(Circulation. 1999;100:e64.)
© 1999 American Heart Association, Inc.
Circulation Electronic Pages |
Mathematical Treatment of Autonomic Oscillations - 2
Assistant Professor of Medicine Harvard Medical School, Director, Laboratory for Cardiovascular Research, HRCA Research and Training Institute, Boston, Mass
Shuman Cardiovascular Research Fellow HRCA Research and Training Institute, Boston, Mass
 |
Introduction |
|---|
 Top Introduction ReferencesIntroduction References |
|---|
To the Editor:
Recently, Montano et al1 claimed a central vagotonic effect of high-dose atropine was evidenced in peroneal nerve muscle sympathetic outflow (MSNA). However, the authors' conclusions critically depend on "normalized units" to quantify low-frequency (LF) and high-frequency (HF) oscillations, a practice that can impart significance to the fluctuations beyond the regulatory mechanisms they subserve. This led to the conclusion that insight into parasympathetic nervous outflow can be gleaned from activity in a sympathetic nerve. We take issue with the interpretation of these data and believe that despite cautionary argument,2 this approach subsumes the physiological meaning of cardiovascular oscillations to their spectral measures.
Heart period oscillations primarily derive from beat-by-beat autonomic control of systemic hemodynamics, ultimately buffering or augmenting arterial pressure fluctuations.3 Vascular sympathetic rhythms have been identified also, although they may or may not be related directly to pressure fluctuations.4 5 Spectral analysis conveniently quantifies these rhythms but in itself does not reveal their source. The findings of Montano et al rely solely on "normalizing" power spectral data, a technique that uncouples the oscillations from their physiological significance by measuring LF and HF relative to each other and making absolute amplitude irrelevant. In the present study, average heart period variance after atropine was <1% of control, representing almost complete elimination of beat-by-beat cardiac autonomic regulation. However, normalized units indicated that high-dose atropine reduced HF variability by only two thirds and increased LF variability by one third, divorcing spectral measures from the oscillations' minimal physiological significance. It is unclear how normalized units affected measures of MSNA variability, since absolute values were not provided.
The use of normalized units seems to presume that cardiovascular oscillations are rather than derive from autonomic outflows; that is, that HF is parasympathetic outflow and LF is sympathetic outflow. Furthermore, the authors apply this assumption to direct MSNA recordings, arriving at the curious conclusion that parasympathetic effects may be "revealed only by examination of the HF oscillation of MSNA."1 Outflow from sympathetic nerves measured by peroneal microneurography is simply sympathetic outflow, regardless of the frequency at which it oscillates. However, "normalizing" HF and LF oscillations to one another and equating HF oscillations with parasympathetic outflow lead to a conclusion that ignores this simple fact.
|
References |
|---|
|
TopIntroductionReferencesIntroduction References |
|---|
-
Montano N, Cogliati C, Porta A, Pagani M, Malliani
A, Narkiewicz K, Abboud FM, Birkett C, Somers VK. Central vagotonic
effects of atropine modulate spectral oscillations of
sympathetic nerve activity. Circulation. 1998;98:1394–1399.
[Abstract/Free Full Text]
-
Eckberg DL. Sympathovagal balance: a critical
appraisal. Circulation. 1997;96:3224–3232.
[Free Full Text]
-
Taylor JA, Eckberg DL. Fundamental relations between
short-term RR interval and arterial pressure
oscillations in humans. Circulation. 1996;93:1527–1532.
[Abstract/Free Full Text]
-
Taylor JA, Williams TD, Seals DR, Davy KP.
Low-frequency arterial pressure fluctuations do not reflect
sympathetic outflow: gender and age differences. Am J
Physiol. 1998;274:H1194–H1201.
[Abstract/Free Full Text]
-
Pagani M, Montano N, Porta A, Malliani A, Abboud FM,
Birkett C, Somers VK. Relationship between spectral components of
cardiovascular variabilities and direct measures of
muscle sympathetic nerve activity in humans. Circulation. 1997;95:1441–1448.
[Abstract/Free Full Text]
Response
Centro L.I.T.A. di Vialba, Centro Ricerche Cardiovascolari, CNR, Medicina Interna II, Ospedale L. Sacco, Università di Milano, Milano, Italy
Cardiovascular Division, Department of Internal Medicine, College of Medicine, University of Iowa, Iowa City, Iowa
|
Introduction |
|---|
|
TopIntroductionReferencesIntroduction References |
|---|
We thank Dr Eckberg for his comments on our work.
