| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Department of Surgery and Anaesthesia, Wellington School of Medicine and Health Sciences, Wellington, New Zealand
Submitted 25 September 2006 ; accepted in final form 6 December 2006
 |
ABSTRACT |
|---|
 TOP ABSTRACT METHODSRESULTSDISCUSSIONREFERENCES |
|---|
respiratory period variability; respiratory rhythm; arterial baroreceptors; cardiorespiratory synchronization
In previous studies, examination of RI plots for horizontal bands, indicating a constant temporal alignment between R and I times, has led to the classification of CVC into patterns I, II, III, and IV on the basis of criteria related to the RI–1 interval and the ratio of HR to observed respiratory rate (fo; Fig. 1). Pattern I coupling is associated with a constant RI–1 interval and HR-to-fo ratio. Pattern II coupling is also associated with a constant RI–1 interval; however, the HR-to-fo ratio varies from breath to breath. In contrast, the RI–1 interval and HR-to-fo ratio alternate between two values during pattern III coupling. Pattern IV coupling is characterized by alternation between a "coupled phase," where the RI–1 interval is temporally aligned, as during pattern I, and a "transitional phase," where the RI–1 interval slowly increases (pattern IVhi) or decreases (pattern IVlo) from breath to breath (12). Plots of HR vs. fo for each coupling pattern showed that coupling patterns were associated with specific regions of the HR-to-fo map, suggesting that the CVC pattern might be causally related to the HR-to-fo ratio (10).
|
However, critical questions remain. What is the origin of the cardiovascular afferent signal implicated by the cardiac-trigger hypothesis? Which neural pathways contribute to this process? A prime candidate for the putative cardiac afferent signal was thought to be the arterial baroreceptors, because baroreflex afferents are known to terminate in the nucleus tractus solitarius, with secondary connections to brain stem areas involved in respiratory rhythm generation (13, 26, 31–33, 40, 41, 43). Another important question relates to the role of vagal modulation of the heart in CVC, which has yet to be fully established (2, 3, 22, 24).
Unfortunately, interpretation of experimental data is challenging. In particular, modeling studies highlighted a critical distinction between observed CVC, indicated by the presence of horizontal bands on the RI plot, and the strength of central coupling between the cardiac afferent signal and the respiratory oscillator (13). According to the model framework, if HR or fi is altered, the presence of CVC may vary without a change in strength of the cardiac afferent input. However, because neither the strength of the cardiac signal nor fi can be easily measured in vivo, it would be necessary to examine the time interval between each R wave and the following I (RI intervals) relations across entire HR-to-fo ratio domains to reliably ascertain truly absent CVC (13). This was not well appreciated by early investigators and potentially explains why previous studies of the role of the vagus nerve in rabbits produced conflicting results, with CVC apparently "abolished" in one study (3), but not in another (24), after bilateral vagotomy.
One way to deal with this challenge may be to alter the HR by electrically pacing the heart and examining CVC and CDI-I across a range of HR-to-fo ratios. This approach would enable a more definitive examination of the role of the vagus nerve and arterial baroreceptors in CVC after selective nerve sectioning. We hypothesize that elimination of the relevant neural pathway(s) should completely abolish CVC, such that no CVC will be observed at any HR-to-fo ratio.
|
METHODS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
With the animal in a fixed supine position, a tracheostomy was performed and an endotracheal tube was inserted. The right internal jugular vein was dissected free of surrounding tissue, and a bipolar cardiac pacing wire was inserted and advanced toward the heart (23). The pacing wire was connected to a square-pulse stimulator (model S88, Grass Telefactor, W. Warwick, RI) in series with a stimulus isolation unit (model SIU7, Grass Telefactor) and left on standby. A femoral artery and vein, dissected free of surrounding tissue, were cannulated (24-gauge intravenous cannulas, Becton Dickinson) for continuous recording of blood pressure, intravenous fluid replacement, and administration of vasoactive drugs. All catheters were filled with heparinized saline and secured to the surrounding skin flap with surgical sutures. After cannulation, animals were left undisturbed for 
30 min for hemodynamic stabilization. Anesthesia was thereafter maintained at adequate levels with isoflurane (2%) and supplemental O2. Body temperature was monitored with a rectal probe and maintained at 
38°C using a heat lamp.
