Absence of Low-Frequency Variability of Sympathetic Nerve Activity in Severe Heart Failure
Background In normal humans, variability of blood pressure, RR interval, and sympathetic activity occurs predominantly at a low frequency (LF; 0.04 to 0.14 Hz) and a high frequency (HF; ±0.25 Hz). In conditions that increase sympathetic activation in normal humans, the LF component is increased relative to the HF component. Patients with heart failure have high levels of sympathetic activity. We tested the hypothesis that the LF component of sympathetic nerve activity variability is increased in heart failure.
Methods and Results. We performed spectral analysis of simultaneous recordings of resting muscle sympathetic nerve activity (MSNA) and RR interval in 21 patients with chronic heart failure and 12 age-matched control subjects. MSNA was higher in patients with heart failure (62±4 bursts per minute) than in the normal subjects (39±4 bursts per minute; P<.01). LF components of RR interval and MSNA variability were lower in the heart failure patients versus the control subjects (P<.01). HF variability of RR interval and MSNA was preserved, at least in part, in heart failure. There was close coherence between variability patterns of RR interval and MSNA. Furthermore, in 14 heart failure patients who had no LF variability in MSNA compared with 7 heart failure patients who did manifest LF variability in MSNA, RR interval was shorter, the variance of RR interval was lower, MSNA was higher, respiratory rate was faster, and left ventricular ejection fraction was lower (all P<.05). At a median follow-up of 12 months, 4 heart failure patients had died, all of whom had had absent LF oscillations in MSNA and RR interval.
Conclusions The LF variability of sympathetic nerve activity is absent in patients with severe heart failure. This disturbed pattern of variability is closely coherent with the abnormal variability of RR interval. These disturbances of rhythmic oscillations of autonomic outflow, evident in both RR interval and MSNA, suggest a central autonomic regulatory impairment in heart failure and may have important prognostic implications.
In normal humans, RR interval variability occurs predominantly at an LF (0.04 to 0.14 Hz) and an HF (±0.25 Hz, synchronous with the respiratory frequency).1 2 3 4 5 6 7 A similar variability profile is present in direct recordings of sympathetic nerve traffic8 9 10 11 12 and intra-arterial blood pressure.1 2 5 It has been suggested that assessment of the variability of heart rate could provide a noninvasive measure of cardiac sympathetic and vagal modulation.1 2 3 4 5 6 7 A reliable noninvasive measure of cardiac sympathetic activity would likely have prognostic implications, such as in the prediction of increased mortality after myocardial infarction or in chronic heart failure.13 14 15
In normal subjects, the relative power of the LF to HF variability of RR interval increases during conditions in which sympathetic activity rises, such as during tilt, moderate hypotension, mild physical exercise, mental stress, coronary occlusion, and rapid-eye-movement sleep.1 2 3 4 5 6 7 These observations have indicated that the ratio of the LF to the HF components of RR interval variability might provide an index of the dynamics of cardiac sympathetic-vagal balance. Sympathetic activation is a key component of the pathophysiology of chronic heart failure.16 17 Previous studies in normal subjects have demonstrated that MSNA contains oscillatory components at LF and HF almost identical to those present in heart rate variability.8 9 10 11 12 The high sympathetic drive in heart failure would therefore be reasonably expected to manifest with a relative increase in the LF oscillatory component. This has not previously been studied. What is known is that the variability of RR interval in heart failure shows a decreased LF component,18 19 20 21 possibly because of limitations in sinus node responsiveness to high levels of cardiac sympathetic activation22 and β-adrenoceptor downregulation.23 24
However, whether heart failure affects the variability of sympathetic activity and how this relates to the abnormal RR interval variability seen in these patients are unknown.
In this study, we tested the hypothesis that the normal variability of sympathetic activity is altered by heart failure and examined the relationship between MSNA variability and RR interval variability in these patients. We therefore performed spectral analysis of simultaneous recordings of MSNA, RR interval, and respiration in a large group of patients with severe chronic heart failure and in healthy control subjects.
Chronic Heart Failure Patients
We studied 21 chronic heart failure patients (14 men, 7 women) 56±13 years of age. All had supporting clinical, chest roentgenographic, and echocardiographic evidence of impaired ventricular function. The cause of heart failure was ischemic heart disease (n=9) or idiopathic (n=12). Left ventricular ejection fraction, determined by a resting radionuclide ventriculogram, was 23±2% (mean±SE). All patients were in sinus rhythm and were receiving a combination of diuretics, nitrates, converting enzyme inhibitors, and digitalis. Drug therapy was withheld the morning of the study to minimize the effects of medications on our measurements, without compromising the health status of patients.
