Effects of Repetitive Bursts of Vagal Activity on Heart Rate

Transcription

1 Effects of Repetitive Bursts of Vagal Activity on Heart Rate By Matthew N. Levy, Thelma lano, and Harrison Zieske ABSTRACT The effect of the timing of discrete bursts of efferent vagal impulses on heart rate was determined in anesthetized dogs. Two modes of stimulation were employed. In the first mode, one stimulus burst was delivered per cardiac cycle. As the time from the beginning of the P wave to the vagal stimulus (P-St interval) was progressively increased, there was first a progressive lengthening of the cardiac cycle (P-P interval), then a rapid decrease in P-P interval, and finally a gradual augmentation of the P-P interval. The amplitudes of curves of P-P interval as a function of P-St interval increased as the number of stimuli per burst was augmented, with 10 stimuli/burst yielding nearly maximal effects. Vagal stimuli applied at P-St intervals which coincided with the negative-slope (d[p-p]/d[p-st]) region of such curves tended to evoke sinus arrhythmias. With 5 stimuli/burst or more, these arrhythmias were pronounced and consisted of alternate short and long P-P intervals. In the second mode of stimulation, bursts of stimuli were delivered to the vagus nerve at a frequency independent of heart rate. The cardiac pacemaker tended to become synchronized in some fixed ratio of vagal stimuli to P waves, and this tendency became greater the larger the number of stimuli per burst of impulses. Within any range of synchronization, a paradoxical effect was manifest increasing frequencies of vagal stimulation produced increasing rather than decreasing heart rates. KEY WORDS analog computer autonomic nervous system cardiac nerves electrocardiogram sinoatrial node cardiac synchronization paradoxical heart-rate response parasympathetic nervous system sinoatrial arrhythmia anesthetized dogs In recent studies of the neural regulation of heart rate, the importance of the timing of the arrival of vagal impulses at the sinoatrial (SA) node has been emphasized (1-5). In these studies, the stimuli which were applied experimentally to the vagi were generally either repetitive, uniformly spaced electrical pulses (1-3) or single bursts of one (4) or multiple (5) pulses. The natural impulse traffic in the cardiac efferent vagal fibers is nonuniform in timing, with a distinct tendency toward grouping of From the Department of Investigative Medicine, Mt. Sinai Hospital, Cleveland, Ohio This work was supported by U. S. Public Health Service Grant HE from the National Heart and Lung Institute. Dr. lano is a Fellow of the Samuel T. Haas Charitable Trust Fund. Received October 4, Accepted for publication December 16, impulses at certain times in the cardiac cycle (6, 7). Hence, the importance of obtaining information on the influence of repetitive bursts of vagal activity on the cardiac pacemaker is obvious. The present study was designed to acquire such data. Methods All experiments were conducted on mongrel dogs which were anesthetized with morphine sulfate, 2 mg/kg, i.m., followed 30 minutes later by chloralose, 75 mg/kg, dissolved in polyethylene glycol. A tracheal cannula was inserted through a midline cervical incision. Both cervical vagi were transected, and a bipolar shielded platinum electrode was applied to the cardiac end of the right nerve. The chest was opened through an incision in the fourth right intercostal space. A bipolar electrode eatheter was introduced into the right atrial cavity through a small incision in the tip of the right auricle. The right atrial electrogram and 186

2 HEART RATE CONTROL 187 mean arterial pressure were recorded on a Brush Mark 200 eight-channel oscillograph and on a Honeywell analog tape recorder, model LAR Propranolol, 1 mg/kg, i.v., was given, and the completeness of the blockade of cardiac adrenergic receptors was verified by observing the absence of a detectable change in heart rate during stimulation of the right stellate ganglion with 2-msec square pulses of 10-v amplitude at a frequency of 10 Hz for 30 seconds. The atrial electrogram served as an input to a parallel-logic analog computer (EAI 580). The computer was also used to generate square pulses (10 v, 1-msec duration) which were used to stimulate the right vagus nerve. By setting a counter, bursts of any desired number of such pulses could be delivered with a spacing of 2 msec between pulses. The frequency at which these bursts were delivered was controlled by the computer in one of two ways: (1) at a frequency of 1 burst/cardiac cycle, with a continuous change in the time from the beginning of the P wave to the first stimulus in the burst of impulses (P-St interval); (2) at a constantly changing frequency, independent of the heart rate. The P-P and P-St intervals of each beat and the stimulus burst frequencies were calculated by the analog computer and were recorded by the direct-writing oscillograph. The analog computer program was a modification of that described previously (3). The P and St pulses also triggered two of the clock-interrupt channels of a digital computer (PDP-12), and the P-P and P-St intervals were stored and subsequently printed on a beat-bybeat basis. STIM/BURST 1 P-P 1.5- to-e 0.5 P-St O- Results RAMP OF P-St INTERVALS Figure 1 shows the results of a representative experiment in which one burst of stimuli was delivered each cardiac cycle. In segment 1, each burst consisted of only one stimulus. Initially, the stimulus was delivered to the vagus nerve each beat at about the beginning of the P wave (P-St interval ss 0). After about 10 seconds, the stimulus was delivered later in each subsequent cardiac cycle by a fixed time increment, i.e., there was an ascending ramp of P-St intervals (Fig. 1, bottom). As the P-St interval was progressively increased, the vagal stimuli became more effective at first, as manifested by an increase in the P-P interval (Fig. 1, top). However, at a P-St interval of approximately 0.22 seconds, the vagal stimuli were maximally inhibitory; stimuli delivered each beat at greater P-St intervals produced progressively smaller P-P intervals. Stimuli delivered at a P-St interval of about 0.36 seconds were minimally inhibitory; stimuli delivered later than this in the cardiac cycle again produced a progressively greater prolongation of the cardiac cycle. For each number of stimuli per burst, the P-St interval was changed by a fixed increment each beat, and a ramp of increasing and then decreasing P-St intervals was obtained for each of the FIGURE 1 The changes in length of the cardiac cycle (P-P interval) produced by variations in the time from the beginning of the P wave to the vagal stimulus (P-St interval) when there were 1, 2, 3, 6, or 10 stimuli/burst of pulses delivered to the transected right vagus nerve. The P-St interval was altered by a fixed amount each beat, first in an ascending ramp and then in a descending ramp, but only one direction is shown for any given number of stimuli per burst. The results of ascending and descending ramps were not significantly different. The sequence of applying the different numbers of stimuli per burst was chosen randomly; in this experiment, the sequence was as follows: 1, 10, 3, 2, and 6 stimuli/burst. The downward blips of the time marker indicate 10-second intervals. 10

3 188 LEVY, IANO, ZIESKE P-P MSEC, P-St,M FIGURE The changes in P-P interval as functions of the P-St interval when there were 1, 2, 3, and 10 stimuli/burst of pulses delivered to the transected right vagus nerve. These curves are a smoothed retracing from an x-ij plot of the same experiment shown in Figure 1. The middle portion of the curve for 10 stimuli/burst was deleted, because it contained the arrhythmia evident in Figure 1 and would have obscured portions of other curves. The curve for 6 stimuli/burst was not included, because it was just below that for 10 stimuli/burst and almost coincident with it. The vertical broken line indicates the changes in P-P interval which would be expected when the number of stimuli per burst was changed while P-St was held constant at 345 msec; this corresponds to the experiment shown in Figure 4. The horizontal broken line indicates the changes in P-St interval which would be expected when the number of stimuli per burst was changed while the stimulus frequency was held constant at 1.43 bursts/sec (St-St = 700 msec). This corresponds to the experiment shown in Figure 9, and the discrepancies represent the changes in the preparation with time. See Figure 1 for abbreviations. levels of stimuli per burst employed in the experiment. In Figure 1, the ascending sequence is shown for 1, 2, and 3 stimuli/burst and the descending sequence for 6 and 10 stimuli/burst. Several features of these tracings change characteristically as the number of stimuli per burst is increased. These features, which are apparent in Figure 1, arc delineated more clearly when they are depicted (Fig. 2) as superimposed "pacemaker response curves," which are plots of P-P intervals as functions of the P-St interval (3). It is evident that the maximums, minimums, and amplitudes of these curves are all augmented by increases in the number of stimuli per burst. However, there is overlapping of ordinate values so that there are P-P intervals which