During each experiment, ECG (lead CM5, Corometrics Neo-Trak 502), respiratory flow (Hans Rudolph heated Pneumotach 8420 A; Vacumed differential pressure transducer 4500), and continuous arterial blood pressure (P23 pressure transducer, Grass Instruments, Quincy, MA) were recorded. Data were sampled at 2,000 Hz per channel with a 16-bit input-output data acquisition board (PCI-6023E series, National Instruments) and streamed to a Macintosh G4 computer for storage and subsequent offline analysis.
Data Analysis
From the ECG and continuous blood pressure recordings, R wave times and beat-to-beat systolic blood pressures (SBP) were extracted. To ensure that heartbeats did not mechanically produce an artifact in respiratory airflow that would bias the assessment of I times, the raw airflow signal was low-pass filtered using a Butterworth filter. A known drawback of the Butterworth filter is that filtered series are phase shifted. This effect was minimized by extension of the original series at the beginning and end with its own reversed array and back-to-back filtering (25, 35). The filtered signal was superimposed on the raw signal and manually examined to ensure satisfactory removal of any signal artifacts. The I values were obtained by detection of the point of positive signal deflection following the point of zero flow at end expiration. From these time series, the RI intervals were established and then plotted as a time series (RI plot). In the RI plot, horizontal banding indicates the presence of CVC, where R waves fall in constant timing relation with I. The interval between inspiration and the immediately preceding heartbeat is designated the RI–1 interval (28, 29). The RI plots were visually examined, in a blinded fashion without reference to the HR and fo time series or the study protocol, for patterns of CVC, as previously defined, on the basis of the RI–1 interval and the HR-to-fo ratio (Table 1) (28, 29). Episodes of CVC were then verified against the entropy index of CVC (see below) and confirmed only when pattern-consistent changes in the HR-to-fo ratio and CDI-I time series were also observed.
|
) and the mean of the absolute values of consecutive breath-to-breath difference in RI–1 interval (|
|). SH, a measure of RI–1 interval dispersion, has previously been applied as an index of CVC (12, 42). This was achieved by passing an overlapping n-breath moving window over the data series and allocating the n RI–1 intervals within each window into a 10-bin histogram with outer limits set at a minimum of 0 and a maximum of the mean of the R-R intervals covered by n values of I. The 10-bin histogram was chosen to ensure comparability of results obtained with previous descriptions of CVC, whereas the choice of the value n varied depending on the total length of data being analyzed. On the basis of the histogram distribution of RI–1 intervals, SH was calculated according to the following algorithm
 | (1) |
 | (2) |
 | (3) |
of 0; in the complete absence of CVC, RI–1 intervals would be uniformly distributed, giving a maximum SH of 1. In practice, neither single-bin distribution nor complete uniformity of the intervals across the range of histogram bins is likely to occur with small data sets. By nature of its construction, SH is most sensitive in the detection of pattern I, II, and III coupling and less sensitive for pattern IV coupling, where the RI–1 interval alignment changes from breath to breath during the transitional phase. Therefore, to enhance the temporal resolution of the index, using an overlapping 40-point window, we plotted SH as a time series for data epochs containing >40 RI–1 intervals (42). The 40-point moving window was chosen to maximize the sensitivity of the index. For data epochs containing 
40 RI–1 intervals, we used a window of equivalent to the total number of RI–1 intervals in that epoch (n) to calculate SH.
The statistical threshold for significant CVC at the 5% level was determined by applying the entropy algorithm described above to random number series. Since the distribution of SH values varies according to the total number of RI–1 intervals processed (n), threshold values were generated with random series that matched n. Thus we determined the threshold (SHT) for the moving window SH by calculating the SH for 1,000 random data sets, each representing 40 RI–1 intervals (n = 40) between 0 and 1. The resulting series of SH values were ranked in ascending order, and the 50th number (0.90) was taken as SHT. Values on the SH time series that fall below SHT are considered statistically significant. Individual SHT values were calculated for data epochs containing 40 RI–1 intervals. To achieve this, we calculated SH using n random numbers between 0 and 1 matching the total number of RI–1 intervals within the epoch for 1,000 repetitions. The resulting series of SH values were ranked in ascending order, and the 50th number was taken as SHT (see 
Fig. 6).
|
|
|
|
|
|) (10). All data analysis was performed using custom-written software in LabView 7 (National Instruments) on a 1-GHz PowerBook G4 computer.
