Baroreceptor Reflex Control of Heart Rate:A Predictor of Sudden Cardiac Death

GEORGE E. BILLMAN, PH.D., PETER J. SCHWARTZ, M.D., AND H. LOWELL STONE, PH.D.

SUMMARY To explore the possibility that the analysis of autonomic reflexes could identify subgroups athigh risk of ventricular fibrillation, we studied chronically instrumented mongrel dogs randomly dividedinto two groups. Twelve dogs served as controls and 17 were studied 3-4 weeks after anterior wall myocar-dial infarction (MI). After recovery, the dogs were given bolus i.v. injections of phenylephrine, 10 .tg/kg,and nitroprusside, 100 ptg/kg, to raise or lower systolic arterial pressure 30-50 mm Hg. The RR intervalswere plotted against the systolic pressure during the preceding beats, and the slope (baroreflex slope) wasdetermined by least-squares-fit linear regression. On a subsequent day, the left circumflex coronary arterywas occluded for 2 minutes, beginning with the last minute of an exercise stress test and continuing for 1

minute after the cessation of exercise (MI group only). The dogs could be divided into two groups based ontheir response to this test; 11 dogs (65%) had ventricular fibrillation (susceptible), whereas six dogs (35%)did not (resistant). The baroreflex slope (control 20.49 ± 8.59; resistant 10.95 ± 4.68; susceptible 4.60 +

1.77 msec/mm Hg) and the heart rate response to a 30-mm Hg increase in arterial pressure (control - 56.5+ 14.8; resistant -40.0 ± 12.2; susceptible - 12.9 ± 5.0 beats/min) for the susceptible dogs were signifi-cantly different from those of the control and resistant dogs. This may indicate that the resistant dogs have agreater capability to activate strong vagal reflexes, which reduce vulnerability to ventricular fibrillation.We conclude that anterior wall MI significantly attenuates the baroreceptor reflex control of heart rate andthat analysis of the heart rate response to arterial pressure increases allows identification of subgroups ofdogs at higher risk for ventricular fibrillation. A prospective study in patients with MI is warranted.

EARLY IDENTIFICATION of patients with ischemicheart disease at high risk for sudden death is a prerequi-site for the use of aggressive preventive measures.Conversely, a more conservative approach may beused in low-risk patients. Important progress has beenmade in the risk stratification for both sudden andnonsudden cardiovascular deaths.' However, identi-fying patients at a high risk for ventricular fibrillationremains elusive. Some of these patients have beenidentified,4' but the underlying mechanisms, whetherpersistent electrical instability or major hemodynamicimpairment, have not been clarified.A large body of evidence indicates that the auto-

nomic nervous system plays a critical role in triggeringventricular fibrillation."'0 In particular, vagal activa-tion is generally viewed as protective, ' while in-creased sympathetic activity predisposes the heart toventricular fibrillation. 12 3 Conversely, both pharma-cologic'4 and surgical'5 16 antiadrenergic interventionsprotect against sudden death due to ventricularfibrillation.

Surprisingly, there has been no attempt to use auto-nomic reflex changes to discriminate among patients athigh or low risk for sudden death. Changes in heart ratemediated by baroreceptor reflex can provide a mean-ingful way to assess autonomic neural control of theheart. Takeshita et al.'7 found that myocardial ische-mia attenuates baroreflex control of heart rate in re-sponse to arterial pressure increases. Clinical studiesFrom the Department of Phtysiology and Biophysics, The University

of Oklahoma at Oklahoma City Health Sciences Center, OklahomaCity, Oklahoma, and the Uniba per lo Studio delle Aritmie, Centro diFisiologia Clinica e Ipertensione, Istituto di Clinica Medica IV, Univer-sitt di Milano, C.N.R, Milan, Italy.

Supported by NIH grants HL-22154 and HL-07430.Address for correspondence: George E. Billman, Ph.D., Department

of Physiology and Biophysics, The University of Oklahoma at Oklaho-ma City Health Sciences Center, P.O. Box 26901. Oklahoma City.Oklahoma 73190.

Received December 18, 1981; revision accepted March 12, 1982.Circulation 66, No. 4, 1982.

