AJP - Heart Ad Instruments
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Am J Physiol Heart Circ Physiol 284: H584-H597, 2003. First published October 24, 2002; doi:10.1152/ajpheart.00602.2002
0363-6135/03 $5.00
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
284/2/H584    most recent
00602.2002v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Barman, S. M.
Right arrow Articles by Gebber, G. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Barman, S. M.
Right arrow Articles by Gebber, G. L.
Vol. 284, Issue 2, H584-H597, February 2003

Basis for the cardiac-related rhythm in muscle sympathetic nerve activity of humans

Susan M. Barman1, Paul J. Fadel2, Wanpen Vongpatanasin2, Ronald G. Victor2, and Gerard L. Gebber1

1 Department of Pharmacology and Toxicology, Michigan State University, East Lansing, Michigan 48824; and 2 Department of Internal Medicine, University of Texas Southwestern Medical Center, Dallas, Texas 75390


    ABSTRACT
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

We tested the hypothesis that the cardiac-related rhythm in muscle sympathetic nerve activity (MSNA) of humans reflects entrainment of a central oscillator by pulse-synchronous baroreceptor nerve activity. Partial autospectral analysis was used to mathematically remove the portion of cardiac-related power in MSNA autospectra that was attributable to its linear relationship to the ECG. In 54 of 98 cases, >= 15% of cardiac-related power remained after partialization with the ECG; peak residual cardiac-related power was often at a frequency different than heart rate. When assessed on a cardiac-related burst-by-burst basis, there was a progressive and cyclic change in the ECG-MSNA interval (delay from R wave to peak of cardiac-related burst) on the time scale of respiration in four subjects. In these subjects, as well as in some in which the interval appeared to change randomly, there was an inverse relationship between the ECG-MSNA interval and cardiac-related burst amplitude. However, in 45% of the cases, these parameters were not related. These results support the view that the cardiac-related rhythm in MSNA reflects forcing of a nonlinear oscillator rather than periodic inhibition of unstructured, random activity.

baroreceptor-induced entrainment; coherence analysis; nonlinear oscillator; partial autospectral analysis; time-series analysis


    INTRODUCTION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

A HALLMARK of muscle sympathetic nerve activity (MSNA) in resting humans is the appearance of pulse-synchronous bursts of variable amplitude that occur sporadically or in short sequences (9, 10, 25, 33, 43, 44, 46, 50, 52). The waxing and waning in amplitude of these cardiac-related bursts correlates with spontaneous fluctuations in blood pressure, with periods of enhanced MSNA preceded by episodes of reduced blood pressure (44, 46, 51, 52). In some subjects, the variations in cardiac-related burst amplitude are on the time scale of the respiratory cycle (9, 10, 25, 38). Eckberg et al. (10) reported that the maximum MSNA occurred at end expiration and minimum activity occurred at end inspiration.

Although the idea that cyclic movements (locomotion) are generated by sensory feedback from muscles (21) rather than a central pattern generator was dismissed by the early 1980s (23, 24), an analogous theory persists in recent literature (27, 35, 49) for the cardiac-related rhythm in the discharges of sympathetic nerves. Specifically, the cardiac-related rhythm is often attributed simply to the waxing and waning of central inhibition induced by pulse-synchronous baroreceptor nerve activity (1, 9, 15, 22, 27, 35, 49, 51-53). In support of this view, Wallin and colleagues (9, 15, 44) reported that there was a nearly constant delay from the R wave of the ECG to the peak of the corresponding sympathetic burst in MSNA of humans. They noted that the mean ECG-MSNA interval for an individual was stable at rest, independent of heart rate, and dependent on the height of the subject. They equated this interval with the latency of baroreceptor-induced inhibition of MSNA (9, 15, 31, 43). After correction for the reflex delay, the profile of MSNA was described as an inverse mirror image of the arterial pressure signal (44, 50, 51). A modest reduction (8-10%) in the mean reflex latency has been noted during some procedures, including the Valsalva maneuver, deep breathing, and arousal from sleep (14, 39, 55). In an assessment of cardiac-related burst-by-burst changes in the ECG-MSNA interval during prolonged expiratory apnea or lower body negative pressure, Wallin et al. (49) determined that the MSNA burst amplitude increased as the interval became shorter. Although not ruling out changes within central neural circuits, they speculated that the inverse relationship could be explained entirely by recruitment of faster-conducting postganglionic fibers leading to shortening of the ECG-MSNA interval and larger-amplitude sympathetic bursts.

Several characteristics of the relationship between the arterial pulse and sympathetic nerve discharge (SND) in anesthetized cats are not consistent with the view that the cardiac-related rhythm is the mere consequence of periodic inhibition of randomly generated activity by baroreceptor nerve activity [for reviews, see Barman and Gebber (4) and Gebber (18)]. Rather, they are consistent with the generic properties of a forced nonlinear oscillator (11, 20, 40, 41, 48, 54) whose intrinsic frequency is close enough to the heart rate so as to allow for its entrainment in a 1:1 relationship to pulse-synchronous baroreceptor nerve activity. First, as demonstrated by spectral analysis, oscillations with a frequency near that of the heartbeat (2- to 6-Hz range) persist in SND after baroreceptor denervation (2). Second, the interval between the arterial pulse and the peak of the cardiac-related burst of SND is dependent on the heart rate (17). Third, Larsen et al. (36) used partial autospectral analysis to mathematically remove the portion of the cardiac-related power in SND that was attributable to its linear relationship to pulse-synchronous baroreceptor nerve activity ("theoretical baroreceptor denervation"). Importantly, in most cases, partialization of the SND autospectrum with the arterial pulse did not remove all of the power in the cardiac-related band of SND. Moreover, peak residual cardiac-related power was often at a frequency either lower or higher than the heart rate. In such cases, the forcing baroreceptor input may have been too weak and/or the difference between the heart rate and the preferred frequency of the central sympathetic oscillator may have been too great to allow for complete capture (strict phase locking) of the centrally generated rhythm (11, 20, 40, 48). Fourth, Lewis et al. (37) used time-series analysis to measure cycle-by-cycle changes in the phase lag of the peak of the cardiac-related burst in SND relative to the peak of the preceding arterial pulse. They showed that, in many cases, there was a progressive and systematic change in the phase lag of SND relative to the arterial pulse on the time scale of the respiratory cycle. Phase slippage in other systems has been attributed to the loss of entrainment of a nonlinear oscillator to its forcing input, leading to phase and frequency desynchronization (11, 26, 30). Taken together, such findings support the hypothesis that the cardiac-related rhythm in SND of cats reflects baroreceptor-induced entrainment of a nonlinear oscillator by pulse-synchronous baroreceptor nerve activity.

In the current study, we applied analytic techniques similar to those used in studies in cats (36, 37) to test the hypothesis that the cardiac-related rhythm in MSNA of humans reflects baroreceptor-induced entrainment of a central sympathetic oscillator. The composite findings support this view as opposed to the classical theory of the origin of this rhythm.


    METHODS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

Subjects

All data used for analyses were available from studies conducted at the University of Texas Southwestern Medical Center (Dallas, TX). The Institutional Review Board approved the protocols used, and each subject provided their informed written consent to participate. Data were collected from 32 volunteers (12 men and 20 women; means ± SE: age, 42.8 ± 2.6 yr; height, 169.3 ± 1.6 cm; weight, 82.0 ± 3.8 kg). Eleven of the female subjects had stage I hypertension, whereas all other subjects were free of any known cardiovascular or respiratory disease. Subjects were instructed to refrain from smoking cigarettes and drinking alcohol or caffeine-containing beverages for at least 12 h before the experiment.