are common to all curves within this range of numbers of stimuli per burst. It is also apparent that over the range of P-St intervals from zero to that which produces the maximum P-P interval the slope of the pacemaker response curve increases with the number of stimuli per burst. As the amplitudes of the curves become greater with an increase in the number of stimuli per burst, that portion of each curve with the negative slope (d[p-p]/d[p-st]) tends to become steeper. Also, tendencies for arrhythmias become more pronounced over the negative-slope region of the curves. In Figure 1, a slight tendency is seen with 2 stimuli/burst, the instability is somewhat more evident with 3 stimuli/burst, and it is pronounced with 6 stimuli/burst. At 10 stimuli/ burst, the entire negative slope region of the pacemaker response curve is replaced by this arrhythmia. If the P-St interval is held constant during the arrhythmia, it will persist indefinitely. A record of the atrial electrogram and the stimulation pattern made at fast paper speed (Fig. 3) reveals that the pattern of heart rate consists of alternating short and long cardiac cycles. By holding the value of the P-St interval constant, it is possible to show the effect on the P-P interval of changing the number of stimuli per burst. In the experiment illustrated in Figure 4, the P-St interval was 345 msec. When the number of stimuli per burst was st J h Jl L FIGURE 3 The atrial electrogram (P) and vagal stimuli (St) recorded at fast paper speed (25 mm/sec) during the arrhythmia at 10 stimuli/burst in Figure 1. The time marks indicate 1-second intervals.

4 HEART RATE CONTROL STIM/BURSTI 1 P-P P-St 1.5-1" FIGURE The changes in P-P interval produced by changes in the number of stimuli per burst when the P-St interval is fixed at 345 msec. The downward blips of the time marker signify 10-second intervals. This tracing was registered from the same animal used to obtain Figures 1 and 3. See Figure 1 for abbreviations. increased in steps, there was a progressive increase in the corresponding P-P interval. It is evident also that there were irregular fluctuations in the P-P interval with 2-4 stimuli/burst, but the P-P interval was steadier with 1 stimulus/burst and 5 stimuli/burst or more. Composite data on the amplitudes of the pacemaker response curves from 13 animals are presented in Figure 5. It is evident that the amplitude becomes appreciably greater with an increase in the number of stimuli per burst. The amplitude at 10 stimuli/burst is nearly maximal, since increasing the number of stimuli per burst to 20 or more resulted in less than a 10% additional increase in amplitude in several experiments. In almost all cases, the pronounced arrhythmia shown in Figures 1 and 3 occurred during the negativeslope phase of the pacemaker response curve when there were 5 stimuli/burst or more. RAMP OF STIMULUS FREQUENCY In eight animals, the right vagus nerve was stimulated at a constantly increasing burst frequency over a range from 0 to 2.5 bursts/ sec. This was usually repeated three or four times, each time with a different number of stimuli per burst. The sequence of employing the different numbers of stimuli per burst was selected randomly. The principal difference Circulation Research. Vol. XXX, February between this series of experiments and those described in the preceding section is that in these experiments there was no dependence of the timing of the vagal stimuli on the phase of the cardiac cycle, whereas in the preceding series one burst was delivered each cardiac cycle and the timing within each cycle was predetermined by the computer program. The results of a representative experiment in which 10 stimuli/burst were used are shown in Figure 6. It is obvious that the P-P interval was a very complex function of the burst frequency. There was an overall tendency for the prolongation of the P-P interval with increasing burst frequency. However, over several ranges of frequency, the P-P interval decreased progressively and substantially as the frequency of vagal stimulation was increased, a so-called paradoxical effect of vagal stimulation on heart rate (3). Similar responses were observed in two dogs by Dong andreitz (5). Examination of the atrial electrogram and the stimulus record at faster paper speeds (Fig. 7) reveals that the cardiac pacemaker had become synchronized with the efferent vagal activity at one of several ratios of STIMULI PER BURST FIGURE 5 The mean amplitudes (differences between maximums and minimums) ±SE of the pacemaker response curves recorded from 13 animals at 1, 3, 5, and 10 stimuli/ hurst, with one burst per cardiac cycle.

5 190 LEVY, IANO, ZIESKE St:p Ji:2j 1:1 2:1 J 31 P-St H STIM. FREQ. H_ BURST/ _ O FIGURE 6 T/ie e^ect of a progressive increase in the burst frequency (10 stimuli/burst) of vagal stimulation on the P-P and P-St intervals in a representative experiment. The downward blips of the time marker indicate 10-second internals. The numbers at the top of the record indicate the ratio of vagal stimuli to P waves (St:P); each ratio would be sustained (synchronization) if the frequency of vagal stimulation was maintained within the appropriate range. See Figure 1 for abbreviations. St:P 1:2 1:1 2:1 3:1 st 2- p-p - 1- o 2-i P-St H JULJLJULX n FIGURE 7 Atrial electrograms (P), vagal stimuli (St), P-P intervals, and P-St intervals during 1:2, 1:1, 2:1, and 3:1 synchronization (St:P) recorded at a paper speed of 25 mm/sec. These segments were obtained by playing back from the analog tape recorder appropriate sections of the experiment in Figure 6. The time marks indicate 1-second intervals. The St tracing is simply an event marker; it does not represent the true wave form of the vagal stimulus. See Figure 1 for abbreviations.

6 HEART RATE CONTROL ST/B" TP I 1 1 J 1 ix J 1:2 1:1 2:1 SYNCHRONIZATION RATIO FIGURE 8 J -5 ST/B" ^-10 ST/B TTie mean amplitudes ± SE of the ranges of synchronization at 1, 3, 5, and 10 stimuli/burst in eight dogs. 3:1 stimulus bursts to P waves (St:P). For example, over the frequency range of 0.76 to 1.33 bursts/sec, the P-P interval decreased from 1.32 to 0.75 seconds (Fig. 6), which corresponds to a range of atrial activation frequency of 0.76 to 1.33 P waves/sec. Hence, there was a 1:1 synchronization. As the stimulation frequency was increased to slightly above 1.33 bursts/sec, the P-P interval abruptly increased to 1.42 seconds. This is indicated at the top of Figure 6 by the arrow between St:P of 1:1 and 2:1. Continued increase in the stimulation frequency was accompanied by another progressive reduction of the P-P interval from this maximum value. Recordings at fast paper speed (Fig. 7) showed that there were now 2 stimulus bursts/cardiac cycle (2:1 synchronization) over the range of stimulus frequencies from 1.41 to 1.92 bursts/sec. Synchronization ratios of 1 to 2 and 3 to 1 are also evident in Figures 6 and 7. The tendency for the cardiac pacemaker to become synchronized with vagal activity is greater the larger the number of stimuli per burst. For example, in the experiment from which Figures 6 and 7 were recorded, 1:1 synchronization prevailed with 10 stimuli/ burst over a range of P-P intervals from 0.75 to 1.32 seconds, i.e., a range of 570 msec. In that same experiment, 1:1 synchronization occurred over ranges of 475, 375, and 250 msec for 5, 3, and 1 stimuli/burst, respectively. Similar relations occurred for other synchronization ratios as well. Hence, with larger numbers of stimuli per burst, the ranges over which synchronization occurred were larger, and there was a greater tendency to pass directly from one synchronization ratio to another as the burst frequency was progressively increased. Conversely, with lower numbers of stimuli per burst, synchronization tended to occur at fewer ratios and over narrower ranges, and there were greater intervening ranges of burst frequencies over which synchronization did not exist. Figure 8 presents the composite data for the eight dogs in which ramps of increasing burst frequency were employed. Data are shown for synchronization ratios of 1 to 2, 1 to 1, 2 to 1, and 3 to 1. Other ratios (e.g., 1 to 4, 1 to 3, 3 to 2) were also frequently seen but not consistently enough to be included in the analysis of the composite data. It is evident from the figure that for any synchronization ratio, the amplitude of the range of P-P intervals over which synchronization occurs increases with the number of stimuli per burst.

7 192 LEVY, IANO, ZIESKE STIM/BURST I1I2I3UI5I7I 10 P-P P-St 1.0-ZI FIGURE 9 The effect of changes in the number of vagal stimuli per burst on P-P and P-St intervals when the right vagus nerve was stimulated at a constant burst frequency of 1.43 burst/sec (St-St interval =700 msec). The downward blips of the time marker indicate 10- second intervals. This record is from the same animal used in Figures 1 4. See Figure 1 for abbreviations. In a few experiments, trains of 20 stimuli/ burst were also employed, and the amplitudes of these ranges were never more than 10-15% greater than those at 10 stimuli/burst. It is also apparent from Figure 8 that for a given number of stimuli per burst the range amplitudes were approximately the same for 1:2 and 1:1 synchronization but definitely less for 2:1 and 3:1 synchronization. In several of these experiments during the application of a ramp of increasing stimulation frequencies, the rate of change of the ramp was suddenly decreased to zero while 1:1 synchronization prevailed, i.e., a given stimulation frequency was maintained. At this constant burst frequency, the number of stimuli per burst was then changed in steps. The results of such an experiment are shown in Figure 9. At the left, 1:1 synchronization prevailed with a P-P interval (and also an St-St interval) of 700 msec and a P-St interval of msec. After 35 seconds, the stimulation mode was suddenly changed to 2 stimuli/burst. After a transient change, the P- P interval returned to the value of 700 msec, which corresponded to the period between the bursts of vagal stimuli, i.e., synchronization persisted. However, the new steady-state level of the P-St interval was now 680 msec. Similar results were obtained when the number of stimuli was periodically increased to 3, 4, 5, and 7 stimuli /burst. Except for transient changes, the P-P interval remained fixed at 700 msec, but there was a progressive, stepwise shift in the P-St interval. When 10 stimuli/burst were applied, synchronization was momentarily lost twice, but then it was reinstituted. However, the situation was obviously unstable, as shown by the small beatby-beat fluctuations in P-P and P-St intervals. Discussion It is well established that efferent cardiac vagal activity does not occur in either a uniform or a randomly dispersed temporal pattern; there is a definite tendency toward grouping of impulses in each cardiac cycle (6, 7). Such periodic, grouped impulses are probably initiated by stimulation of the arterial baroreceptors by the arterial pulse wave. Iriuchijima and Kumada (8) have shown that a single stimulus applied to the carotid sinus nerves appears subsequently as a discrete burst of impulses in efferent vagal fibers. Since increased mean arterial pressure and increased pulse pressure, within limits, augment the afferent discharges from the baroreceptors (9), it is likely that such stimuli would also increase the number of grouped, pulse-synchronous impulses in the efferent vagal fibers. The present series of experiments shows that an increase in the number of periodic efferent vagal discharges accentuates the tendency of the cardiac pacemaker to become synchronized with the rhythmic bursts of vagal activity, but it also may accentuate the tendency toward SA node instability. Such effects can be appreciated by considering the influence of changes in the number of vagal stimuli per burst on the configuration of the pacemaker response curve. This curve has been defined previously (3) as a plot of the P- P interval as a function of the P-St interval. In Figures 1 and 2, it is apparent that the amplitude of the pacemaker response curve becomes greater with an increase in the number of stimuli per burst. The amplitude of the pacemaker response curve defines the Circulation Research. Vol. XXX, February 1972

8 HEART RATE CONTROL 193 range over which 1:1 synchronization will take place for a given number of stimuli per burst (3). Synchronization of the cardiac pacemaker with efferent vagal activity tends to occur when the vagal burst arrives during that portion of the cardiac cycle in which the slope (d [P-P]/d [P-St]) of the pacemaker response curve is positive, because negative feedback prevails under such conditions (3, 10). The overlapping of ordinate values of the family of pacemaker response curves (Fig. 2) accounts for the tendency for synchronization to persist at a fixed heart rate when the number of stimuli per burst is altered. As shown in Figure 9, the P-P interval remained constant, except for transient changes, when the number of stimuli was varied from 1 to 7 stimuli/burst. The tendency persisted even at 10 stimuli/burst, although the synchronization was much less stable under these conditions. Changes in the number of stimuli per burst at a fixed burst frequency were accompanied by appreciable shifts in the P-St interval, as shown in Figure 9. The changes in P-St interval produced by changing the number of stimuli per burst at a constant burst frequency are represented by the horizontal broken line in Figure 2. On the other hand, when the P-St interval is fixed, an increase in the number of stimuli per burst produces a decidedly greater negative chronotropic effect, as shown in Figure 4 (which was recorded from the same animal). The tracing in Figure 4 reveals that the P-P interval increased by about 450 msec when the P-St interval was fixed and the number of stimuli was increased from 1 to 10 stimuli/burst. This response is represented by the vertical broken line in Figure 2. The negative-slope region of the pacemaker response curve defines a region in which positive feedback exists, as explained previously (3, 10). One manifestation of such positive feedback is the unsteadiness of the P-P intervals in Figure 4 with 2-4 stimuli/burst. Throughout the entire series of stimulations shown in this figure, the P-St interval was fixed at 345 msec. This P-St interval coincided in time with the negative-slope regions of the curves for 2 and 3 stimuli/burst, as illustrated by the vertical broken line in Figure 2. The negative-slope region of the pacemaker response curve also accounts for the SA node arrhythmia which is evident in Figures 1 and 3. Additional insight may be gained concerning the mechanism of this type of arrhythmia if the data are analyzed in an additional way to yield a curve which will be called the "vagal time course." The effects of any given burst of vagal stimuli obviously occur after the burst is delivered. It therefore seems appropriate to consider the resulting P-P interval as a function of the time which elapsed between the application of the stimulus and the P wave which terminated the cardiac cycle, since this is the time over which the released acetylcholine could have had an effect on the pacemaker. The lower curve in Figure 10 is the vagal time course for 1 stimulus/burst derived from the data in segment 1 of Figure 1. It can be seen that the minimal effect occurred in cycles which terminated at about 250 msec after the stimulus and that the maximal effect occurred in cycles which ended approximately 500 msec after the stimulus. p-p A ' ' \ 6STIM/BURST 1STIM /BURST TIME AFTER STIMULATION.msec FIGURE 10 The P-P interval as a function of the time from the beginning of the vagal stimulus to the beginning of the P wave for the experiment depicted in Figure 1. In the upper curve, the broken line represents the discontinuity resulting from the arrhythmia which is evident in Figure 1 (6 stimuli/burst). The dots represent the values of the P-P intervals during that arrhythmia. See Figure 1 for abbreviations.

9 194 LEVY, IANO, ZIESKE This function is analogous to the time course of the effect of a single vagal stimulus on heart rate, as described originally by Brown and Eccles (11). These authors derived the time course of the effect on heart rate of a single volley of vagal impulses by plotting the P-P interval of several beats following a vagal stimulus against the time which elapsed between the application of the stimulus and the P wave terminating the cycle. With a single stimulus, cardiac cycles which ended sooner than about 200 msec after stimulation were not prolonged at all (4, 11). This latency in the response to a stimulus is due at least in part to the delay involved in vagal impulse transmission and in acetylcholine release and diffusion. As the time between a stimulus and the following P wave increased beyond about 200 msec, the P-P interval was progressively more prolonged until a maximal effect was reached in cycles terminating approximately 450 msec after stimulation (4). Cycles terminating at later times were less prolonged, but the effect was usually detectable for more than ten beats after the single stimulus. This knowledge of the effect of a single stimulus can be applied to the interpretation of the response to repetitive stimulation (Fig. 10, 1 stimulus/burst). A cycle which terminates less than approximately 200 msec after the application of a given stimulus is not affected by that stimulus, and the P-P interval is determined by the amount of acetylcholine which still lingers from stimuli given during previous cycles. However, after this latency, the P-P interval begins to be affected by the stimulus given during that cycle, as well as by those given during previous cycles. The effect of previous stimuli continues to fall off with time, but the effect of the immediate stimulus is manifest only after the latency. The net result is that the slope of the curve becomes less negative after 200 msec, and the P-P interval passes through a minimum at about 250 msec. It then becomes progressively more prolonged until a maximum is reached at approximately 500 msec after the stimulus. Cycles terminating at times greater than this are less prolonged, as the effect begins again to fall off with time. The upper curve in Figure 10 shows the vagal time course for 6 stimuli/burst. As can be seen in Figure 1, a pronounced arrhythmia occurs over a certain range of P-St intervals with 6 stimuli/burst so that no continuous relationship appears to exist between the P-P and P-St intervals over the negative-slope region of the pacemaker response curve. However, when the data are plotted as a vagal time course (Fig. 10), the functional relationship for all beats becomes apparent. In Figure 10, the stable beats are plotted as a continuous curve, and each unstable beat is plotted as a single point. The unstable beats apparently define the region of the curve over which the vagal effect is increasing with time from stimulation (indicated in the figure by the broken line). The instability is related to the steepness of the vagal time course. Examination of Figure 3 leads to a further understanding of the instability. The P waves occur in pairs of alternating short and long beats, although each stimulus occurs at the same fixed interval after the preceding P wave. The first stimulus of each pair occurs just shortly enough before atrial depolarization that it has little or no effect on the subsequent P wave; that is, the SA node discharges during the latent period for that vagal stimulus. However, the acetylcholine released lingers long enough to retard the generation of the following P wave, so that the next vagal stimulus arrives at the SA node at a time when it too can be effective on that beat. Hence, the effects of both stimuli combine to cause a marked prolongation of the P-P interval. When the next P-wave generation begins, sufficient time has elapsed for most of the acetylcholine liberated by the preceding vagal barrages to be dissipated. Hence, the next stimulus occurs too late to be effective on that cycle, and another short P-P interval ensues. These alternating short and long beats appear as points at the bottom and top, respectively, of the rising portion (broken line) of the vagal time course in Figure 10. When the slopes of the rising portion of the

10 HEART RATE CONTROL 195 vagal time course curves are not large, as with 1 stimulus/burst, a relatively small delay in the onset of a given P wave has little influence on the effect of the next vagal stimulus, and the arrhythmias are negligible. References 1. SUGA, H., AND OSHIMA, M.: Modulation-characteristics of heart rate by vagal stimulation. Japan J Med Elect 6:465-^71, REID, J.V.O.: Cardiac pacemaker: Effects of regularly spaced nervous input. Am Heart J 78:58-64, LEVY, M.N., MARTIN, P.J., IANO, T., AND ZIESKE, H.: Paradoxical effect of vagus nerve stimulation on heart rate in dogs. Circ Res 25: , LEVY, M.N., MARTIN, P.J., IANO, T., AND ZIESKE, H.: Effects of single vagal stimuli on heart rate and atrioventricular conduction. Am J Physiol 218: , DONG, E., JR., AND REITZ, B.A.: Effect of timing of vagal stimulation on heart rate in the dog. Circ Res 27: , JEWETT, D.L.: Activity of single efferent fibres in the cervical vagus nerve of the dog, with special reference to possible cardioinhibitory fibres. J Physiol (Lond) 175: , KATONA, P.G., POITRAS, J.W., BARNETT, CO., AND TERRY, B.S.: Cardiac vagal efferent activity and heart period in the carotid sinus reflex. Am J Physiol 218: , IRIUCHIJIMA, J., AND KUMADA, M.: Efferent cardiac vagal discharge in response to electrical stimulation of sensory nerves. Jap J Physiol 13: , EAD, H.W., GREEN, J.H., AND NEIL, E.: Comparison of the effects of pulsatile and nonpulsatile blood flow through the carotid sinus on the reflexogenic activity of the sinus baroreceptors in the cat. J Physiol (Lond) 118: , GLAZE, H.J., JR., AND DONG, E., JR.: Mathematical models for heart rate responses to vagal nerve stimulation. Report PH , National Heart and Lung Institute, BROWN, G.L., AND ECCLES, J.C.: Action of a single vagal volley on the rhythm of the heart beat. J Physiol (Lond) 82: , 1934.