874

suggest'8 '9 that baroreflex function may be altereda,fter myocardial infarction (MI), as drug-induced hy-potension may not be associated with tachycardia.The purpose of this study was to determine whether

there is a relationship between autonomic reflex con-trol of heart rate and susceptibility to sudden death thatwould identify subgroups at higher risk.

MethodsThirty-four mongrel dogs that weighed 14.5-25 kg

Were chronically instrumented to measure aortic pres-~ure, left circumflex coronary blood flow, and theFICG. The dogs were randomly assigned to twogroups: 12 controls and 22 in which an MI wasproduced.

Surgical PreparationThe dogs were given thiopental sodium (Pentothal,

Abbott Laboratories), 25 mg/kg i.v., as a preanesthet-ic, and anesthesia was maintained by the inhalation ofa halothane, nitrous oxide and oxygen mixture. A leftthoracotomy was performed in the fourth intercostalspace. The left circumflex coronary artery was dissect-ed from the surrounding epicardial fat, and an 8-mHzcontinuous-wave Doppler flow transducer and a pneu-matic occluder were placed around this vessel. A hep-arinized, saline-filled catheter was placed into the aor-ta through the left common carotid artery. Insulatedsilver-coated copper wires were sutured to the epicar-dial surface of both the left and right ventricles andwere later used to record the ECG.

In 22 dogs, an experimental MI was produced. Amodified two-stage occlusion20 was performed on theleft anterior descending coronary artery approximatelyone-third of the distance from its origin. The vesselwas partially occluded for 20 minutes and then tied off.In addition, two to three branches from the left anteriordescending coronary artery were ligated proximal tothe occlusion. The leads to the cardiovascular instru-

by guest on July 4, 2018http://circ.ahajournals.org/

Dow

nloaded from

BAROREFLEXES AND SUDDEN CARDIAC DEATH/Billman et al.

mentation were tunnelled under the skin to exit on theback of the dog's neck. Pentazocine lactate (TalwinWinthrop laboratories), 30 mg i.m., was given ap-proximately every 8 hours for the first 24 hours tocontrol postoperative pain.

Prevention of Early Cardiac ArrhythmiasThe dogs subjected to MI were placed in an inten-

sive care setting. The therapy described below wasgiven to prevent early cardiac arrhythmias. The dogsreceived 100 mg of lidocaine HCl i. m. (Xylocaine,Astra Laboratories) before surgery, which was supple-mented (60 mg i.v.) during each stage of the Harristwo-stage occlusion. After surgery, the dogs wereplaced in a quiet recovery area, the ECG was moni-tored, and a lidocaine drip infusion (100 ,ug/ml inlactated Ringer's solution) was maintained at the rateof 1 ml/min for the first 24 hours after MI. The dogswere then given 500 mg of procainamide HCl i.m.(Pronestyl, E. R. Squibb & Sons Inc.) twice daily forthe next 3 days. These interventions seemed useful, asonly four dogs died during this period. A fifth dog diedfrom exsanguination through the catheter 2 weeks afterMI.

Baroreceptor Reflex TestingThe dogs were allowed to recover for 3-4 weeks

before baroreceptor reflex testing began. The dogswere placed on a laboratory table and a venous catheterwas percutaneously placed in the cephalic vein to ad-minister the vasoactive drugs. Arterial pressure wasobtained from the aortic catheter with a StathamP23Db pressure transducer. Baroreceptor reflex con-trol of heart rate was then assessed by the method ofSmyth et al.21 The dogs were given injections of sodi-um nitroprusside dihydrate, 100 gug/kg (Nipride,Roche Laboratories), and phenylephrine HCl, 10 gig/kg (Neo-Synephrine, Winthrop Laboratories) to raiseor lower systolic arterial pressure 30-50 mm Hg. EachRR interval (and heart rate) was plotted as a-function ofthe preceding systolic pressure. The analysis was per-formed beat by beat beginning only after the first no-ticeable change in RR interval. Control data were ob-tained by averaging over five successive beatsimmediately before the vasoactive drug was infused. Aleast-squares-fit linear regression was performed, andthe reflex control of heart rate was expressed as theslope of the linear regression line (an index of barore-ceptor reflex sensitivity). The slope was accepted forfurther analysis only if the correlation coefficient was0.80 or greater.