Procedures
Baroreflex testing. The Oxford method was used to test baroreflex sensitivity (BRS) (6, 37). This method involved reflex changes in HR in response to randomized 0.3-ml bolus injections of phenylephrine (1.5–3 µg/kg) or sodium nitroprusside (2–5 µg/kg) over 10 s. To derive BRS, a least squares linear regression was fit to R-R interval change in response to ramp changes in SBP induced by phenylephrine and sodium nitroprusside. The slope of the linear regression was taken as an estimate of BRS (ms/mmHg). BRS estimates were calculated separately for phenylephrine and sodium nitroprusside slopes.
Cardiac pacing.
Cardiac pacing was achieved by delivery of 20-ms pulses at 1- to 3-mA current at rates slightly higher than the intrinsic HR and increasing in 5 beat/min increments every 2 min, for a total of 10–20 min. The intrinsic HR and the individual animal's capacity to maintain blood pressure determined the total number of pacing steps. Successful pacing was confirmed, in real time, using a storage oscilloscope (model 464, Tektronix). Postmortem examination of eight animals confirmed that the tips of the pacing wire were at or near the confluence of the superior vena cava and the right atrium.
Sinoaortic denervation. Sinoaortic denervation (SAD) was achieved in a two-step procedure (27): the sternocleidomastoid muscles were retracted, and the carotid sheath and the underlying carotid triangle were exposed. With the aid of a dissecting microscope (model SZ-STU1, Olympus), the aortic depressor nerve (ADN) was identified prior to its junction with the vagus nerve and cut. In some rats, baroreceptor afferent fibers also coursed through the sympathetic chain and/or the superior laryngeal nerve (SLN). Therefore, the sympathetic trunk was identified at its origin at the nodose ganglion and sectioned, and the SLN was cut at its junction with the vagus nerve. For isolation of the carotid baroreceptors, the carotid sinus nerve (CSN) was severed at its junction with the glossopharyngeal nerve, and the internal, external, and caudal sections of the common carotid artery were completely stripped. Then all carotid branches were painted with 10% ethanol and allowed to stabilize for 60 min. Consistent with previous studies and with the inherent variability in baseline HR and SBP associated with respiration taken into account, SAD was considered complete when baroreflex testing with phenylephrine and sodium nitroprusside produced <10 beat/min reflex changes in HR (38, 39).
Vagotomy. In rats, the vagus nerves travel along the common carotid artery within the carotid sheath. Vagus nerves were dissected free from the carotid sheath and surrounding tissue and sectioned bilaterally at the caudal and rostral ends.
Sham surgery. In eight rats, sham operations were performed to assess the effects of time and surgical stress on CVC. In these experiments, rats were subjected to the same experimental procedure, including tracheostomy, surgical dissection, isolation of baroreceptor afferent nerves, and insertion of the cardiac pacing electrode. The key difference, however, is that denervation was not performed in these animals.
Protocol
A schematic representation summarizing the study protocol is shown in Fig. 2. In all protocols, 15 min of baseline data collection was followed by baroreflex testing to establish the baseline BRS. Then cardiac, respiratory, and blood pressure activity were recorded simultaneously during cardiac pacing, which enabled the examination of the relations between CVC and respiratory period variability over a range of HR-to-fo ratios. At this point, the study was split into three separate protocols, with the effects of vagotomy (protocol I), SAD (protocol II), and SAD + vagotomy (protocol III) assessed in separate experiments. Animals in protocols II and III underwent further BRS testing to determine the completeness of SAD. In all protocols, variable-rate cardiac pacing was performed to ensure that the effects of denervation on CVC and respiratory period variability were assessed across a minimum of one HR-to-fo integer ratio. In addition, sham experiments were conducted separately in eight rats that underwent similar surgical preparation but were not subjected to cardiac pacing or denervation.
Statistical Analysis
Values are means (SD). Paired data were compared using Student's paired t-test, and differences between CVC patterns were compared using Student's unpaired t-test if factorial ANOVA was found to be significant. P < 0.05 was considered statistically significant.
|
RESULTS |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Baseline recordings were visually inspected for each animal before cardiac pacing. Although horizontal banding of RI plots were observed in 48% of recordings (14 of 29 rats), the apparent degree of CVC varied between, as well as within, the data epochs. This variation is illustrated in Fig. 3, which shows 300-s recordings from two experiments. The example in Fig. 3A shows no apparent horizontal structures on the RI plot, which is consistent with the absence of CVC. In contrast, the example in Fig. 3B shows periods of horizontal banding, consistent with the presence of CVC, as well as periods with no visible structure on the RI plot. Comparison of the two epochs shows breath-to-breath HR-to-fo ratio consistently >6:1 throughout the example in Fig. 3A and 4:1 throughout the example in Fig. 3B. || and || were smaller in Fig. 3B, consistent with the presence of CVC. SH was greater than SHT throughout the recording in Fig. 3A, consistent with no apparent structure in the RI plot and relatively large-magnitude variability in CDI-I, which did not vary throughout the recording. In contrast, SH fell below SHT in Fig. 3B during pattern I coupling, which was associated with a step-wise reduction in the magnitude of CDI-I.