Recordings

All experiments were performed with the subject in the supine position, relaxed, and breathing spontaneously. A lead II ECG recording was used to determine heart rate and to reflect timing of pulse-synchronous activity of baroreceptor and other cardiovascular afferent fibers (14, 17, 33, 49). Respiratory movements were recorded with a strain-gauge pneumograph placed over the upper abdomen. Blood pressure was measured with an automated sphygmomanometer (Welch Allen). Postganglionic MSNA was recorded using standard microneurographic techniques (9, 25, 45-47). A tungsten microelectrode with an uninsulated tip diameter of 1-5 µm was inserted into the right or left peroneal nerve near the fibular head. A reference electrode with a larger uninsulated tip was inserted subcutaneously ~2 cm from the recording electrode. The nerve signal was processed by a preamplifier and an amplifier (model 662C-3, Nerve Traffic Analyzer, University of Iowa Bioengineering; Iowa City, IA) with a total gain of 90,000. MSNA was bandpass filtered (700-2,000 Hz), rectified, and integrated by a resistance-capacitance circuit (time constant, 0.1 s). In contrast to skin sympathetic nerve activity, MSNA is characterized by cardiac-related bursts that increase in frequency with end-expiratory breath holds and Valsalva maneuvers, and MSNA is not affected by arousal or skin stroking (46, 51).

Experimental Procedures

Recordings were made at baseline and during an intravenous infusion of sodium nitroprusside (SNP) or a cold pressor test (CPT). SNP infusions started at 0.5 µg/min and were increased by 0.5 µg/min every 3 min until mean blood pressure decreased 15 mmHg or a maximum dose of 4 µg/min was reached. CPT was performed by immersing the subject's left hand up to the wrist in ice water for 2 min. Infusion of SNP increases MSNA as a consequence of baroreceptor unloading (7, 45), whereas CPT is used both clinically and experimentally to evaluate nonbaroreceptor reflex-mediated activation of MSNA (7, 13, 34, 47).

Data Analyses

Data were originally saved using a model 4000 PCM Recording Adaptor (AR Vetter, Rebersburg, PA) and a model SLV-750HF Sony videocassette recorder. Data were played back and acquired (1- or 5-ms sampling intervals) with software and an analog-to-digital converter board from RC Electronics (Santa Barbara, CA) or Axon Instruments (Union City, CA) onto a Dell Optiplex GX400 computer. Data were then processed with Datapac software (Run Technologies; Mission Viejo, CA); linear smoothing (time constant, 100 ms) was applied to the MSNA signal, and dynamic demeaning was used to remove baseline shifts in the ECG signal. These manipulations minimized errors in detecting the timing of peaks in MSNA and the R wave of the ECG for time-series analysis (see Time-series analysis).

Frequency-domain analysis. Fast Fourier transform was used to construct autospectra of MSNA, ECG, and the respiratory signal and coherence functions (normalized cross-spectra) relating pairs of these signals (19, 32). A 50-Hz sample rate was used to construct spectra having a resolution of 0.05 Hz/bin. The autospectra and coherence functions, displayed on a scale of 0-5 Hz, were averages of 20-85 20-s data windows. For data blocks <400 s, data windows were overlapped to have a minimum of 20 windows in the average. The autospectrum of a signal shows how much power (voltage squared) is present at each frequency, and the coherence function measures the strength of linear correlation of two signals. According to Benignus (5), a coherence value of 0.1 signifies a statistically significant relationship between two signals when at least 20 windows are averaged. De Boer et al. (8) used a coherence value of 0.5 to indicate significance. In the current study, the ECG-MSNA coherence value at the frequency of the heartbeat was >= 0.5 in 94 of 98 data blocks. A perfect linear relationship would have a coherence value of 1.0; a statistically significant coherence value <1.0 would occur if either the linear relationship was obscured to some extent by noise in the system or the relationship between the signals was, in part, nonlinear. These alternatives can be distinguished by using the algorithm of Jenkins and Watts (29) to perform partial autospectral analysis.

Partial autospectral analysis involves the mathematical elimination of the portion of one signal (S1) that is predictable on the basis of its linear relationship to a second signal (S2). The partial autospectrum at a given frequency (f) is defined as
AS<SUB>S1&cjs0823;  S2</SUB>(f) = AS<SUB>S1</SUB>(f)[1 − Coh<SUB>S1-S2</SUB>(f)]
where ASS1/S2 is the autospectrum of S1 partialized by S2, ASS1 is the ordinary autospectrum of S1, and CohS1-S2 is the ordinary coherence between signals S1 and S2. If the power in ASS1 at f is entirely predicted by S2, then partialization with S2 will remove all the power in S1 at that frequency. If, however, power in ASS1 at f is not fully predicted by S2, then residual power will be present in ASS1/S2 at that frequency (29, 42). Residual power in a peak exceeding background level indicates that the relationship between S1 and S2 is not strictly linear. Such would be expected in the case when the frequency of a nonlinear oscillator is attracted to but does not reach that of its forcing input (36, 42).

A macro written in Microsoft Excel 7.0 was used to measure the power above background activity in the cardiac-related band of the MSNA autospectrum before and after partialization with the ECG. A line was fitted to connect the left and right limits of the peak surrounding the heart rate in the MSNA autospectrum (in the range over which the ECG-MSNA coherence function also contained a peak), and power in this band was calculated as the area above this line (see Fig. 5A).

Time-series analysis. Software developed in our laboratory (37) was used to construct time series of the measurements shown in Fig. 1A. We measured 1) the interval (in ms) between R waves of the ECG (R-R interval) (Fig. 1A,a), 2) the interval (in ms) between the R wave and the next peak of MSNA (Fig. 1A,b), 3) the interval (in ms) between bursts of MSNA (MSNA interburst interval) (Fig. 1A,c), and 4) the trough-to-peak amplitude of MSNA bursts (MSNA amplitude, in V) (Fig. 1A,d). An average value of MSNA amplitude was obtained for each data block (i.e., during baseline, SNP infusion, and CPT). MSNA burst amplitude was also normalized on a scale of 0-1.0, with 1.0 representing the amplitude of the largest burst in a data block.


View larger version (24K):
[in this window]
[in a new window]
  		Fig. 1.   Measurements used for time-series analysis of R-R intervals (Int) of the ECG and cardiac-related bursts of muscle sympathetic nerve activity (MSNA). A: top and middle traces show a 5-s recording of ECG and MSNA, respectively, and the bottom trace shows a 20-s record of MSNA during a baseline recording session. a, Interval between R waves of the ECG (R-R interval); b, interval between ECG and the next peak of MSNA; c, interval between bursts of MSNA (MSNA interburst intervals); d, trough-to-peak amplitude of MSNA bursts. B: histograms of the distribution of R-R intervals and MSNA interburst intervals constructed from a time series containing 1,553 cardiac cycles. C: initial peaks in histograms shown on an expanded time base.

To align the ECG with the appropriate burst of MSNA, the values of Fig. 1A,a and b, were summed (10, 14, 15, 38, 49, 55). In some cases (e.g., if heart rate was near 2 Hz), it was necessary to add the values of the two preceding cardiac intervals to the value of Fig. 1A,b to obtain a value consistent with the expected ECG-MSNA interval of ~1-1.5 s (9, 10, 15, 38, 43, 49, 53, 55).