Sudden Death TestingSusceptibility to ventricular fibrillation was evaluat-

ed with a new experimental model in which malignantarrhythmias were consistently induced by a combina-tion of exercise, acute myocardial ischemia and auto-nomic reflexes.22 The 17 dogs that survived MI weresubjected to a submaximal exercise stress test. Thedogs ran on a motor-driven treadmill for 15 minutes,while work load increased every 3 minutes (4.8 kph,

during the last 3 minutes). During the last minute of theexercise test, the left circumflex coronary artery wasoccluded; the treadmill was suddenly stopped and theocclusion maintained for an additional minute. Theocclusion lasted 2 minutes. Large steel plates wereplaced across the dog's chest so that electrical defibri-lation could be performed with minimal delay. Arterialpressure, ECG and heart rate were monitored through-out the exercise test.

Data AnalysisAll data were recorded on a Beckman RM recorder.

The slope, as determined by linear regression, wastreated as a response variable and analyzed by one-wayanalysis of variance. Newman-Keul's multiple-rangetest23 was used to make comparisons. Specifically, thehypothesis that anterior wall myocardial infarction at-tentuated the baroreceptor reflex control was tested(null hypothesis- anterior wall myocardial infarctiondoes not alter the baroreceptor reflex control of heartrate). The hypothesis that sudden cardiac death couldbe predicted from the baroreflex slope was also tested(null hypothesis - the baroreflex slopes in the dogsthat had ventricular fibrillation are not different fromthose in dogs that did not). To avoid possibie bias, wecalculated the baroreflex slopes before the dogs weretested on the treadmill. Control heart rate and systolicarterial pressure were obtained by averaging over thelast five beats before the drug infusion. The changes inheart rate for a 30-mm Hg increase and a 30-mm Hgdecrease in systolic aortic pressure were also calculat-ed. The myocardial infarct size was determined by thenitroblue tetrazolium enzymatic staining technique24and reported as percent of the left ventricle involved.These data were compared as described above.

ResultsEffect of Myocardial InfarctionECG and arterial pressure recordings after phenyl-

ephrine injection for a control dog and a dog that sur-vived MI are shown in figure 1. For any given arterialpressure, the RR interval was more prolonged in thecontrol dog. A similar pattern was observed in all thedogs. For a 30-mm Hg increase in systolic aortic pres-sure, heart rate decreased by 56.5 + 14.8 beats/min,whereas after MI the decrease was only 26.0 + 8.6beats/min. Similarly, the baroreflex slope was sig-nificantly reduced after MI (controls 20.49 ± 8.59vs post-MI 6.84 + 4.31 msec/mm Hg). Resting heartrate and blood pressure were similar in both groups(table 1).

Susceptibility to Ventricular FibrillationThe dogs that survived MI could be divided into two

groups based on their response to the combined exer-cise and acute myocardial ischemia test. Eleven dogs(65%) developed ventricular fibrillation and were con-sidered susceptible to sudden death. Six dogs (35%)did not have life-threatening arrhythmias and wereconsidered resistant. In five of the six resistant dogs,the heart rate decreased during the exercise-ischemia

875

0% grade during first 3 minutes; 6.4 kph, 12% grade

by guest on July 4, 2018http://circ.ahajournals.org/

Dow

nloaded from

VOL 66, No 4. OCTOBER 198.2

Non-Infarcted Animal

2040 m sc 400 moc

I7 H

ECG 0 _

t50 mm Hg

(mm Hg) x:

Infarcted Animal

low m mm

1na mm Hg

400 m

H

FIGURE 1. ECG and systolic arterial bloodpressure (BP) recordings from a control clog

and a dog that survived infarction (suscepti-ble) in response to phenvlephrine. The RR in-terval (in bracket) wcas plotted and the systolicpressure marked with an asterisk. For thesame systolic presslure the control dog hatdmuch larger RR interval than the dog that sir-

vived infarction. A sinus arrhythmia at rest

was ojten seen in both the resistant (and sus-

ceptible dogs.

BPHg)(mm Hg)x _

test. This contrasts with the further increase in heartrate invariably observed at the time of the coronaryocclusion in the susceptible dogs.