Although CVC was not visually apparent in 15 of 29 baseline recordings, during cardiac pacing, 9 of these 15 "noncoupled" animals (60%) subsequently showed clear evidence of CVC, as illustrated in Fig. 4, which shows results obtained from one such animal before and during cardiac pacing. Before pacing (Fig. 4A), no horizontal structures were observed on the RI plot, consistent with the absence of apparent CVC. In contrast, the same animal during cardiac pacing (Fig. 4B) demonstrated banding structures on the RI plot that are consistent with pattern I and IVhi coupling. Comparison of the two sets of results shows HR-to-fo ratio consistently >4:1 throughout the recording in Fig. 4A, with relatively large-magnitude variability in the CDI-I time series that did not vary throughout the recording. In contrast, the recording in Fig. 4B shows periods of pattern I coupling with an HR-to-fo ratio corresponding to 5:1, while the period of pattern IVhi coupling was associated with HR-to-fo ratio slightly >5:1. Although || during cardiac pacing was similar, there was significant variation in magnitude throughout the epoch, with significant step-wise reductions in CDI-I during pattern I coupling. || is greater in Fig. 4A than in Fig. 4B. SH is greater than SHT in Fig. 4A, which is consistent with no apparent structure in the RI plot or variation in the magnitude of CDI-I throughout the recording. In contrast, SH fell below SHT in Fig. 4B during pattern I coupling, which was associated with a step-wise reduction in CDI-I.
The changes in CVC pattern and CDI-I with HR alterations in an animal with clear evidence of CVC are illustrated in greater detail in Fig. 5. HR was increased from 319 to 360 beats/min over seven different 120- to 180-s HR steps. The RI plot shows sequential transitions between different CVC patterns, as well as periods without CVC. From the beginning of the recording, at 6 s, a 30-s period of pattern I coupling was observed. This was followed by a period of pattern IVlo coupling between 36 and 210 s. Between 210 and 360 s, three transitions from pattern I to pattern IVhi and back to pattern I were followed by a period of pattern I coupling between 385 and 420 s and pattern IVhi coupling between 420 and 600 s. The epoch appeared uncoupled from 600 s onward. A distinct relation between the pattern of CVC and the HR-to-fo ratio can be observed. Pattern I coupling was associated with an HR-to-fo ratio of 5, consistent with five heartbeats elapsing for every breath. Pattern IVlo coupling was associated with an HR-to-fo ratio slightly below a whole number, whereas pattern IVhi coupling was associated with an HR-to-fo ratio slightly greater than a whole number. In contrast, the HR-to-fo ratios during uncoupled segments were noninteger, with values that were generally a little greater or smaller than the HR-to-fo ratios associated with pattern IV coupling. Figure 5 shows that transitions between different CVC patterns were accompanied by sharp step-wise changes in fo and CDI-I, again with pattern I coupling associated with SH less than SHT and step-wise reduction in CDI-I. These observations are consistent with consecutive breaths during pattern I coupling having similar cycle durations that abruptly shift to a pattern of greater breath-to-breath respiratory period variability with pattern IV coupling and without coupling.
In six animals, visual inspection of the RI plot showed no clear horizontal structures, even with cardiac pacing. In these experiments, incremental changes in HR did not appear to influence CDI-I, with the degree of variability at integer HR-to-fo ratios appearing similar to variability magnitude observed at noninteger HR-to-fo ratio. Moreover, the sharp step-wise changes in fo and CDI-I clearly apparent in Fig. 5 were not observed in these experiments, which is consistent with very weak or truly absent CVC.
The relations between CVC pattern, HR-to-fo ratio, and breath-to-breath CDI-I highlighted by Figs. 3–5 were consistently observed throughout the study and are statistically summarized in Table 2. Table 2 shows that pattern I was associated with whole number integer HR-to-fo ratios with smaller || and ||, indicating minimal RI–1 interval and respiratory period variability compared with pattern IV and uncoupled data. In the present study, pattern III coupling was present in only two experiments for 30 s on each occasion (see Fig. 7). Even less prevalent was pattern II coupling, which was not observed in the nerve-intact state. Because of small numbers, patterns II and III were excluded from the statistical comparison in Table 2.
|
|
was less than SHT, as well as the relation between CVC pattern, HR-to-fo ratio, and CDI-I illustrated in Fig. 5. Examples of CVC and associated respiratory period variability after vagotomy are shown in Fig. 6.