We constructed histograms of R-R intervals and MSNA interburst intervals. As shown in Fig. 1B, the first peak in the histogram of MSNA interburst intervals closely coincided with the single peak in the histogram of R-R intervals, and subsequent peaks in the MSNA interburst interval histogram were at multiples of the cardiac cycle time. Such histograms show that the bursts of MSNA used for time-series analysis were indeed cardiac related and that there were cardiac cycles in which bursts did not occur. The total number of counts in these histograms was used to obtain a ratio of the number of cardiac-related bursts of MSNA to the number of cardiac cycles (MSNA burst/ECG).

Statistical Analysis

Data are expressed as means ± SE. Statistical analyses were done using Prism 3.0 or GraphPad Instat software. P <=  0.05 indicated statistical significance. Student's paired t-test was used to compare values of ECG-MSNA coherence at the frequency of the heartbeat, MSNA burst/ECG, MSNA amplitude, root mean square (RMS) voltage (a measure of total power in the 0- to 25-Hz band of MSNA), mean blood pressure, and heart rate during baseline and SNP or CPT. Coherence values were subjected to z-transformation before this analysis. A paired t-test was also used to compare the SDs of the distributions of R-R intervals and MSNA interburst intervals. An unpaired t-test was used to compare the differences between minimum and maximum ECG-MSNA intervals in individuals during baseline, infusion of SNP, and CPT. The Pearson product-moment correlation coefficient (r value) was used to test for a relationship between the ECG-MSNA interval and MSNA amplitude. This method was also used to test for a dependency of residual cardiac-related power in the MSNA autospectrum after partialization with the ECG (MSNA/ECG autospectrum) on the ECG-MSNA coherence value, MSNA burst/ECG, mean heart rate, mean blood pressure, and MSNA amplitude. Fisher's exact test was used to compare the likelihood of having residual cardiac-related power in the MSNA/ECG autospectrum during baseline versus infusion of SNP or CPT. chi 2-Analysis was used to compare the likelihood of having a peak at a frequency other than the heart rate in the MSNA and MSNA/ECG autospectra during baseline, SNP infusion, and CPT.


    RESULTS
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

Characteristics of MSNA

MSNA, ECG, and respiration were recorded in 27 subjects during baseline and SNP infusion and in 12 subjects during baseline and CPT. Seven subjects were studied on more than 1 day, and ten subjects were studied during both SNP infusion and CPT. Analyses were done on a total of 98 data blocks that ranged in length from 3-28 min (baseline), 3-10 min (SNP infusion), and 2 min (CPT). There were no remarkable differences in results from normotensive and hypertensive subjects, so the data from both groups were pooled.

Figures 2 and 3 show examples of the patterns of MSNA seen in these subjects. In the traces shown in Fig. 2A, cardiac-related bursts of MSNA occurred during most of the cardiac cycles in a 40-s epoch while SNP was being infused. The amplitudes of cardiac-related bursts waxed and waned on the time scale of the respiratory cycle with peak amplitude occurring during inspiration (downward deflection). The record of MSNA was not advanced relative to the respiratory signal to account for peripheral conduction delays, as is often done by investigators interested in assessing the temporal relationship between central respiratory and sympathetic circuits (10, 38). Figure 2B shows the results of spectral analysis of a 547-s data block from this individual during infusion of SNP. The two peaks in the MSNA autospectrum (Fig. 2B,a) are at the frequency of respiration (0.3 Hz) and at the frequency of the heartbeat (1.5 Hz). Coherence analysis revealed that MSNA was strongly correlated to the ECG at these two frequencies as well as at a harmonic of heart rate (Fig. 2B,b). The significant coherence between MSNA and the ECG at 0.3 Hz shows that both signals were correlated to the respiratory signal. As expected, the respiratory-MSNA coherence function contained a sharp peak at the frequency of respiration (Fig. 2B,c).


View larger version (46K):
[in this window]
[in a new window]
  		Fig. 2.   MSNA showing both a cardiac-related and respiratory-related rhythm in 1 subject during an infusion of sodium nitroprusside (SNP). A: traces (a-c) are recordings of the respiratory signal (inspiration is downward), MSNA, and ECG; time scale is 10 s/division. B: frequency-domain analysis of a 9.1-min data block. Plots (a-c) are autospectra (AS) of MSNA and coherence functions relating MSNA to ECG (ECG-MSNA) and respiration. Spectra are based on 32 20-s windows with 15% overlap; frequency resolution is 0.05 Hz/bin.



View larger version (38K):
[in this window]
[in a new window]
  		Fig. 3.   MSNA showing a cardiac-related but not a respiratory-related rhythm in 1 subject under baseline conditions. A and B: same format as in Fig. 2 except spectra are for a 5-min data block and are based on 29 20-s windows with 50% overlap. Time scale in A is 11.25 s/division.

The traces in Fig. 3A are from a baseline recording session for another individual. There were marked variations in the amplitude of cardiac-related bursts in MSNA, and the bursts occurred in clusters that were interspersed with periods of little or no activity during this 45-s epoch. MSNA was not correlated to respiration in this subject. The existence of a cardiac-related but not a respiratory-related rhythm in MSNA was formally demonstrated by spectral analysis of a 300-s data block (Fig. 3B). The MSNA autospectrum (Fig. 3B,a) contained a single sharp peak at the frequency of the heartbeat (1.05 Hz); the coherence value relating MSNA to the ECG at this frequency was 0.89 (Fig. 3B,b). Although there was considerable power in the low-frequency range of the MSNA autospectrum, the respiratory-MSNA coherence value (0.09) at the frequency of respiration (0.25 Hz) was not significantly different than zero (Fig. 3B,c).

In most cases (87 of 98), the peak in the cardiac-related band of the MSNA autospectrum was at the frequency of the heartbeat. However, in 11 cases, these two values deviated by >= 0.1 Hz (resolution of measurement, 0.05 Hz). Figure 4A compares the heart rate (as determined by the peak in the ECG autospectrum) with the frequency of the peak in the cardiac-related band of the MSNA autospectrum for these data blocks. The peak in the MSNA autospectrum could be at a frequency lower (n = 8) or higher (n = 3) than the heart rate. chi 2-Analysis showed that there was not a significant difference (P = 0.6308) in the likelihood of having a peak at a frequency distinct from the heart rate during baseline (4 of 49), SNP infusion (5 of 35), and CPT (2 of 14).


View larger version (15K):
[in this window]
[in a new window]
  		Fig. 4.   Differences between heart rate and peak frequency in cardiac-related band of MSNA. A: data from ECG and MSNA AS (11 data blocks). B: data from ECG and partialized MSNA AS (MSNA/ECG, 30 data blocks). Two of the MSNA/ECG AS contained two peaks. Solid line, line of equality; SNP, sodium nitroprusside; CPT, cold pressor test.

Counts in the histograms of R-R intervals and MSNA interburst intervals (see Fig. 1B) were used to determine MSNA burst/ECG. This value ranged from 0.09 to 0.92 in the 98 data blocks analyzed, which is comparable with the incidence of MSNA bursts/100 heartbeats reported by others (31, 43, 46, 52). The MSNA burst/ECG differed by only 0.08 ± 0.01 during multiple baseline recording sessions for 12 subjects studied on more than 1 day or twice on the same day. Likewise, the MSNA burst/ECG differed by only 0.09 ± 0.02 during infusion of SNP on 2 days in seven individuals. The reproducibility in the frequency of occurrence of cardiac-related bursts during multiple recording sessions for an individual is a common feature in studies evaluating MSNA in humans (31, 43, 46, 52). An additional indication of reproducibility is that the ECG-MSNA coherence value differed by only 0.11 ± 0.02 during multiple baseline recording sessions for 12 individuals.

There was considerably more variation in the distribution of MSNA interburst intervals than in the distribution of R-R intervals. This was the case even when one limited the comparison to instances in which there was a burst of MSNA in consecutive cardiac cycles (the first peak in the histogram of MSNA interburst intervals). This point is illustrated in Fig. 1C, which shows this segment of the histograms of R-R intervals and MSNA interburst intervals on an expanded time base. We compared this portion of the histograms for 42 data blocks in which bursts of MSNA frequently occurred in consecutive cardiac cycles. There was a statistically significant difference (P < 0.001, paired t-test) between the SDs of the distributions of R-R intervals and MSNA interburst intervals (53 ± 3 vs. 87 ± 3 ms, respectively).

Table 1 shows the effects of infusion of SNP and CPT on mean blood pressure, mean heart rate, and MSNA. Mean blood pressure was significantly decreased during infusion of SNP and significantly increased during CPT, whereas heart rate was significantly increased by both procedures. In all 98 data blocks, there was a cardiac-related rhythm in MSNA, as evidenced by an ECG-MSNA coherence value at the frequency of the heartbeat that ranged from 0.27 to 0.92. The coherence value was >= 0.5 in 94 of 98 data blocks (see Fig. 6). The mean ECG-MSNA coherence value at the frequency of the heartbeat during SNP infusion and CPT was not significantly different from that during baseline. Both procedures significantly increased the MSNA burst/ECG and MSNA trough-to-peak amplitude (in V), which led to significant increases in RMS voltage (a reflection of total power) in MSNA.

                              
View this table:
[in this window]
[in a new window]
  		Table 1.   Effects of SNP infusion and CPT on cardiovascular parameters and MSNA

Although there was often a peak in the MSNA autospectrum at a frequency <0.5 Hz, this band of activity was significantly correlated to the respiratory signal (Resp-MSNA coherence value: 0.47 ± 0.03) in only 36 of 98 data blocks analyzed. These cases included recordings during baseline (n = 13), SNP infusion (n = 20), and CPT (n = 3).

Partialization of the MSNA Autospectrum with the ECG

Figure 5 shows two examples of partialization of the MSNA autospectrum with the ECG. The MSNA autospectrum (Fig. 5A,b, solid line) contained a band of power with a peak at the frequency of the heartbeat (1.55 Hz), as discerned from the ECG autospectrum (Fig. 5A,a); the ECG-MSNA coherence value at this frequency was 0.61 (Fig. 5A,c). After partialization of the MSNA autospectrum with the ECG, there was residual cardiac-related power (46% of control). There were two peaks in this frequency band of the MSNA/ECG autospectrum (Fig. 5A,b, shaded area): one at a value lower (1.35 Hz) and one at a value higher (1.70 Hz) than the heart rate. Residual cardiac-related power in the MSNA/ECG autospectrum was arbitrarily defined as >= 15% of the power that occurred in this band in the MSNA autospectrum before partialization. Such was the case for 54 of 98 data blocks; residual cardiac-related power was 36 ± 2% of control (range: 15-69% of control). In the other cases, partialization of the MSNA autospectrum with the ECG virtually eliminated cardiac-related power (6 ± 1% of control, n = 44). In the example shown in Fig. 5B, the ECG-MSNA coherence value at the frequency of the heartbeat (1.10 Hz) was 0.87. 


View larger version (28K):
[in this window]
[in a new window]
  		Fig. 5.   Partialization of MSNA AS with ECG in 2 subjects (A and B). Plots show ECG AS (a), superposition of MSNA (solid line) and MSNA/ECG (thin dashed line and shaded area) AS (b), and coherence function relating MSNA to ECG (c). The area above the thick dashed line in A,b is the cardiac-related power in AS of MSNA. Spectra in A are for a 3-min data block during baseline and are based on 21 20-s windows with 55% overlap; spectra in B are for a 18.3-min data block during baseline and are based on 55 nonoverlapping 20-s windows. Data in B are from same recording session as in Fig. 1.

There was a greater likelihood (P = 0.0213, Fisher's exact test) of having residual cardiac-related power in the MSNA/ECG autospectrum during CPT (11 of 14 cases) than during baseline (4 of 14 cases). There was a tendency (P = 0.0534) for SNP infusion to increase the likelihood of having residual cardiac-related power (15 of 35 cases at baseline vs. 24 of 35 cases during SNP). There was no difference (P = 0.5185) in the likelihood of having residual cardiac-related power during baseline for these two groups. The values of residual cardiac-related power were similar (within 8 ± 2% of each other, n = 18 pairings) during multiple baseline recording sessions for individuals studied on more than 1 day or twice on the same day; it was also similar (within 10 ± 2%) during infusion of SNP on different days (n = 8 pairings).

As shown in Fig. 4B, in 30 of 54 MSNA/ECG autospectra, the peak(s) in the band of residual cardiac-related power occurred at a frequency that differed from heart rate by at least 0.1 Hz. The peak could be at a frequency lower (n = 27) or higher (n = 5) than the heart rate; two of the MSNA/ECG autospectra contained two peaks. chi 2-Analysis showed that there was not a significant difference (P = 0.2400) in the likelihood of having a peak at a frequency distinct from the heart rate during baseline (13 of 19 spectra), SNP infusion (10 of 24 spectra), and CPT (7 of 11 spectra).

We assessed the impact of several factors on the magnitude of cardiac-related power in MSNA/ECG autospectrum. As shown in Fig. 6, there was a significant inverse relationship (r = -0.6223, P < 0.0001) between the ECG-MSNA coherence value at the frequency of the heartbeat and the magnitude of cardiac-related power in the MSNA/ECG autospectra. However, there was no relationship between the magnitude of cardiac-related power and MSNA burst/ECG (r = -0.1122, P = 0.2713), mean heart rate (r = 0.1699, P = 0.0945), mean blood pressure (r = -0.0182, P = 0.8591), or MSNA amplitude (r = 0.0208, P = 0.8392).


View larger version (18K):
[in this window]
[in a new window]
  		Fig. 6.   Relationship between ECG-MSNA coherence value at frequency of the heartbeat and cardiac-related power in MSNA/ECG AS expressed as a percentage of that in the control MSNA AS (98 data blocks). The Pearson correlation coefficient (r), level of statistical significance (P), and regression line (solid line) are shown.

Measurements of the ECG-MSNA Interval

We made cardiac-related burst-by-burst measurements of the ECG-MSNA interval (sum of Fig. 1A,a and b) in 42 data blocks (2-30 min) from 22 subjects during baseline (n = 14), infusion of SNP (n = 23), and CPT (n = 5). This analysis was limited to data blocks in which the MSNA burst/ECG was at least 0.6. The mean ECG-MSNA interval averaged 1,302 ± 8 ms (range: 1,172-1,399 ms). We quantified the variability of the ECG-MSNA interval within each of the 42 data blocks. The difference between the minimum and maximum values of ECG-MSNA interval averaged 556 ± 27 ms (range: 205-806 ms). The difference was similar during baseline (610 ± 46 ms) and infusion of SNP (560 ± 35 ms); however, it was significantly less (P <=  0.01, unpaired t-tests) during CPT (387 ± 67 ms).

Patterns of Change in the ECG-MSNA Interval

The 42 data blocks used for time-series analysis were divided into smaller segments (n = 223) of at least 40 s containing 36-100 ECG-MSNA intervals. Two patterns of cardiac-related burst-by-burst changes in the ECG-MSNA interval were seen. Figures 7C and 8C show two examples of the most common pattern, which was characterized by apparently random variations in the interval. The second pattern was seen in four subjects and was characterized by a progressive and systematic change in ECG-MSNA interval on the time scale of respiration. An example of this pattern of cardiac-related burst-by-burst variation in the ECG-MSNA interval is shown in Fig. 9C for a portion of the data segment shown in Fig. 2A. Note that the shortest interval during each respiratory cycle occurred near end-inspiration. The R-R interval remained relatively constant (Fig. 9B).


View larger version (13K):
[in this window]
[in a new window]
  		Fig. 7.   Time series of the ECG-MSNA interval (C) and trough-to-peak MSNA amplitude (MSNA Amp; D) for a subject in which these parameters changed in an apparently random fashion during baseline. B: R-R intervals included in the measurements of ECG-MSNA intervals. A: respiratory (Resp) signal for same 80-s data block. Plot in E shows that the ECG-MSNA interval and MSNA amplitude were inversely related; the Pearson correlation coefficient, level of statistical significance, and regression line (solid line) are shown.



View larger version (12K):
[in this window]
[in a new window]
  		Fig. 8.   Time series in which the ECG-MSNA interval and MSNA amplitude changed in an apparently random fashion and were not correlated. The sequence of traces is the same as in Fig. 7; data are from same baseline recording session as in Figs. 1 and 5B.



View larger version (12K):
[in this window]
[in a new window]
  		Fig. 9.   Time series in which the ECG-MSNA interval and MSNA amplitude were inversely related and changed on the time scale of the respiratory cycle. The sequence of traces is the same as in Fig. 7. Plot in E is for a 40-s data block (same as in Fig. 2A) that includes the 15-s time series in C and D. Data were collected during an infusion of SNP.

Relationship Between the ECG-MSNA Interval and MSNA Cardiac-Related Burst Amplitude

ECG-MSNA interval (sum of Fig. 1A,a and b) and MSNA cardiac-related burst amplitude (Fig. 1A,d) were inversely related in 123 of 223 data segments; r values averaged -0.5296 ± 0.0134 (range: from -0.2607 to -0.8365). The example of an inverse relationship between these parameters shown in Fig. 7E is for a subject in which both the ECG-MSNA interval (Fig. 7C) and cardiac-related burst amplitude (Fig. 7D) changed in an apparently random fashion, unrelated to the phases of the respiratory cycle. The example of an inverse relationship shown in Fig. 9E is for a subject in which the changes in the ECG-MSNA interval and cardiac-related burst amplitude were respiratory related. The plot of 51 data points shown in Fig. 9E is for a 40-s data block; however, to more easily visualize the respiratory-related pattern of the changes in the ECG-MSNA interval and cardiac-related burst amplitude, the time series in Fig. 9, C and D, shows only 15 s of this data block.

In the other 100 data blocks, there was not a significant relationship between the ECG-MSNA interval and cardiac-related burst amplitude. As shown by the representative example in Fig. 8, the changes in these parameters occurred in an apparently random fashion.


    DISCUSSION
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

This is the first study to specifically address the hypothesis that the cardiac-related rhythm in MSNA of humans is due to baroreceptor-induced forcing of a central sympathetic oscillator whose frequency is close to the heart rate. A key observation in support of this view is the existence of residual cardiac-related power in the MSNA/ECG autospectrum in over 50% of the data blocks analyzed. Residual cardiac-related power rising above background results from activity in this frequency band that is not linearly related to pulse-synchronous activity of baroreceptor and other cardiovascular afferent fibers as reflected by the ECG (30, 36). Importantly, the peak in the cardiac-related band of the MSNA/ECG autospectrum was often at a frequency distinct from (usually lower than) the heart rate. On occasion, this was also the case for the peak in the MSNA autospectrum. Such findings imply that a central sympathetic oscillator was attracted to the frequency of the heartbeat, although strict phase locking was not achieved. This type of relationship is generally referred to as "relative coordination" (30, 48), a condition that may arise as a consequence of a relatively weak forcing input and/or too great of a difference between heart rate and the preferred intrinsic frequency of the central oscillator. We also found cases in which partialization of the MSNA autospectrum with the ECG virtually eliminated cardiac-related power. Within the framework of our hypothesis, such cases would reflect a state of strict phase locking or "absolute coordination" of a nonlinear oscillator to its forcing input (30, 48).

Other findings of the current study are also consistent with the view that cardiac-related bursts are forced oscillations rather than the consequence of a near-constant onset latency of baroreceptor-induced inhibition of randomly generated sympathetic activity, as has been suggested by others (9, 15, 49, 51-53). In four subjects, an analysis of cardiac-related burst-by-burst changes in the ECG-MSNA interval revealed a progressive and systematic change in the interval on the time scale of the respiratory cycle. Within the context of the forced oscillator hypothesis, this could be explained in two ways. Episodes of progressive, cyclic changes in the ECG-MSNA interval might reflect respiratory-related resetting of the angle of strict phase locking of the sympathetic oscillator to its forcing baroreceptor input. The degree of resetting would change on a cardiac-related burst-by-burst basis as a function of an ever-changing level of respiratory input to the circuit governing baroreceptor-sympathetic coordination. Alternatively, progressive changes in the ECG-MSNA interval might reflect phase slippage of the sympathetic oscillator relative to its forcing baroreceptor input, as might occur in relative coordination. Progressive phase slippage is expected when the frequency of an oscillator is stable but somewhat different from that of its weak, periodic forcing input (30, 48).

Variations in the ECG-MSNA interval most often occurred in an apparently random fashion. In cases in which the peak in the MSNA and MSNA/ECG autospectra was close to but not the same as heart rate, we presume that an unstable central oscillator with a fluctuating frequency was attracted to but not strictly phase locked to its forcing baroreceptor input. Under this condition, residual cardiac-related power would be expected in the MSNA/ECG autospectra, but the ECG-MSNA interval would vary erratically on a cardiac-related burst-by-burst basis depending on the momentary frequency of the unstable central oscillator. Although the characteristics of human MSNA deprived of baroreceptor control have not been well described (16), the frequency of sympathetic bursts in baroreceptor-denervated cats varies between 2 and 6 Hz (2, 4, 18). Thus the central oscillator that is normally entrained 1:1 to the cardiac cycle in the cat is inherently unstable in the absence of baroreceptor input.

The difference between the minimum and maximum ECG-MSNA interval in individual data blocks was considerable, averaging 552 ms. This is far from a "near-constant latency," which would be predicted on the basis of the classic model of generation of the cardiac-related rhythm by periodic inhibition of randomly generated MSNA (9, 15, 49, 51-53). Others (14, 39, 49, 55) have reported smaller, albeit significant, variations (200-300 ms) in this interval during expiratory apnea, the Valsalva maneuver, and spontaneous arousals from sleep. To reconcile these data with the classic view for generation of cardiac-related bursts in MSNA, Wallin et al. (49) suggested that the variations in the ECG-MSNA interval could be explained by the "size principle" originally used by Henneman and Mendell (28) to account for the order of recruitment of somatomotor neurons. As described by Wallin and colleagues (14, 15, 49), the ECG-MSNA interval has several components. It is estimated that 100-150 ms of this interval is the time between the R wave of the ECG and arrival of pulse-synchronous baroreceptor nerve activity in the brain stem (14, 15). Central processing of baroreceptor information in the brainstem of humans has been estimated to take 250-300 ms (6, 15). There are no direct measures for spinal pathway conduction time to preganglionic sympathetic neurons to the peroneal nerve, but it is estimated to take 250-300 ms (15, 49). The remaining portion of the ECG-MSNA interval (averaging ~600 ms) is attributed to conduction along the sympathetic fibers of the peroneal nerve, with an average conduction velocity of ~1 m/s (15, 25). Because the difference between minimal and maximal ECG-MSNA intervals during lower body negative pressure was larger for the peroneal nerve in the leg (120 ms) than the radial nerve in the arm (90 ms), for which conductance distance is shorter, Wallin et al. (49) concluded that changes in the time involved in central processing, which should be the same for both nerves, could not account for the variability in the ECG-MSNA interval. Rather, they suggested that the variability in the interval was attributable to the range of axonal conduction velocities in postganglionic sympathetic fibers. Specifically, in a manner analogous to the order of recruitment of alpha -motoneurons by their excitatory inputs (28), Wallin et al. (49) proposed that as central sympathetic drive is increased, postganglionic neurons with more rapidly conducting axons are recruited in an orderly fashion. This would lead to a reduction in ECG-MSNA interval and a concomitant increase in burst amplitude. An important corollary of this is that there would be an inverse relationship between the ECG-MSNA interval and cardiac-related burst amplitude. They reported that this was a common feature of their data. In contrast, we noted that these parameters were unrelated in 45% of the data blocks analyzed. Thus the size principle cannot explain the cardiac-related burst-by-burst variations in the ECG-MSNA interval seen in many of the data blocks that we analyzed. Morgan et al. (39) and Xie et al. (55) came to the same conclusion when they reported only a very weak correlation between the ECG-MSNA interval and cardiac-related burst amplitude in studies dealing with changes in MSNA during sleep and arousal in humans.

As already mentioned, the variability of ECG-MSNA intervals during baseline or infusion of SNP in our experiments was, on the average, considerably greater than that reported by others (14, 39, 49, 55) during various procedures. It is unlikely that the variability we noted was due to inadvertent counting of noncardiac-related bursts in MSNA. First, the mean value of the ECG-MSNA interval for individuals ranged from 1,172 to 1,396, which is virtually identical to the range reported by others (9, 14, 15, 43, 49, 55). Second, as shown by the example in Fig. 1B, counts in the MSNA interburst interval histograms fell within peaks that corresponded to the R-R interval or its multiples.

Three additional findings of the current study are also consistent with the forced oscillator theory for generation of cardiac-related bursts in MSNA. First, there was considerably more variability in the MSNA interburst intervals within consecutive cardiac cycles than in R-R intervals in individual data blocks. We doubt that this difference is artifactual because measurements of the timing of MSNA peaks were made after linear smoothing, which minimizes errors in peak detection (see METHODS). Second, the ECG-MSNA coherence value at the frequency of the heartbeat never reached 1.0. Although this could be due to the existence of physiological noise in MSNA in the frequency range of the heartbeat (as reflected by background power), it can also be explained by imperfect entrainment of a nonlinear oscillator to its forcing input. Finally, the magnitude of residual cardiac-related power in the MSNA/ECG autospectra was inversely related to the ECG-MSNA coherence value at the frequency of the heartbeat. This too is consistent with the forced oscillator hypothesis because relatively high ECG-MSNA coherence values would be expected when the oscillator is strongly entrained to its forcing baroreceptor input. Strong entrainment would cause the sympathetic oscillator to have the same frequency as the heartbeat. As a result, partialization of the MSNA autospectra with the ECG would essentially eliminate cardiac-related power. In contrast, a relatively low ECG-MSNA coherence value signifies a state of weaker entrainment that allows a higher proportion of sympathetic bursts to occur at a frequency somewhat different than heart rate. Under this situation, there would be considerable residual cardiac-related power in the MSNA/ECG autospectra.

We expected that the fall in mean blood pressure during the infusion of SNP would lead to a decrease in the ECG-MSNA coherence value at the frequency of the heartbeat and an increase in the percent of residual cardiac-related power. However, this was not the case. Perhaps the ECG-MSNA coherence value did not change because the reduced level of pulse-synchronous baroreceptor nerve activity during the fall in mean blood pressure was counterbalanced by an increase in the number of cardiac cycles in which there was a burst of MSNA (i.e., MSNA burst/ECG). An increase in the ratio by itself would increase the coherence value (3). Nonetheless, there was a tendency (although not statistically significant) for SNP to increase the likelihood (24 of 35 cases) of having residual cardiac-related power compared with baseline conditions (15 of 35 cases).

The fact that the ECG-MSNA coherence value at the frequency of the heartbeat was not increased during the rise in mean blood pressure produced by CPT might be explained as follows. Enhanced pulse-synchronous baroreceptor nerve activity during the pressor response would be expected to increase the coherence value. However, this effect might have been counterbalanced by increased central sympathetic drive leading to the increase in mean blood pressure that would have made it more difficult for baroreceptor nerve activity to entrain a "hyperexcitable" central oscillator. Nevertheless, we did note that there was a significantly greater likelihood of having residual cardiac-related power during CPT (11 of 14 cases) than during baseline (4 of 14 cases). Also, in most (7 of 11) cases during CPT, the peak of the residual cardiac-related power was at a frequency lower or higher than heart rate. Both of these observations are consistent with an increased tendency for a "hyperexcitable" oscillator to escape entrainment to its forcing input.

Although ours is the first study to specifically address the possibility that cardiac-related bursts in MSNA of humans reflect the entrainment of an oscillator, evidence in support of this notion can be found in studies from other laboratories. If cardiac-related bursts in MSNA resulted simply as the consequence of periodic inhibition of randomly generated activity (9, 15, 49, 51-53), then interruption of the baroreceptor input should result in the elimination of synchronous bursts of MSNA. In contrast, Fagius et al. (16) reported that MSNA still occurred in bursts that were not synchronized to the heartbeat during short-lasting baroreceptor deafferentation (bilateral local anesthesia of the glossopharyngeal and vagus nerves in the neck). On this basis, Wallin and Fagius (50) suggested that the characteristic pattern of MSNA "is brought about by recurrent baroreceptor inhibition entraining the bursts in the cardiac rhythm." The predominant burst frequency of MSNA during interruption of baroreceptor reflexes was 0.4-0.7 Hz in these subjects whose baseline heart rates were 0.8 and 1.1 Hz. This finding is consistent with our observation that the peak in the band of residual cardiac-related power in the MSNA/ECG autospectra was most often at a frequency less than the heart rate. Fagius (12) also suggested that MSNA had an inherent characteristic of occurring in bursts because, without fail in 10 subjects, the sympathetic burst after an extrasystole was terminated considerably earlier than expected on the basis of the baroreceptor reflex latency. The ECG-MSNA interval was reduced by as much as 51% under these conditions. Wallin et al. (49) also noted a similar phenomenon at the end of a prolonged expiratory apnea in some subjects. Xie et al. (55) reported that the mean ECG-MSNA interval was shortened upon spontaneous arousals from slow wave sleep and subsequent to the appearance of spontaneous K complexes in the electroencephalogram. Finally, Kocsis et al. (33) reported that a peak often remained at or near the frequency of the heartbeat in the left MSNA-to-right MSNA coherence function after partialization with the ECG or the arterial pulse. This suggests that the correlation between these two signals in the cardiac-related frequency band was not solely attributable to shared pulse-synchronous baroreceptor input.