The regression analyses for two MI dogs, one resis-

tant and one susceptible to ventricular fibrillation, are

presented in figure 2. The slopes for all 11 MI dogs are

displayed in figure 3. The baroreflex slope was signifi-cantly reduced in the susceptible dogs compared withthe resistant dogs (figure 4, table 1). Both the totalslope (response to increases and decreases in arterialpressure plotted as one function) and the response tophenylephrine (increased arterial presure) were signifi-cantly attenuated (F,26 = 20.04, p < 0.001 and F.,,

111.71, p < 0.001, respectively). In fact, there was

very little overlap between the groups: 90% of theresistant group had a slope greater than 10.72 msec/mm Hg, and 90% of the susceptible dogs had slopesless than 6.17 msec/mm Hg. However. the RR-inter-val response to an arterial pressure decrease was muchmore variable, and there was no significant differencebetween the groups (F.n, = 3.31, NS).The average heart rate reduction for a 30-mm Hg

increase in systolic arterial pressure was significantly

reduced in the susceptible dogs compared with both thecontrol and resistant dogs (F, = 41.89, p < 0.001)(fig. 5). The heart rate responses to arterial hypoten-sion, however, were not significantly different be-tween the groups (F - 0.40, NS). When heart raterather than RR interval was plotted aganst systolicarterial pressure, the results were similar (phenyleph-rine response: control 6.63 + 1.6 beats/mimmmHg; resistant -3.22 1.0 beats/min/mm Hg; sus-

ceptible -0.49 + 0.2 beats/min/mm Hg).The mean predrug heat rate and the systolic arterial

pressure were similar in all groups. The resistant andsusceptible dogs had MIs of similar size (table 1).

DiscussionThis study presents a very reliable means of identi-

fying subpopulations of dogs vulnerable to sudden car-diac death. Baroreceptor reflex control of heart rate, as

measured by the slope of the RR interval plottedagainst systolic arterial pressure, was significantly re-duced by anterior wall MI. The dogs that exhibited thegrreatest reduction in slope (smallest slope) invariablyhad ventricular fibrillation during the exercise-ische-

TABLE 1. The Effects ofMyocardial Infstrction otn Heart Rate, Ss stoliw Aortic Pressture. an2d Bar-orefiex- SlopeControl dogs Resistant dogs Susceptible dogs(n = 12) (n 6) (ti 11)

Slope (msec/mm Hg)

Total 20.49 88.59 10.95 + 4.68- 4.60 + 1.77`-;-:PhE only 26.08 15.37 17.91 ± 8.61 4.41 22.26'-1:NP only 11.17 + 7.77 7.25 5.13 4.53 2.74

HR (beats,min) 115.9 + 26.6 128.2 ± 15.0 109.4 + 23.4APsys (mm Hg) 122.6 17.7 125.2 18.8 125.4 21.4Infarct size (%4

of left ventricle) 23.2+ 5.5 21.7 t 7.4AHR (beats/min)

30 mm Hg APsys increase -56.5 + 14.8 -40.0 12.2; 12. 9 + 5 -30 mm Hg APsys decrease 87.4 + 26.7 76.4 + 47.2' 75.4 + 33.6

Values are mean + SD.--p < 0.05 vs control.tp < 0.01 vs control.<,pK 0.01 vs resistant.Abbreviations: PhE = phenylephrine; NP - nitroprusside; HR heart rate; APsvs systolic arterial pressure.

876 CIRCULATION

by guest on July 4, 2018http://circ.ahajournals.org/

Dow

nloaded from

BAROREFLEXES AND SUDDEN CARDIAC DEATH/Billman et al.

so0 .

40

0-y=10.12x-391.r =0.82

0

Ay=2.51x+

A r=0.91

CDSE

.67 E 30U0

E

0L 200-jin

A X

-I 10143.1 Lw

0

m n

O Control n= 12

E3 Resistant n: 6

M Susceptible n : 11

Total PhE only NP only

80 100 120 140 160

APsys (mmHg)

FIGURE 2. Regression analyses for tvo dogs that survivedinfarction, one resistant and one susceptible. There is a largeslope difference and the correlation coefficients are relativelyhigh. The response to phenylephrine and nitroprusside is ex-

pressed as one regression line (total slope). AP sys = systolicarterial pressure.