The effect of SAD was examined in nine experiments. In this cohort, CVC was observed in all RI plots obtained before SAD with the heart beating spontaneously. This was verified by at least one episode of statistically significant CVC on the SH time series and features consistent with CVC on the HR-to-fo ratio and CDI-I time series. After SAD, RI plots from spontaneous epochs were again examined, and brief periods of CVC were detected in two animals. An example from one animal is shown in Fig. 7. The post-SAD BRS was 8 beats/min (phenylephrine) and 9 beats/min (sodium nitroprusside) for one rat and 8 beats/min (phenylephrine) and 8 beats/min (sodium nitroprusside) for the other rat showing CVC. In the seven rats that demonstrated no CVC after SAD, cardiac pacing was performed to ensure that the absence of observable CVC was not simply due to noninteger HR-to-fo ratios. Examination of the RI plot in these rats showed no RI plot structure consistent with CVC, even with approximately whole number integer HR-to-fo ratios. The SH in these recordings was consistently greater than SHT, and CDI-I time series showed no variation in relation to changes in the HR-to-fo ratio.
In seven experiments, we examined the effect of vagotomy + SAD. In this cohort, CVC was observed in all animals before denervation with the heart beating spontaneously. After denervation, none of the seven animals demonstrated CVC. Visual inspection of the RI plot for recordings obtained from these seven animals during cardiac pacing verified the absence of any RI plot structure consistent with CVC, even with approximately whole number integer HR-to-fo ratio. SH values for these animals were consistently greater than SHT values, and the CDI-I time series showed no variation in relation to changes in the HR-to-fo ratio (Fig. 8).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
The role of the vagus nerve in CVC has previously been examined, but results have been inconclusive. Kenner et al. (24) used an epinephrine infusion to achieve a whole number HR-to-fo ratio and observed periods of coupling with transient periods of "escape" that were associated with increased respiratory variability. The loss of these features after vagal denervation led Kenner et al. to conclude that vagal afferents were critical to CVC. However, Bucher et al. (2), using a variable pneumothorax and cooling of the sinoatrial node to achieve whole number HR-to-fo ratio, described CVC that persisted after bilateral vagotomy. This discrepancy may be due to failure in both studies to adequately consider the effects of changes in the HR-to-fo ratio and CDI-I associated with vagal denervation.
In this study, bilateral vagal denervation (with the CSN and ADN intact) was associated with no significant changes in CVC. Bilateral SAD (with the vagus nerves intact) was associated with a marked reduction in CVC, although in a small number of animals CVC persisted. In contrast, CVC was completely abolished in all animals after bilateral SAD and vagal denervation, even after we accounted for the HR-to-fo ratio by cardiac pacing. These observations are consistent with arterial baroreceptors playing a critical role in the CVC process with negligible contribution from vagal afferents.
A potential explanation for the persistence of CVC in two rats after SAD may relate to the completeness of SAD. In rats, the ADN generally runs separately from the vagus nerve but can also run closely alongside the vagus nerve (27, 39). Even though SAD was considered complete in all rats included in protocol II of the present study, this was based on an arbitrary definition of a <10 beat/min change in HR after phenylephrine and sodium nitroprusside challenge (27, 38, 39). Thus it remains possible that not all fibers were eliminated and that some residual ADN fibers traveling along the vagus sheath remained intact to produce slightly <10 beat/min reflex changes in HR.
Although all patterns of CVC were observed in this study, patterns II and III were relatively rare. In the nerve-intact state, pattern III was observed only briefly in two experiments, whereas pattern II was not observed. The rarity of these patterns may relate to the relatively small areas occupied by these two patterns on the HR-to-fi map described in our previous study (13). In addition, noisy dynamics inherent in the cardiorespiratory system may adversely affect these patterns and cause them to appear "uncoupled" where they may be present.
Two clear epochs of pattern II coupling were observed after vagotomy, which in this study was associated with a significant reduction in fo. An association between pattern II coupling and slow fo in anesthetized humans has been described (12). One explanation for this association is that pattern II is a variant of pattern I coupling in which small variations in the fi of the respiratory oscillator cause variations in the HR-to-fo ratio. According to our model, the impact of such small variations in fi is greatest at low frequencies and may explain why pattern II was observed only after vagal denervation.