Normalization of burst amplitudes and normalization of spectral power: Low-dose atropine did not cause a significant change in MSNA (-15±14%). High-dose atropine decreased MSNA by 62±7% (P<0.03). Dr Eckberg correctly refers to potential effects of "uncontrollable differences" in burst amplitude between subjects. These bursts are arbitrary measures and are best understood when they are normalized, an approach clearly favored by Dr Eckberg. Later in his comments, however, he suggests that our normalization of LF and HF powers of MSNA may not be appropriate. Since absolute LF and HF spectral measures are derived from absolute and arbitrary measures of MSNA, we believe that normalization helps mitigate the problem of variability between individuals and is consistent with the principles inherent in normalization of burst amplitudes.
Comparisons of our data with those of Ikuta et al1: They found a slight increase in LF RR spectral power after low-dose atropine, whereas we found no change. This difference in our studies needs to be kept in perspective. First, all subjects in our study were male. Ikuta et al studied only females. Second, we used a bolus dose of atropine; Ikuta et al used steady-state infusions increasing every 4 to 5 minutes to a total duration of 24 minutes. Third, we used an autoregressive algorithm, whereas Ikuta et al used a fast Fourier transform with an LF band between 0.05 and 0.15 Hz.
"Your observation is at variance with one of your principal conclusions": There is no conflict between the results and the conclusions in our Abstract. The Abstract refers exclusively to normalized data. It would then seem logical that the conclusion would also refer to normalized data.
Reduced normalized LF RR power simply signifies increased sinus arrhythmia: This suggestion highlights the importance of simultaneous measurements of both RR and MSNA spectral powers. Sinus arrhythmia (breathing-related changes in RR interval) occurs in the HF range. During low-dose atropine, the decrease in normalized LF RR in our study was accompanied by a decreased normalized LF of MSNA, and as described later, a decreased absolute LF power of MSNA. Thus, the similar effects of low-dose atropine not only on heart rate but also on MSNA demonstrate very clearly that effects of low-dose atropine on RR interval and on other measures of cardiovascular variability involve mechanisms other than sinus arrhythmia alone.
"Can you explain the disparity between your results and ours?": Our data are consistent with human studies by others2 demonstrating that atropine increases blood pressure and decreases norepinephrine, which is also at odds with Dr Eckberg's findings. In Dr Eckberg's study, only descriptive information of the qualitative absence of changes in blood pressure, MSNA (in bursts per minute), and norepinephrine was provided. No actual measurements of these variables were reported. It is surprising that blood pressure did not increase after high-dose atropine. Blood pressure in his study was measured noninvasively by use of an intermittent blood pressure monitor. By contrast, we report actual measures of continuous recordings of intra-arterial blood pressure and MSNA. We are very comfortable with our data showing that tachycardia after high-dose atropine is associated with an increase in intra-arterial systolic pressure and that this increase in systolic pressure is associated with an unequivocal reduction in MSNA, whether expressed as bursts per minute, bursts per 100 heart beats, arbitrary units, or normalized units. In view of the unreported data and qualitative descriptions of changes in blood pressure and MSNA referred to by Dr Eckberg, we are not comfortable speculating on theoretical reasons for inconsistencies in the findings in his study compared with more explicit data from our and other investigators' studies.2
"You present only `normalized' MSNA": The following are the absolute values, in arbitrary units squared (au2): MSNA LF: baseline 277±179, low-dose atropine 107±80 (P<0.05 versus baseline), and high-dose atropine 117±103. MSNA HF: baseline 189±124, low-dose atropine 108±66, and high-dose atropine 161±153. Thus, the decrease in normalized LF of MSNA after low-dose atropine is accompanied by a decrease in absolute LF of MSNA, demonstrating that the effect of low-dose atropine on the variability profile of MSNA is not simply a function of normalization.
"When you divide one by the other ... , you enter largely uncharted territory": We are in uncharted territory whether we refer to ratios or absolute values, since we do not know the precise mechanism of the oscillations. By looking at simultaneous oscillatory characteristics of 2 different autonomic outflows, RR and MSNA, we may arrive at more reasonable interpretations. The only certainty is that methodological and conceptual paradigms will change as new experimental knowledge is gleaned.
We also appreciate Drs Taylor and Myers' interest in our findings:
"Heart period oscillations primarily derive from beat-by-beat autonomic control of systemic hemodynamics": This thesis is flawed for several reasons. First, LF oscillations have been demonstrated in the discharge of single brain stem neurons recorded in sinoaortic denervated cats; furthermore, this LF oscillation was present even in the absence of similar blood pressure fluctuations.3 Second, in patients with heart failure studied before and after implantation of a left ventricular assist device, there is a striking, newly evident LF oscillation in RR interval of the native heart after device implantation.4 This LF oscillation is manifest in the absence of any similar oscillation in blood pressure (which is dependent on the artificial heart output and independent of RR characteristics of the native heart). Third, in our study of atropine, low-dose atropine induced significant changes in the spectral patterns of both RR and MSNA in the absence of any changes in absolute or spectral components of intra-arterial blood pressure. Fourth, Dr Taylor himself concludes in one of his recent studies5 that "respiratory sinus arrhythmia does not represent simple baroreflex buffering of arterial pressure."
"HF is parasympathetic outflow and LF is sympathetic outflow": Drs Taylor and Myers have misinterpreted and misrepresented our Results and Discussion in their last paragraph. We refer them to our extensive experimental evidence showing, for example, that despite high sympathetic drive, patients with heart failure have decreased or absent LF powers of RR and MSNA variability.6 We also refer them to our unequivocal statements that "our data do not imply that the frequency composition of an oscillatory signal can be equated with the strength of that signal"7 and "[our] findings should not be misinterpreted as implying that power spectral variability can be equated to direct measurements of sympathetic or other autonomic function."8
"Outflow from sympathetic nerves ... is simply sympathetic outflow, regardless of the frequency at which it oscillates": It is not clear why Drs Taylor and Myers presume that central effects of low-dose atropine would affect heart rate exclusively and that sympathetic and parasympathetic outflows are mutually exclusive and devoid of interaction. Cholinergic muscarinic receptor blockade modulates adrenergic neurotransmission and norepinephrine release. Parasympathetic mechanisms therefore exert inhibitory effects on both cardiac9 and vascular10 sympathetic activity.
"The authors ... [arrive] at the curious conclusion that parasympathetic effects `may be revealed only by examination of the HF oscillation of MSNA.'" Any central parasympathetic muscarinic influence of high-dose atropine on RR variability would be masked by the peripheral (sinoatrial nodal) muscarinic blockade by atropine and the ensuing tachycardia. Changes in the MSNA oscillatory profile are consistent with central vagotonic effects of atropine, are evident whether one considers normalized or absolute measures, and occur even in the absence of any change in overall MSNA. These effects cannot be ignored. We welcome a better strategy for assessing in humans the influence of central muscarinic modulation of cardiovascular oscillations.
|
References |
|---|
|
TopIntroductionReferencesIntroduction References |
|---|
-
Ikuta Y, Shimoda O, Kano T. Quantitative
assessment of the autonomic nervous system activities during
atropine-induced bradycardia by heart rate spectral analysis.
J Auton Nerv Syst. 1995;52:71–76.[Medline]
[Order article via Infotrieve]
-
Goldstein DS, Keiser HR. Pressor and depressor
responses after cholinergic blockade in humans. Am Heart
J. 1984;107:974–979.[Medline]
[Order article via Infotrieve]
-
Montano N, Gnecchi Ruscone T, Porta A, Lombardi F,
Malliani A, Barman SM. Presence of vasomotor and respiratory rhythms in
the discharge of single medullary neurons involved in the regulation of
cardiovascular system. J Auton Nerv Syst. 1996;57:116–122.[Medline]
[Order article via Infotrieve]
-
Cooley R, Montano N, Cogliati C, van de Borne P, Oren
R, Richenbacher W, Somers VK. Evidence for a central origin of the low
frequency oscillation in RR interval variability.
Circulation. 1998;98:556–561.
[Abstract/Free Full Text]
-
Taylor JA, Eckberg DL. Fundamental relations between
short-term RR interval and arterial pressure
oscillations in humans. Circulation. 1996;93:1527–1532.
-
van de Borne P, Montano N, Pagani M, Oren R, Somers
VK. Absence of low frequency variability of sympathetic nerve activity
in severe heart failure. Circulation. 1997;95:1449–1454.
[Abstract/Free Full Text]
-
Pagani M, Montano N, Porta A, Milliani A, Abboud FM,
Birkett C, Somers VK. Relationship between spectral components of
cardiovascular variabilities and direct measures of
muscle sympathetic nerve activity in humans. Circulation. 1997;95:1441–1448.
-
van de Borne P, Montano N, Pagani M, Zimmerman B,
Somers VK. Relationship between repeated measures of
hemodynamics, muscle sympathetic nerve activity, and
their spectral oscillations. Circulation. 1997;96:4326–4332.
[Abstract/Free Full Text]
-
Haunstetter A, Haass M, Yi X, Krüger C,
Kübler W. Muscarinic inhibition of cardiac
norepinephrine and neuropeptide Y release during
ischemia and reperfusion. Am J Physiol. 1994;267:R1552–R1558.
[Abstract/Free Full Text]
-
Rorie DK, Rusch NJ, Shepherd JT, Vanhoutte PM, Tyce GM.
Prejunctional inhibition of norepinephrine release caused
by acetylcholine in the human saphenous vein. Circ Res. 1981;49:337–341.
[Abstract/Free Full Text]
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||