We also studied 12 age- and sex-matched healthy control subjects (mean age, 55±17 years; 9 men, 3 women). None were receiving any medications.
Informed written consent was obtained from all patients and control subjects. The study was approved by the Institutional Human Subjects Review Committee.
Sympathetic nerve activity to muscle circulation was recorded continuously by obtaining multiunit recordings of postganglionic sympathetic activity to muscle, measured from a nerve fascicle in the peroneal nerve posterior to the fibular head, as described previously.8 9 10 Electrical activity in the nerve fascicle was measured with tungsten microelectrodes (shaft diameter, 200 μm, tapering to an uninsulated tip of 1 to 5 μm). A subcutaneous reference electrode was first inserted 2 to 3 cm away from the recording electrode, which was inserted into the nerve fascicle. The neural signals were amplified, filtered, rectified, and integrated to obtain a mean voltage display of sympathetic nerve activity.
ECG, respiration (pneumograph), and MSNA were recorded for 10 minutes on a Gould 2800 S recorder and an IBM 433DX/T computer in all the patients and control subjects.
Sympathetic bursts were identified by a single observer (Dr van de Borne). The intraobserver and interobserver variabilities in our laboratory have been reported to be 4.3±0.3%25 and 5.4±0.5%.26 MSNA was calculated as bursts per minute after careful inspection of the mean voltage neurogram.
Analog-to-digital conversion was performed in real time at 600 samples per second per channel. The data were then analyzed off-line with a personal computer (433DX/T, IBM). The principles of the software for data acquisition and spectral analysis have been described elsewhere.1 In summary, a derivative/threshold algorithm provided the continuous series of RR intervals (tachogram) derived from the ECG. Isolated artifacts and rhythm disturbances were detected and removed. All interpolated values were visually checked. Stationary segments devoid of arrhythmias (150 to 300 RR intervals) were analyzed with autoregressive algorithms. These algorithms provided the number, center frequency, and power of the oscillatory components. A potential advantage of the autoregressive method is that it allows an accurate spectral estimation on short segments of data, which are more likely to be stationary. Furthermore, statistical criteria,1 27 28 such as Akaike’s test and Anderson’s test, allowed us to determine the optimal model order (ranging between 8 and 12) fitting the data and verified that all information contained in the time series had been extracted in the computation.
Previous studies1 2 3 4 5 6 7 have shown that two major oscillatory components are usually detectable in the RR interval. One of these oscillatory components (the HF component) is synchronous with respiration (usually ≈0.25 Hz). The other component is described as the LF. Its center frequency is ≈0.10 Hz but can vary considerably (from 0.04 to 0.13 Hz).
In this study, the LF and HF components were expressed in absolute (milliseconds squared) and normalized units. The normalized units were obtained by calculating the percentage LF and HF variability with respect to the total power after subtracting the power of the VLF component (frequencies <0.03 Hz).1 2 3 4 5 6 7
The signals of sympathetic nerve activity and respiratory activity were sampled once every cardiac cycle, thus providing a neurogram and a respirogram synchronized with the tachogram. Before sampling, the neurogram was preprocessed to provide for each cardiac cycle the time-integrated value of the signal. The respirogram was sampled in correspondence with the R wave of the ECG. These time series underwent an analysis similar to that described above for the tachogram.
A coherence (K2) function was then used to determine the amount of linear coupling between the series of RR interval, MSNA, and respiration.7 This measure has the same meaning as the squared correlation coefficient (explained variance) in a linear regression equation and allows determination of the amount of linear coupling between the oscillations present in different time series.
Results are expressed as mean±SEM. The data of the heart failure patients were analyzed blinded to the degree of severity of heart failure. Statistical analysis consisted of Mann-Whitney U tests corrected for ties and χ2 tests (level of significance assumed at P<.05).
MSNA was markedly increased in the patients with heart failure (62±4 bursts per minute) compared with the normal subjects (39±4 bursts per minute; P<.01; Table 1⇓ and Fig 1⇓). Heart failure was also accompanied by a shorter RR interval (P<.05) and a fivefold reduction in the variance of RR interval (P<.01).
LF and HF components were found in the RR interval and MSNA recordings of all the control subjects (Figs 1⇑ and 2⇓). In contrast, only 4 of the 21 RR interval recordings and 7 of the 21 MSNA recordings of the heart failure patients disclosed an LF component (P<.001 by χ2 test compared with the control subjects; Figs 1⇑ and 2⇓). The four heart failure patients who did have an LF component in their RR interval also had an LF component in sympathetic activity. All 21 heart failure patients had HF components in both the RR interval and MSNA variabilities. The center frequencies of the LF and HF components of the heart failure patients did not differ significantly from those of the control subjects. Both control subjects and heart failure patients had a VLF component in their variability profiles. The significance of this VLF oscillation is not known and is not addressed in this report.