Mechanism of CVC
In 1967, Hinderling (21) described the tendency of the HR-to-fo ratio to assume integer values during cardiac pacing in human subjects. Hinderling proposed that the tendency for cardiac and respiratory rhythm to couple was the result of a cardiac influence on respiration but did not further elaborate on the possible nature of this influence. The present study extends Hinderling's observations to the anesthetized rat and is the first to show that CVC during cardiac pacing is accompanied by sharp step-wise changes in fo and CDI-I that are related to the HR-to-fo ratio. For example, in the case of pattern I coupling, where we have proposed that each breath is triggered by the preceding heartbeat, the CDI-I is consistently smaller than in the other patterns of CVC or in the case of uncoupled epochs. Therefore, dynamic transitions from pattern IV or an uncoupled data segment to pattern I coupling would be accompanied by a step-wise reduction in CDI-I. This was clearly demonstrated in this study and is consistent with the hypothesis that CVC results from the entrainment, or resetting, of the respiratory rhythm by some cardiac cycle-related trigger.
This cardiac-trigger hypothesis adequately explains why CVC persists during atrial fibrillation (28), which is a pathological condition of the heart associated with irregular HR and diminished efferent vagal modulation of the sinoatrial node. Taken together, these observations clearly preclude any significant role of respiratory modulation of the heart in the CVC process.
Recent experiments recording neuronal activity in the nucleus tractus solitarius and in the rostral and caudal ventral respiratory groups detected and quantified beat-to-beat cardiac modulation of neurons within central respiratory networks (4, 5). It was concluded that a cardiac cycle-related modulation is associated with the arterial pulse, which is evident in respiratory premotor, as well as motor, output. It was also suggested that the respiratory pulse modulation is likely to be peripheral and sensory, rather than of central origin (4, 5), which is consistent with our observation that CVC was abolished after SAD. Indeed, it is well established that neural stimulation of the CSN and the SLN alters respiratory timing (7–9, 33, 36, 40, 41, 43).
Although the respiratory influence on cardiac timing (respiratory sinus arrhythmia) is well known, the ability of the heart to variably influence respiratory timing is not well appreciated or understood (4, 5, 17, 19, 20). Such an interaction may have potential clinical relevance in conditions of impaired respiratory rhythm control and may help explain observations such as the onset of tachycardia with exercise, dyspnea associated with tachycardia, including patients in fast atrial fibrillation, hyperventilation associated with anxiety, and reduced incidence of apnea episodes during atrial overdrive pacing in patients with central and obstructive sleep apnea (15–17, 19, 20).
Study Limitations
These findings should be interpreted in the context of the following limitations, many of which are common to all locking studies. One limitation relates to the plasticity of biological systems, where preparations may be affected, in unaccountable ways, by the stimulation itself (18). In addition, complex interactions between the various parameters and restrictions on the ability to measure these variables in biological systems mean that experiments were conducted with less control than in the ideal situation, where all interacting variables are constant. In the present study, it was elected to modify HR, inasmuch as HR is readily amenable to physiological manipulation. However, the fi of the respiratory oscillator, or the true strength of the cardiac stimulus, cannot be directly measured, and although these variables can be estimated, these procedures carry their own set of limitations (13). In the present study, it has been assumed that the fi and cardiac burst magnitude are relatively constant throughout a given experiment. Another limitation relates to the use of cardiac pacing as a means to alter HR. Because cardiac pacing cannot slow HR, the ability to explore regions of coupling associated with lower HR-to-fo ratios before vagotomy was limited. In addition, the present study was limited by how finely the HR and, therefore, the HR-to-fo ratio could be altered without excessive prolongation of the time needed to perform the study protocol. Finally, although we have demonstrated a dependency of respiratory period variability on cardiac activity, this study dealt entirely with peripheral pathways of CVC and does not provide further insights on central processes.
In summary, the present study shows that induced changes in HR resulted in changes in CVC pattern and CDI-I that are consistent with our model. This study has demonstrated the critical role of baroreceptor afferent nerves in CVC, and, therefore, it can be concluded that the putative cardiac afferent signals responsible for CVC are of baroreceptor origin. However, detail regarding the central processes of CVC requires further study.
|
FOOTNOTES |
|---|
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.
|
REFERENCES |
|---|
|
TOPABSTRACTMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE |
