse activity of single sympathetic or vagal efferent fibers isolated from the same nerve impinging on the heart. An opposite effect was obtained by stimulating cardiac vagal afferents.12
The limitations of spectral analysis of HRV in several physiological and pathophysiological states have always been recognized by the proponents of this approach5 11 ; however no mention is made of their caution.
Concerning heart failure patients, it is astonishing that Dr Eckberg did not quote the paper by van de Borne et al,13 companion of Reference 7, in which it was shown that the patients who had no LF component in RR variability also did not present this spectral component in MSNA.
Concerning the effects of graded tilt,8 why does LFnu or LF/HF correlate better with tilt angle than HF in absolute values?
The remaining issues to be analyzed are still quite numerous. Thus, we shall rather attempt to incorporate them in a more general perspective. In our opinion, what prevails in Eckberg’s view of cardiovascular rhythmicity is a reductionistic model attempting to equate a rhythm to 1 or few reflexes, the sovereign of which is obviously the baroreflex (see, for example, Figure 5 in Reference 1). This view should address the following facts (none of them mentioned by Eckberg): (1) conscious dogs,4 5 14 when quiet and acquainted with the laboratory, most often present only an HF component in RR variability, although an LF component is present in SAP variability, and even though the baroreflexes are known to be extremely active in this species; (2) transient coronary occlusion in the same model5 14 elicits a marked increase in LF, probably as a result of an excitatory cardiac sympathetic reflex,15 which can occur in the absence of arterial pressure changes and which is known to acutely reduce the baroreflex gain; (3) exercising dogs increase their LFnu RR component while abating the baroreflex gain5 ; and (4) some tetraplegic patients have an LF RR component in the absence of an LF SAP component.16
We think that these are not just details but almost insurmountable barriers against the rigid interpretation of a baroreflex-dependent and vagally mediated LF RR component.
We obviously recognize the possibility that baroreflex mechanisms might participate substantially in the genesis of cardiovascular rhythmicity17 but have always questioned5 the exclusiveness attributed to them in the genesis of the LF component of RR variability.
In the attempt to contribute to a new way of thinking more adequate to neural complexity, we have advanced a hypothesis based on the interaction of patterns, subserved by multiple reflexes. It is a fact that major neural patterns, like wakefulness and sleep, are characterized by distinct and recognizable rhythmic activities.
What we propose is that in closed-loop conditions,5 2 main rhythms, 1 marker of excitation and linked to sympathetic excitation (LF) and 1 marker of inhibition and quiet and linked to vagal predominance (HF), would be organized, in physiological conditions, in a reciprocal manner. The recent article by Jasson et al18 provides a remarkable demonstration of this hypothesis. Its Figure 4 shows, in the time and frequency domains, a perfect coincidence between an increase of LF, a decrease of HF, a shift of the instant center frequency of the whole spectrum, and an increase of heart rate at the beginning of and during tilt. With this technique, no normalization procedure is necessary, and therefore the doubt that the reciprocal relationship between LFnu and HFnu might reflect only a simplistic mathematical model1 should be removed. This could be an example of how the 2 rhythms “constantly interact.”1
However, it is also of paramount importance to realize how diffuse these rhythms are. Both LF and HF components were found in the variability of the discharge of single medullary neurons recorded in sinoaortic denervated cats.19 In addition, acute spinal cats have an LF component in sympathetic cardiac nerve activity in RR and SAP variability after thoracic dorsal root section, suggesting the existence of an intrinsic spinal rhythmicity, independent even of the spinal afferent input (Montano et al, unpublished data, 1997). In Eckberg’s view, although a sympathetic modulation of the Mayer waves is likely to exist in tetraplegic patients, the LF RR oscillations are due to a vagally mediated baroreflex.1 9 Would a simpler hypothesis not be that the LF rhythmicity, intrinsic to the pattern of sympathetic excitation, affects both RR and SAP variability? In this case, the increase in the LF RR component, induced by phenylephrine in quadriplegic patients,1 could be due to a sympathetic spinal excitatory reflex20 released from baroreceptor restraint.
The similarities between the somatic and the autonomic nervous system have become progressively more evident, especially regarding spinal mechanisms. On this basis, a revision of the autonomic integrative properties was recently attempted.20 Thus, the comparison with somatic reflexes does not derive from philosophy, as criticized by Eckberg,1 but from physiology.
The sympathovagal balance is nothing but a concept. Although it is not totally quantifiable (like other useful concepts, such as homeostasis or intelligence), it should be judged for its heuristic value. The normalization procedure was not the result of serendipity but rather a result of this concept. Similarly, LF/HF ratio was proposed4 5 8 11 to assess the fractional distribution of power especially when simpler algorithms were used, such as FFT. The new approach was tested not only with subtractive strategies (ie, atropine administration14 ) but with observational studies, which usually constitute the main path to the study of neural complexity. We selected patterns well known to exist, such as the sympathetic excitation and the vagal withdrawal during standing, comprising a cohort of reflexes and modified gains. Circadian rhythmicity was another pattern that could be assessed clearly by spectral analysis of recordings obtained with either high-fidelity measurement of arterial pressure or usual Holter devices21 (another topic ignored by Eckberg’s article1 ).
In our last publication,22 we demonstrated that body posture (supine or upright) can be predicted in ≈85% of the cases by using 10 spectral variables extracted from short-term series of RR intervals. However, similar results were also obtained using only 3 variables, ie, RR, LFnu, and HFnu. Inconsistent results were provided when only 2 variables were used, including RR and HF in absolute units (ie, the 2 variables considered by Eckberg to carry the relevant information). This accomplishment is probably the best answer to Eckberg’s skepticism.