The mean LF powers in RR interval and MSNA were lower in the 21 heart failure patients than in the 12 control subjects (LF RR interval power: absolute, 54±34 versus 863±407 ms2; normalized units, 11±5 versus 55±7; LF sympathetic power: absolute, 39±23 versus 93±52 AU2; normalized units: 9±3 versus 27±3, respectively, in the patients with heart failure and the control subjects; Table 1⇑; all P<.01).
In contrast, the reduction in the absolute HF variability of RR interval in heart failure patients fell short of statistical significance (P=.052; Table 1⇑). Both heart failure patients and control subjects displayed similar absolute HF powers of MSNA variability. The normalized HF powers of RR interval and MSNA variability were increased in the heart failure patients (P<.05). The overall decrease in RR interval variability in the heart failure patients was largely reflected by a nearly absent LF variability. The marked sympathetic excitation in heart failure, confirmed by direct intraneural recordings obtained from microneurography and resting tachycardia, was associated with a reduction in the ratio of LF to HF variability of RR interval in the patients with heart failure (LF/HF ratio of RR interval, 0.6±0.3 in the heart failure patients versus 3.3±0.8 in the control subjects; P<.0001). The ratio of LF to HF variability of MSNA was also reduced in the patients with heart failure (LF/HF ratio of sympathetic activity, 0.3±0.5 in the heart failure patients versus 1.0±0.7 in the control subjects: P=.001).
Post hoc comparison of heart failure patients with absent LF variability versus those in whom LF variability was present in MSNA revealed that the RR interval was shorter, the variance of the RR interval was lower, MSNA was higher, and left ventricular ejection fraction was lower in the patients who did not have any LF variability in sympathetic activity compared with those who did (all P<.05; Fig 3⇓). The patients who did not present an LF component in MSNA were also more tachypneic (respiratory rate, 0.34±0.02 Hz) compared with those in whom an LF component was present (respiratory rate, 0.26±0.02 Hz; P<.05). At a median follow-up of 12 months, 4 heart failure patients had died, all of whom had had absent LF oscillations in MSNA and RR interval. In heart failure patients in whom an LF component of MSNA was present (n=7), the LF/HF ratio was not significantly different from normal control subjects.
A coherence function was used to determine the amount of linear coupling between the series of RR interval, MSNA, and respiration. The coherence between the LF variability of the RR interval and MSNA in the 4 patients who did have these oscillations was similar to that presented by the 12 control subjects (Table 2⇓). Also, the linear couplings between the HF component of RR interval variability, MSNA, and respiration were similar for the heart failure patients and the control subjects.
One major finding of this study is that in patients with severe heart failure, oscillations of both MSNA and RR interval have distinct and closely coupled abnormalities. These abnormalities include an attenuation of the LF variability component. The HF variability component, indicative of the influence of respiration, is preserved at least in part. Another finding is that the complete absence of the LF component in MSNA and RR interval variability characterizes those patients with the highest level of MSNA and the most severe heart failure and may have important prognostic implications.
In normal humans, cardiovascular measures show a short-term variability that is consistently represented by two major oscillation frequencies. The HF oscillation at ≈0.25 Hz is in phase with respiration, and the LF oscillation at ≈0.1 Hz is in phase with vasomotor waves.1 2 3 4 5 6 7 These oscillations are present in normal conditions and closely coherent in simultaneous recordings of the RR interval, intra-arterial blood pressure, and MSNA.1 2 3 4 5 6 7 8 9 10 11 12
Traditional paradigms of spectral analysis of cardiovascular variability hold that increased tonic sympathetic activity and reduced tonic vagal activity are accompanied by an increase in the relative predominance of phasic LF variability.1 2 3 4 5 6 7 12 Conversely, quiet relaxation with consequent sympathetic inhibition and high vagal tone is manifest by a relative predominance of the HF component of cardiovascular variability.1 2 3 4 5 6 7 With nitroprusside used to lower blood pressure and increase sympathetic drive and phenylephrine used to raise blood pressure and inhibit sympathetic drive, this paradigm appears to hold true for comparisons within normal individuals during the same session.12 The use of normalized units or the LF/HF ratio provides numerical indexes for following changes in the balance between sympathetic and vagal cardiovascular drives.1 2 3 4 5 6 7 12
Heart failure is characterized by a high sympathetic drive to both the heart and peripheral blood vessels. Thus, spectral analysis of the RR interval and sympathetic neural discharge would be reasonably expected to manifest predominantly LF oscillations. Previous studies revealed that the RR interval variability in heart failure had a decreased LF component,18 19 20 21 perhaps because saturation of the sinus node by a very high sympathetic drive would make the sinus node less capable of maintaining a rhythmic modulation.22 β-Adrenoceptor downregulation could also contribute to this finding.23 24
Our patients with heart failure had resting tachycardia and high levels of MSNA, confirming the high sympathetic drive characteristic of heart failure. Spectral analysis of these measures, however, shows that a diminished LF component characterizes both the RR interval variability and the variability of intraneural recordings of MSNA to muscle blood vessels. An LF component of RR interval and MSNA was either present or absent simultaneously in 91% of our patients. Thus, this oscillatory abnormality affects central autonomic outflow in heart failure patients and indicates that the absence of an LF component in the variability of the RR interval in heart failure patients is not exclusively a consequence of peripheral changes in target organ responsiveness to autonomic drives.
Heart failure patients in whom the LF component was absent in both the RR interval and MSNA variabilities also had faster heart rates, faster respiratory rates, higher MSNA, and lower ejection fractions than those heart failure patients in whom LF oscillation was present. Short-term mortality (>25%) was limited exclusively to heart failure patients in whom LF variability of the RR interval and MSNA was absent. Thus, this abnormality in cardiovascular variability patterns may have prognostic implications.
The cause of a reduced or absent LF cardiovascular variability in heart failure is not known. Arterial baroreflex mechanisms may contribute to the modulation of the LF oscillations in cardiovascular rhythms.8 10 29 30 Hence, one possible explanation is an impairment in baroreflex circulatory regulation, well documented in heart failure.31 32 33 Alternatively, there may be saturation of LF oscillatory systems caused by the high sympathetic drive, with LF oscillations maintained in chronic heart failure patients with lower sympathetic activity. Other mechanisms may include a central effect of neurohumoral excitation, whereby excessive circulating levels of catecholamines, angiotensin, or vasopressin may compromise autonomic regulatory functions at a central level.
Heart failure tended to blunt the HF variability of the RR interval but did not affect the absolute HF variability of MSNA. HF variability components in the RR interval and MSNA were evident in all 21 heart failure patients, in contrast to the LF variability of the RR interval and MSNA, which was evident in only 4 of 21 heart failure patients. Thus, despite the rapid heart rate and high sympathetic tone, there is some preservation of respiratory modulation of autonomic drive8 9 10 11 34 in heart failure, which is in agreement with earlier findings by other investigators.35 Absent LF oscillations thus contributed in large part to the diminished cardiovascular variability in heart failure patients.
Despite the diverse neurohumoral, metabolic, and other abnormalities that characterize heart failure, there is a preservation of the coherence between the variabilities of RR interval and sympathetic neural discharge. The common variability patterns and abnormalities present in two separate autonomic nervous outflows, namely the sympathetic-vagal balance modulating the SA node and sympathetic neural discharge directed at the muscle vasculature, suggest that both vagal and sympathetic autonomic modulatory nuclei manifest unitary abnormalities in oscillatory characteristics in patients with heart failure. We therefore speculate that the pathophysiology of heart failure may include primary central abnormalities in autonomic modulation.
Increased sympathetic activation in heart failure but a decreased LF variability component in RR interval and MSNA and a decreased LF/HF ratio do not follow the traditional paradigm used to explain the HF and LF oscillations in cardiovascular variability. If it did, we would expect a relatively increased LF component in the heart failure patients. Why should there be this dissociation between high sympathetic drive and the LF components of MSNA and RR variability? We believe that although the traditional paradigm may hold true for a physiological range of autonomic drives in normal humans, it cannot be simply extrapolated to include such pathological conditions as severe heart failure. The ability of the cardiovascular system to maintain a high level of variability is a sign of health. In extremes of stress, such as heavy physical exercise36 or severe heart failure, when all physiological mechanisms are mobilized to the maximum to maintain homeostasis, the system has no reserve to maintain its variability. Consequently, spectral analysis of cardiovascular measures in heart failure provides a reliable index of patterns and absolute values of variability but is not a marker of the mean level of sympathetic drive. Thus, relationships between the tonic and phasic characteristics of sympathetic drive, evident in studies within normal subjects,12 are lost in patients with severe heart failure.
In conclusion, we have shown that the LF component of oscillations in the RR interval and MSNA variability is absent in patients with severe heart failure. The absence of the LF component identifies heart failure patients with the highest level of sympathetic activation and perhaps those with increased mortality risk. Central autonomic regulatory impairment in patients with heart failure may account for the observation that in contrast to physiological perturbations in normal subjects, the heightened sympathetic drive in heart failure does not manifest as an increase in the LF component of either sympathetic or overall cardiovascular variability.
Selected Abbreviations and Acronyms
|MSNA||=||muscle sympathetic nerve activity|
|VLF||=||very low frequency|
Dr van de Borne, a visiting research scientist from the Hypertension Clinic, Department of Cardiology, Free University of Brussels, Belgium, is a recipient of an International Research John E. Fogarty Fellowship (NIH 1F05-TW05181-01); a Belgian NATO Research Fellowship (17/B/94/BE); the Dr André Loicq Foundation Award, Belgium; a Bekales Research Award, Belgium; and a Michael J. Brody Fellowship from the University of Iowa. These studies were also supported by an American Heart Association Grant-in-Aid, the Council for Tobacco Research, NIH grant HL-14388, and an NIH Sleep Academic Award (V.K.S.). We are indebted to Linda Bang for expert typing of this manuscript, and to Mary Clary, RN, for technical assistance.
- Received June 5, 1996.
- Revision received October 3, 1996.
- Accepted October 23, 1996.
- Copyright © 1997 by American Heart Association
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 sympatho-vagal interaction in man and conscious dog. Circ Res.. 1986;59:178-193.
Malliani A, Pagani M, Lombardi F, Cerutti S. Cardiovascular neural regulation explored in the frequency domain. Circulation.. 1991;84:482-492. Research Advances Series.
Malliani A, Pagani M, Lombardi F. Power spectrum analysis of heart rate variability: a tool to explore neural regulatory mechanisms. Br Heart J.. 1994;71:1-2, Editorial.
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.
van de Borne P, Nguyen H, Biston P, Linkowski P, Degaute JP. Effects of wake and sleep stages to the 24-hour autonomic control of blood pressure and heart rate in recumbent men. Am J Physiol. 1994;266(Heart Circ Physiol 35):H548-H554.
Montano N, Ruscone TG, 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.
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.
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.
Leimbach WN, Wallin BG, Victor RG, Aylward PE, Sundlof G, Mark AL. Direct evidence from intraneural recordings for increased central sympathetic outflow in patients with heart failure. Circulation.. 1986;73:913-919.
Mortara A, La Rovere MT, Signorini MG, Pantaleo P, Pinna G, Martinelli L, Ceconi C, Cerutti S, Tavazzi L. Can power spectral analysis of heart rate variability identify a high risk subgroup of congestive heart failure patients with excessive sympathetic activation? A pilot study before and after heart transplantation. Br Heart J.. 1994;71:422-430.
Guzzetti S, Cogliati C, Turiel M, Crema C, Lombardi F, Malliani A. Sympathetic predominance followed by functional denervation in the progression of chronic heart failure. Eur Heart J.. 1995;16:1100-1107.
Nielson CP, Vestal RE. α-Adrenoceptors and aging. In: Amery A, Staessen J, eds. Handbook of Hypertension. New York, NY: Elsevier Science Publishing Co; 1989;12:51-62.
Bristow MR, Andreson FL, Port DJ, Sker IL, Hershberger RE. Differences in β-adrenergic neuroeffector mechanisms in ischemic versus idiopathic dilated cardiomyopathy. Circulation.. 1991;84:1024-1039.
Mark AL, Victor RG, Nerhed G, Wallin BG. Microneurographic studies of the mechanisms of sympathetic nerve responses to static exercise in humans. Circ Res.. 1985;57:461-469.
Anderson EA, Sinkey CA, Lawton WJ, Mark AL. Elevated sympathetic nerve activity in borderline hypertensive humans. Hypertension.. 1989;14:177-183.
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. Comput Cardiol. 1983:205-208.
Ferguson DW, Abboud FM, Mark AL. Selective impairment of baroreflex-mediated vasoconstrictor responses in patients with ventricular dysfunction. Circulation.. 1984;69:451-460.
Casadei B, Conway J, Forfar C, Sleight P. Effect of low doses of scopolamine on RR interval variability, baroreflex sensitivity, and exercise performance in patients with chronic heart failure. Br Heart J.. 1996;75:274-280.
Seals DR, Suwarno NO, Dempsey JA. Influence of lung volume on sympathetic nerve discharge in normal humans. Circ Res.. 1990;167:130-141.
Porter TR, Eckberg DL, Fritsch JM, Rea RF, Beightol LA, Schmedtje JF, Mohanty PK. Autonomic pathophysiology in heart failure patients: sympathetic-cholinergic interrelations. J Clin Invest.. 1990;85:1362-1371.