120O,

>

0

z

1000

800

600

400

80 120 160 200

APsys mmHg

FIGURE 3. Regression analysis for all dogs that survived in-

farction. The solid lines represent susceptible dogs, the dashedlines resistant dogs. The response to phenylephrine and nitro-

prusside is expressed as one regression line (total slope). AP sys= systolic arterial pressure.

FIGURE 4. Composite slopes for all dogs. Total RR intervalresponse to both an increase and a decrease in pressure. PhE= phenylephrine (10 pglkg) response; NP = nitroprusside(100 p.glkg) response. *p < 0.05 vs control dogs. **p < 0.01vs control dogs. + + p < 0.01 vs resistant dogs. Plotted as mean

+ SD.

mia testing. The heart rate reduction to a given arterialpressure increase was virtually absent in these dogs(30-mm Hg increase, 56.5 + 14.8 beats/min for con-

trol dogs vs 12.9 5.0 beats/min for the susceptibledogs).

Myocardial Infarction and BaroreflexThese results are consistent with the findings of both

experimental and clinical studies. 1719,25, 26 Takeshita etal.'7 found that acute myocardial ischemia attenuatedthe baroreflex-mediated reduction in heart rate re-sponse to arterial pressure increases. The baroreflexslope was reduced more than 50% during ischemia,and returned to control levels after the release of theocclusion. Toubes and Brody26 examined the peripher-al vascular component of the baroreflex and found thatafter experimental coronary embolization, hindlimbvascular resistance no longer increased in response tosystemic arterial hypotension.

Indirect clinical evidence suggests that baroreflexfunction is impaired by myocardial infarction. Severalstudies18 19 have shown that reflex tachycardia in re-sponse to drug-induced hypotension is often absent or

reduced in ischemic heart disease patients, as is reflexvasoconstriction.27 Bennett et al.25 found that the heartrate response to both the Valsalva maneuver and lowerbody negative pressure was attenuated in patients 1

year after MI, and baroreflex slope was demonstrablyreduced (control 11.5 msec/mm Hg, MI patients 2.5msec/mm Hg). Eckberg et al.28 used the pressor drugtechnique to evaluate baroreflex-mediated changes inheart rate and found that patients with the most ad-vanced heart disease had the smallest heart ratereductions.

* = RESISTANTA =SUSCEPTIBLE

1000 F

900 I

800 F

R-RINTERVAL(msec)

700 F

600 F

.500 [

400

300

877

by guest on July 4, 2018http://circ.ahajournals.org/

Dow

nloaded from

VOL 66, No 4, OCTOBER 1982

b/min+ 120 -

100 -

80 -

I-

u

a

60 -

40

20

O I 1

0

CZ,

U-

20

40

60

-80

B

**

++

Control Resistant Susceptible

FIGURE 5. Heart rate (HR) responses to 30-mm Hg increase(A) and decrease (B) systolic arterial pressure-increase. **p <0.01 vs control animals. + +p < 0.01 vs resistant animals.Plotted as mean + SD.

In contrast to these findings, an exaggerated heartrate response to carotid sinus massage has been notedin patients with coronary artery disease.29 These seem-

ingly disparate observations may result from the tech-nique used to activate the baroreceptors. Eckberg30reported that carotid massage, although simple to per-

form, is difficult to quantify. There is no standardreproducible method that can be applied during repeat-ed tests in the same patients or during tests in differentpatients, thereby making interpretation of these datadifficult.More than one mechanism may be involved with the

attenuation of baroreflex control of heart rate. Acutemyocardial ischemia excites vagal3' 32 and sympathet-ic33 cardiac afferent fibers. The resulting reflexes de-pend to some extent on the site of ischemia; anteriorischemia leads to predominant sympathetic reflexesand inferoposterior ischemia leads to predominant va-

gal reflexes, as suggested by both clinical34 and experi-mental35 studies. An inhibition of the sympathetic out-flow mediated by vagal afferent nerve fibers may

explain the observations of Toubes and Brody.26 Theresults of Takeshita et al. 17 can be best explained by theeffects of sympathetic afferent fiber excitation,33 whichleads simultaneously to an excitatory cardio-cardiacsympathetic reflex36 and to an inhibitory sympathova-gal reflex.37 Both reflexes would contribute to themaintenance of heart rate at a higher level than wouldbe produced by baroreceptor reflexes not associatedwith acute myocardial ischemia.