In summary, the combined results of frequency-domain and time-domain analyses in the current study support the view that, as is the case for SND in the cat (4, 18), the cardiac-related rhythm in MSNA of humans reflects entrainment of a central sympathetic oscillator to pulse-synchronous baroreceptor nerve activity.


    ACKNOWLEDGEMENTS

The authors thank Lisa M. Braybrook for assistance with data compilation.


    FOOTNOTES

This study was supported by National Institutes of Health Grants HL-33266 (to S. M. Barman), HL-06296 (to R. G. Victor), DA-10064 (to R. G. Victor), HL-69648 (an Individual National Research Service Award to P. J. Fadel), and K23-RR-016321 (to W. Vongpatanasin).

Address for reprint requests and other correspondence: S. M. Barman, Dept. of Pharmacology and Toxicology, Michigan State Univ., East Lansing, MI 48824 (E-mail: barman{at}msu.edu).

The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

First published October 31, 2002;10.1152/ajpheart.00602.2002

Received 15 July 2002; accepted in final form 18 October 2002.


    REFERENCES
TOP
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
REFERENCES

1.   Adrian, ED, Bronk DW, and Phillips G. Discharges in mammalian sympathetic nerves. J Physiol 74: 155-133, 1932.

2.   Barman, SM, and Gebber GL. Sympathetic nerve rhythm of brain stem origin. Am J Physiol Regul Integr Comp Physiol 239: R42-R47, 1980[Free Full Text].

3.   Barman, SM, and Gebber GL. Rostral ventrolateral medullary and caudal medullary raphe neurons with activity correlated to the 10-Hz rhythm in sympathetic nerve discharge. J Neurophysiol 68: 1535-1547, 1992[Abstract/Free Full Text].

4.   Barman, SM, and Gebber GL. `Rapid' rhythmic discharges of sympathetic nerves: sources, mechanisms of generation, and physiological relevance. J Biol Rhythms 15: 365-379, 2000[Abstract/Free Full Text].

5.  Benignus VA. Estimation of the coherence spectrum and its confidence interval using the fast Fourier transform. AU-17: 145-150, 1969.

6.   Borst, C, and Karemaker JM. Time delays in the human baroreceptor reflex. J Auton Nerv Syst 9: 399-409, 1983[Web of Science][Medline].

7.   Cui, J, Wilson THE, and Crandall CG. Baroreflex modulation of muscle sympathetic nerve activity during cold pressor test in humans. Am J Physiol Heart Circ Physiol 282: H1717-H1723, 2002[Abstract/Free Full Text].

8.   De Boer, RW, Karemaker JM, and Strackee J. Relationships between short-term blood-pressure fluctuations and heart-rate variability in resting subjects. I: a spectral analysis approach. Med Biol Eng Comput 23: 352-358, 1985[Web of Science][Medline].

9.   Delius, W, Hagbarth KE, and Wallin BG. General characteristics of sympathetic activity in human muscle nerves. Acta Physiol Scand 84: 65-81, 1972[Web of Science][Medline].

10.   Eckberg, DL, Nerhed C, and Wallin BG. Respiratory modulation of muscle sympathetic and vagal cardiac outflow in man. J Physiol 365: 181-196, 1985[Abstract/Free Full Text].

11.   Ermentrout, GB, and Rinzel J. Beyond a pacemaker's entrainment limit: phase walkthrough. Am J Physiol Regul Integr Comp Physiol 246: R102-R106, 1984[Abstract/Free Full Text].

12.   Fagius, J. Muscle nerve sympathetic activity following ectopic heart beats-a note on the burst pattern of sympathetic impulses. J Auton Nerv Syst 22: 243-245, 1998.

13.   Fagius, J, Karhuvaara S, and Sundlöf G. The cold pressor test: effects on sympathetic nerve activity in human muscle and skin nerve fascicles. Acta Physiol Scand 137: 325-334, 1989[Web of Science][Medline].

14.   Fagius, J, Sundlöf G, and Wallin BG. Variation of sympathetic reflex latency in man. J Auton Nerv Syst 21: 157-165, 1987[Web of Science][Medline].

15.   Fagius, J, and Wallin BG. Sympathetic reflex latencies and conduction velocities in normal man. J Neurol Sci 47: 433-448, 1980[Web of Science][Medline].

16.   Fagius, J, Wallin BG, Sundlöf G, Nerhed C, and Englesson S. Sympathetic outflow in man after anesthesia of the glossopharyngeal and vagus nerves. Brain 108: 423-438, 1985[Abstract/Free Full Text].

17.   Gebber, GL. Basis for phase relations between baroreceptor and sympathetic nerve discharge. Am J Physiol 230: 263-270, 1976[Abstract/Free Full Text].

18.   Gebber, GL. Central determinants of sympathetic nerve discharge. In: Central Regulation of Autonomic Functions, edited by Loewy AD, and Spyer KM.. New York: Oxford University Press, 1990, p. 126-144.

19.   Gebber, GL, Zhong S, Barman SM, Paitel Y, and Orer HS. Differential relationships among the 10-Hz rhythmic discharges of sympathetic nerves with different targets. Am J Physiol Regul Integr Comp Physiol 267: R387-R399, 1994[Abstract/Free Full Text].

20.   Glass, L, and Mackey MC. From Clocks to Chaos: the Rhythms of Life. Princeton, NJ: Princeton University Press, 1988.

21.   Gray, J. The role of peripheral sense organs during locomotion in the vertebrates. In: Physiological Mechanisms in Animal Behavior. Cambridge, MA: Cambridge University Press, 1950, vol. 4, p. 112-126.

22.   Green, JH, and Heffron PF. Studies upon the relationship between baroreceptor and sympathetic activity. Q J Exp Physiol 53: 23-32, 1968.

23.   Grillner, S, and Zangger P. On the central generation of locomotion in the low spinal cat. Exp Brain Res 34: 241-261, 1979[Web of Science][Medline].

24.   Grillner, S, and Zangger P. The effect of dorsal root transection on the efferent motor pattern in the cat's hindlimb during locomotion. Acta Physiol Scand 120: 393-405, 1984[Web of Science][Medline].

25.   Hagbarth, KE, and Vallbo ÅB. Pulse and respiratory grouping of sympathetic impulses in human sympathetic nerves. Acta Physiol Scand 74: 96-108, 1968[Web of Science][Medline].

26.   Hanson, F. Comparative studies of firefly pacemakers. Fed Proc 37: 2158-2164, 1978[Web of Science][Medline].

27.   Hedman, AE, Matsukawa K, and Ninomiya I. Origin of cardiac-related synchronized cardiac sympathetic nerve activity in anaesthetized cats. J Auton Nerv Syst 47: 131-140, 1994[Web of Science][Medline].

28.   Henneman, E, and Mendell LM. Functional organization of motoneuron pool and its inputs. In: Handbook of Physiology. The Nervous System. Motor Control. Bethesda, MD: Am. Physiol. Soc, 1981, sect. 1, vol. II, pt. 1, chapt. 11, p. 423-507.

29.   Jenkins, GM, and Watts DG. Spectral Analysis and its Applications. San Francisco, CA: Holden Day, 1968.

30.   Kelso, JAS Dynamic Patterns: the Self-Organization of Brain and Behavior. Cambridge, MA: MIT Press, 1995.

31.   Kienbaum, P, Karlsson T, Sverrisdottir YB, Elam M, and Wallin BG. Two sites of modulation of human sympathetic nerve activity by arterial baroreceptors? J Physiol 531: 86-869, 2001.

32.   Kocsis, B, Gebber GL, Barman SM, and Kenney MJ. Relationships between activity of sympathetic nerve pairs: phase and coherence. Am J Physiol Regul Integr Comp Physiol 259: R549-R560, 1990[Abstract/Free Full Text].

33.   Kocsis, B, Karlsson T, and Wallin BG. Cardiac- and noncardiac-related coherence between sympathetic drives to muscles of different human limbs. Am J Physiol Regul Integr Comp Physiol 276: R1608-R1616, 1999[Abstract/Free Full Text].

34.   Kregel, KC, Seals DR, and Callister R. Sympathetic nervous system activity during skin cooling in humans: relationship to stimulus intensity and pain sensation. J Physiol 454: 359-371, 1992[Abstract/Free Full Text].

35.   Kunitake, T, and Hiroshi K. Discharge pattern of renal sympathetic nerve activity in the conscious trat: spectral analysis of integrated activity. J Neurophysiol 84: 2859-2867, 2000[Abstract/Free Full Text].

36.   Larsen, PD, Lewis CD, Gebber GL, and Zhong S. Partial spectral analysis of cardiac-related sympathetic nerve discharge. J Neurophysiol 84: 1168-1179, 2000[Abstract/Free Full Text].

37.   Lewis, CD, Gebber GL, Zhong S, Larsen PD, and Barman SM. Modes of baroreceptor-sympathetic coordination. J Neurophysiol 84: 1157-1167, 2000[Abstract/Free Full Text].

38.   Macefield, VG, and Wallin BG. Modulation of muscle sympathetic activity during spontaneous and artificial ventilation and apnoea in humans. J Auton Nerv Syst 53: 137-147, 1995[Web of Science][Medline].

39.   Morgan, BJ, Crabtree DC, Puleo DS, Badr MS, Toiber F, and Skatrud JB. Neurocirculatory consequences of abrupt change in sleep state in humans. J Appl Physiol 80: 1627-1636, 1996[Abstract/Free Full Text].

40.   Pavlidis, TC. Biological Oscillators: Their Mathematical Analysis. New York: Academic, 1973.

41.   Rinzel, J, and Ermentrout GB. Analysis of neural excitability and oscillations. In: Methods of Neuronal Modeling, edited by Koch C, and Segev I.. Cambridge, MA: MIT Press, 1998, chapt. 7, p. 381-447.

42.   Rosenberg, JR, Halliday DM, Breeze P, and Conway BA. Identification of patterns of neuronal connectivity: partial spectra, partial coherence and neuronal interactions. J Neurosci Methods 83: 57-72, 1998[Web of Science][Medline].

43.   Sundlöf, G, and Wallin BG. The variability of muscle nerve sympathetic activity in resting recumbent man. J Physiol 272: 383-397, 1977[Abstract/Free Full Text].

44.   Sundlöf, G, and Wallin BG. Human muscle nerve sympathetic activity at rest: relationship to blood pressure and age. J Physiol 274: 621-637, 1978[Abstract/Free Full Text].

45.   Tuncel, M, Wang Z, Arbique D, Fadel PJ, Victor RG, and Vongpatanasin W. Mechanism of the blood pressure-raising effect of cocaine in humans. Circulation 105: 1054-1059, 2002[Abstract/Free Full Text].

46.   Vallbo, ÅB, Hagbarth KE, Torebjork HE, and Hagbarth KE. Somatosensory, proprioceptive, and sympathetic activity in human peripheral nerves. Physiol Rev 59: 919-957, 1979[Free Full Text].

47.   Victor, RG, Leimbach WN, Jr, Seals DR, Wallin BG, and Mark AL. Effects of the cold pressor test on muscle sympathetic nerve activity in humans. Hypertension 9: 429-436, 1987[Abstract/Free Full Text].

48.   Von Holst, E. Relative coordination as a phenomenon and as a method of analysis of central nervous functions. In: The Collected Papers of Erich von Holst, edited by Martin R.. Coral Gables, FL: University of Miami Press, 1973, p. 33-135.

49.   Wallin, BG, Burke D, and Gandevia S. Coupling between variations in strength and baroreflex latency of sympathetic discharges in human muscle nerves. J Physiol 474: 331-338, 1994[Abstract/Free Full Text].

50.   Wallin, BG, Delius W, and Sundlöf G. Human muscle nerve sympathetic activity in cardiac arrhythmias. J Clin Lab Invest 34: 293-300, 1974[Web of Science][Medline].

51.   Wallin, BG, Delius W, and Hagbarth KE. Comparison of sympathetic nerve activity in normotensive and hypertensive subjects. Circ Res 33: 9-21, 1973[Abstract/Free Full Text].

52.   Wallin, BG, and Fagius J. The sympathetic nervous system in man-aspects derived from microelectrode recordings. Trends Neurosci 9: 63-67, 1986.

53.   Wallin, BG, Sundlöf G, and Delius W. The effect of carotid sinus nerve stimulation on muscle and skin nerve sympathetic activity in man. Pflügers Arch 358: 101-110, 1975[Web of Science][Medline].

54.   Winfree, AT. When Time Breaks Down: the Three-Dimensional Dynamics of Electrochemical Waves and Cardiac Arhythmias. Princeton, NJ: Princeton University Press, 1987.

55.   Xie, A, Skatrud JB, Puleo DS, and Morgan BJ. Arousal from sleep shortens sympathetic burst latency in humans. J Physiol 515: 621-628, 1999[Abstract/Free Full Text].

Am J Physiol Heart Circ Physiol 284(2):H584-H597
0363-6135/03 $5.00 Copyright © 2003 the American Physiological Society


This article has been cited by other articles:


Home page
 J. Physiol.Home page
W. H. Cooke, C. A. Rickards, K. L. Ryan, T. A. Kuusela, and V. A. Convertino
Muscle sympathetic nerve activity during intense lower body negative pressure to presyncope in humans
J. Physiol., October 15, 2009; 587(20): 4987 - 4999.
[Abstract] [Full Text] [PDF]

Home page
 Am. J. Physiol. Heart Circ. Physiol.Home page
P. J. Fadel, H. S. Orer, S. M. Barman, W. Vongpatanasin, R. G. Victor, and G. L. Gebber
Fractal properties of human muscle sympathetic nerve activity
Am J Physiol Heart Circ Physiol, March 1, 2004; 286(3): H1076 - H1087.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
284/2/H584    most recent
00602.2002v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Barman, S. M.
Right arrow Articles by Gebber, G. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Barman, S. M.
Right arrow Articles by Gebber, G. L.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Visit Other APS Journals Online