Regarding our language (an aspect analyzed throughout 42 lines of text), we think that the metaphor comparing “RR-interval fluctuations” to “ripples on a sea of varying depths”1 is more remote from physiology than the flexor-extensor interaction.5 But it is a fact that any new way of thinking has to generate its own language.
To the warning that “calculations of sympathovagal balance may obscure rather than illuminate human physiology and pathophysiology”1 we would like to reply that this concept might instead have furnished the Rosetta stone that made decipherable the puzzle of the rhythmic components of HRV.
Finally, because the Task Force article11 still represents the only document on HRV endorsed by an international panel of experts in the various fields, its complete disregard requires some reliable explanation for the scientific community.
- Copyright © 1998 by American Heart Association
Eckberg DL. Sympathovagal balance: a critical appraisal. Circulation. 1997;96:3224–3232.
Brovelli M, Baselli G, Cerutti S, Guzzetti S, Liberati D, Lombardi F, Malliani A, Pagani M, Pizzinelli P. Computerized analysis for an experimental validation of neurophysiological models of heart rate control. In: Computers in Cardiology. Silver Spring, Md: IEEE Computer Society Press; 1983:205–208.
Pomeranz B, Macaulay JB, Caudill MA, Kutz I, Adam D, Gordon D, Kilborn KM, Barger AC, Shannon DC, Choen RJ, Benson H. Assessment of autonomic function in humans by heart rate spectral analysis. Am J Physiol. 1985;248:H151–H153.
Pagani M, Lombardi F, Guzzetti S, Rimoldi O, Furlan R, Pizzinelli P, Sandrone G, Malfatto G, Dell’Orto S, Piccaluga E, Turiel M, Baselli G, Cerutti S, Malliani A. Power spectral analysis of heart rate and arterial pressure variabilities as a marker of sympathovagal interaction in man and conscious dog. Circ Res. 1986;58:178–193.
Malliani A, Pagani M, Lombardi F, Cerutti S. Cardiovascular neural regulation explored in the frequency domain. Circulation. 1991;84:482–492.
Pagani M, Montano N, Porta A, Malliani A, Abboud FM, Birkett CL, Somers VK. Relationship between spectral components of cardiovascular variabilities and direct measures of muscle sympathetic nerve activity in humans. Circulation. 1997;95:1441–1448.
Montano N, Gnecchi Ruscone T, Porta A, Lombardi F, Pagani M, Malliani A. Power spectrum analysis of heart rate variability to assess the changes in sympathovagal balance during graded orthostatic tilt. Circulation. 1994;90:1826–1831.
Koh J, Brown TE, Beightol LA, Ha CY, Eckberg DL. Human autonomic rhythms: vagal cardiac mechanisms in tetraplegic subjects. J Physiol. 1994;474:3:483–495.
Saul JP, Rea RF, Eckberg DL, Berger RD, Cohen RJ. Heart rate and muscle sympathetic nerve variability during reflex changes of autonomic activity. Am J Physiol. 1990;258:H713–H721.
Task Force of the European Society of Cardiology and the North American Society of Pacing and Electrophysiology. Heart rate variability: standards of measurement, physiological interpretation, and clinical use. Circulation. 1996;93:1043–1065.
Schwartz PJ, Pagani M, Lombardi F, Malliani A, Brown AM. A cardiocardiac sympathovagal reflex in the cat. Circ Res. 1973;32:215–220.
van de Borne P, Montano N, Pagani M, Oren RM, Somers VK. Absence of low-frequency variability of sympathetic nerve activity in severe heart failure. Circulation. 1997;95:1449–1454.
Rimoldi O, Pierini S, Ferrari A, Cerutti S, Pagani M, Malliani A. Analysis of short-term oscillations of RR and arterial pressure in conscious dogs. Am J Physiol. 1990;258:H967–H976.
Malliani A, Schwartz PJ, Zanchetti A. A sympathetic reflex elicited by experimental coronary occlusion. Am J Physiol. 1969;217:703–709.
Guzzetti S, Cogliati C, Broggi C, Carozzi C, Caldiroli D, Lombardi F, Malliani A. Influences of neural mechanisms on heart period and arterial pressure variabilities in quadriplegic patients. Am J Physiol. 1994;35:H1112–H1120.
Piepoli M, Sleight P, Leuzzi S, Valle F, Spadacini G, Passino C, Jonhston J, Bernardi L. Origin of respiratory sinus arrhythmia in conscious humans: an important role for the arterial carotid baroreceptors. Circulation. 1997;95:1813–1821.
Jasson S, Médigue C, Maison-Blanche P, Montano N, Meyer L, Verneiren C, Mansier P, Coumel P, Malliani A, Swynghedauw B. Instant power spectrum analysis of heart rate variability during orthostatic tilt using a time-/frequency-domain method. Circulation. 1997;96:3521–3526.
Furlan R, Guzzetti S, Crivellaro W, Dassi S, Tinelli M, Baselli G, Cerutti S, Lombardi F, Pagani M, Malliani A. Continuous 24-hour assessment of the neural regulation of systemic arterial pressure and RR variabilities in ambulant subjects. Circulation. 1990;81:537–547.
Malliani A, Pagani M, Furlan R, Guzzetti S, Lucini D, Montano N, Cerutti S, Mela GS. Individual recognition by heart rate variability of two different autonomic profiles related to posture. Circulation. 1997;96:4143–4145.