Although these reflex alterations in baroreflex func-

tion may be important during acute myocardial ische-mia, they probably contribute very little to the changesafter MI. The hemodynamic consequences of MI maycontribute more to this baroreflex slope reduction. MIdecreases the upstroke velocity of systolic arterialpressure,38' 39 which is an important determinant of ca-rotid sinus nerve activity40 and, hence, of the barore-ceptor reflex. Alterations in upstroke velocity maytherefore contribute significantly to the reduction ofbaroreflex slope after MI.Our results demonstrate that the capability of reflex-

ly decreasing heart rate in response to increases inarterial pressure is markedly reduced after MI.

Baroreflexes and Sudden DeathThe baroreflex slopes for the resistant and suscepti-

ble dogs were strikingly different, which immediatelysuggested a difference in the autonomic control of theheart in these two groups of dogs. Despite the novelapproach used in this study, our findings were notunexpected, as a strict relationship between barorecep-tor reflexes and vulnerability to ventricular fibrillationis logical on the basis of current knowledge. A steepbaroreflex slope in response to an increase in bloodpressure reflects primarily the activation of vagal effer-ent nerve fibers30 coupled, to some extent, with re-duced sympathetic efferent activity. Thus, the resistantdogs seem to have a greater capability to reflexly in-crease their vagal tone than the susceptible dogs.

Verrier, Lown and associates showed that enhancedvagal tone decreases susceptibility to ventricular fibril-lation. They also showed that vagal activation can op-pose the decrease in the ventricular fibrillation thresh-old induced by sympathetic stimulation.4" Moreover,they found that vagal stimulation antagonizes the in-creases in ventricular vulnerability associated with ei-

42~~~~~~~~~~1ther coronary occlusion" or psychological stress.43Five of six resistant dogs had a marked decrease inheart rate during the brief coronary occlusion duringexercise, which further demonstrates a propensity forpowerful vagal reflexes, as disclosed by the baroreflextesting, in these dogs.

Conversely, the almost flat baroreflex slopes in thesusceptible dogs may indicate a relatively greater sym-pathetic activation. Preliminary experiments indicatethat 3-adrenergic blockade substantially increases thebaroreflex slope in the susceptible dogs, but onlyslightly alters the slope in both the resistant and controldogs. The association between high sympathetic activ-ity and increased vulnerability to ventricular fibrilla-tion is well established.9

Thus, the analysis of the autonomic reflexes inducedby baroreceptor activation allows for the identificationof subgroups of dogs that are either more or less vul-nerable to ventricular fibrillation triggered by acutemyocardial ischemia.

Critique of the StudyThe pharmacologic method for evaluating barore-

flex function has been widely used since its descriptionby Smyth et al."' and has provided valuable insight into

878 CIRCULATION

by guest on July 4, 2018http://circ.ahajournals.org/

Dow

nloaded from

BAROREFLEXES AND SUDDEN CARDIAC DEATH/Billman et al.

circulatory reflex regulation. This technique is botheasy to perform and yields reliable quantitative data.However, as Eckberg30 pointed out, this approach hastwo inherent limitations.

First, it treats the RR interval arterial pressure curveas a linear rather than a sigmoidal function. Therefore,arterial pressure elevations that occur within either thethreshold or the saturation portions of the curve pro-duce smaller calculated slopes than do pressurechanges confined to the linear portion of the curve.Second, drugs that increase arterial pressure by vascu-lar smooth muscle excitation will also cause thesmooth muscle layers that contain the pressure recep-tor units to contract, thereby altering the baroreceptorfiring pattern or frequency.To circumvent these problems, we used only regres-

sion lines that displayed a large correlation coefficient(> 0.80) during an induced pressure change of 30-50mm Hg. Pressure changes of this magnitude have beenroutinely reported to extend well within the linear por-tion of the baroreceptor curve.40 In addition, similarslope reductions were noted when heart rates ratherthan RR interval were plotted against systolic pressure.The heart rate reductions in response to 30-mm Hgpressure increases were significantly different in thegroups.

Presumably, the effects of phenylephrine at the re-ceptor sites would be similar in both the control and MIdogs. Thus, drug-induced changes in receptor functionshould be independent of myocardial ischemic injuryand vulnerability to sudden death. This problemshould therefore not be a major concern in this study.Whenever susceptibility to sudden death is assessed

experimentally, a critical issue is the relevance to theclinical condition. The occurrence of ventricular fibril-lation in our model depends solely on the interactionbetween the following clinically relevant factors: acutemyocardial ischemia, exercise and its cessation and theinteraction between vagal and sympathetic reflexes.'Baroreflex testing can be used to predict the resultsof the interaction between vagal and sympatheticreflexes.

Clinical SignificanceExtrapolation of these data to the clinical setting

seems possible even when one considers species differ-ences.' The baroreflex slope and the heart rate reduc-tions associated with phenylephrine injections mayprove useful for identifying patients vulnerable toventricular fibrillation. Identifying MI patients at highrisk for sudden death by their cardiac autonomic re-sponse to changes in blood pressure merits carefulexamination.

AcknowledgmentThe authors express their sincerest thanks to Irene McMichael for

typing this manuscript. We also thank D. Thomas Dickey and GaryStout for their technical assistance.

References1. Moss AJ, Davis HT, DeCamilla J, Bayer LW: Ventricular ectopic

beats and their relation to sudden and nonsudden cardiac death after

myocardial infarction. Circulation 60: 998, 19792. Davis HT, DeCamila J, Bayer LW, Moss AJ: Survivorship pat-

terns in the posthospital phase of myocardial intarction. Circula-tion 60: 1252, 1979

3. Ruberman W, Weinblatt E, Goldberg JD, Frank CW, ChaudharyBS, Shapiro S: Ventricular premature complexes and sudden deathafter myocardial infarction. Circulation 64: 297, 1981

4. Schwartz PJ, Wolf S: Q-T interval prolongation as predictor ofsudden death in patients with myocardial infarction. Circulation57: 1074, 1978

5. Zappa P, Grazi S, Lotto A, Schwartz PJ: Excessive 5 year mortal-ity of anterior myocardial infarction complicated by ventricularfibrillation. (abstr) Circulation 56 (suppl III): 111-565, 1977

6. Lown B, Verrier RL: Neural activity and ventricular fibrillation. NEngl J Med 294: 1165, 1976

7. Schwartz PJ, Brown AM, Malliani A, Zanchetti A (eds): NeuralMechanisms in Cardiac Arrhythmias. New York, Raven Press,1978, p 442

8. Malliani A, Schwartz PJ, Zanchetti A: Neural mechanisms in life-threatening arrhythmias. Am Heart J 100: 705, 1980

9. Verrier RL, Lown B: Autonomic nervous system and malignantcardiac arrhythmias. In Brain, Behavior, and Bodily Disease, edit-ed by Weiner H, Hofer MA, Stunkard AJ. New York, RavenPress, 1981, p 273

10. Schwartz PJ, Stone HL: The role of the autonomic nervous systemin sudden coronary death. Ann NY Acad Sci 382: 162, 1982

11. Verrier RL: Neural factors and ventricular electrical instability. IJ7Sudden Death, edited by Wellens HJJ, Kulbertus HE. The Hague,M Nijhof. 1980, p 137

12. Schwartz PJ, Snebold NG, Brown AM: Effects of unilateral cardiacsympathetic denervation on the ventricular fibrillation threshold.Am J Cardiol 37: 1034, 1976

13. Schwartz PJ. Vanoli E: Cardiac arrhythmias elicited by interactionbetween acute myocardial ischemia and sympathetic hyperactivity:a new experimental model for the study of antiarrhythmic drugs. JCardiovasc Pharmacol 3: 1251, 1981

14. The Norwegian Multicenter Study Group: Tinmolol-induced reduc-tion in mortality and reinfarction in patients surviving acute myo-cardial infarction. N Engl J Med 304: 801, 1981

15. Schwartz PJ, Stone HL: Left stellectomy in the prevention of ven-tricular fibrillation caused by acute myocardial ischemia in con-scious dogs with anterior myocardial infarction. Circulation 62:1256, 1980

16. Schwartz PJ: The long QT syndrome. In Sudden Death, edited byWellens HJJ. Kulbertus HE. The Hague. M Nijhof, 1980, p 358

17. Takeshita A. Matsuguchi H. Nakamura M: Effect of coronaryocclusion on arterial baroreflex control of heart rate. CardiovascRes 14: 303. 1980

18. Chatterjee K, Parmley WW, Ganz W, Forester J. Walinsky P.Crexalls C. Swan HJC: Hemodynamic and metabolic responses tovasodilator therapy in acute myocardial infarction. Circulation 48:1183, 1973

19. Kelly DT, Delgado CE. Taylor DR, Pitt B. Ross RS: Use ofphentolamine in acute myocardial intarction associated with hyper-tension and left ventricular failure. Circulation 47: 729, 1973

20. Harris AS: Delayed development of ventricular ectopic rhythmsfollowing experimental coronary occlusion. Circulation 1: 1318,1950

21. Smyth HS, Sleight P. Pickering GW: Reflex regulation of arterialpressure during sleep in man. Circ Res 24: 109, 1969

22. Schwartz PJ. Billman GE, Stone HL: Autonomic interventions in anew animal model for sudden death. (abstr) Circulation 64 (supplIV): IV-289, 1981

23. Winer BJ: Statistical Principles in Experimental Design. 2nd ed.New York. McGraw-Hill, 1971

24. Nachlas MM, Shnitka TK: Macroscopic identification of earlymyocardial infarct by alterations in dehydrogenase activity. Am JPathol 42: 379, 1963

25. Bennett T. Wilcox RG.. Hampton JR: Cardiovascular reflexes inpatients after myocardial infarction: effect of long term treatment

with beta-adrenoceptor antagonists. Br Heart J 44: 265, 198026. Toubes DB. Brody MJ: Inhibition of reflex vasoconstriction after

experimental coronary embolization in the dog. Circ Res 26: 21 1,1970

879

by guest on July 4, 2018http://circ.ahajournals.org/

Dow

nloaded from

VOL 66, No 4, OCTOBER 1982

27. Kirby BJ: Circulatory reflexes in myocardial infarction. Br Heart J39: 168, 1977

28. Eckberg DL, Drabinsky M, Braunwald E: Defective cardiac para-sympathetic control in patients with heart disease. N Engl J Med285: 877, 1971

29. Brown KA, Maloney JD, Smith HC, Hartzler GO, Ilstrup DM:Carotid sinus reflex in patients undergoing coronary angiography:relationship of degree and location of coronary artery disease toresponse to carotid sinus massage. Circulation 62: 697, 1980

30. Eckberg DL: Parasympathetic cardiovascular control in human dis-ease: a critical review of methods and results. Am J Physiol 239:H581, 1980

31. Recordati G, Schwartz PJ, Pagani M, Malliani A, Brown AM:Activation of cardiac vagal receptors during myocardial ischemia.Experientia 27: 1423, 1971

32. Thoren PN: Activation of left ventricular receptors with non-medullated vagal afferent fibers during occlusion of a coronaryartery in the cat. Am J Cardiol 37: 1046, 1976

33. Malliani A, Recordati G, Schwartz PJ: Nervous activity of afferentcardiac sympathetic fibres with atrial and ventricular endings. JPhysiol (Lond) 229: 457, 1973

34. Webb SW, Adgey AA, Pantridge JF: Autonomic disturbance atonset of acute myocardial infarction. Br Med J 3: 89, 1972

35. Thames MD, Klopfenstein HS, Abboud FM, Mark AL, Walker JL:Preferential distribution of inhibitory cardiac receptors with vagalafferents to the inferoposterior wall of the left ventricle activated

during coronary occlusion in the dog. Circ Res 43: 512, 197836. Malliani A, Schwartz PJ, Zanchetti A: A sympthetic reflex elicited

by experimental coronary occlusion. Am JIPhysiol 217: 703, 196937. Schwartz PJ, Pagani M, Lombardi F, Malliani A, Brown AM: A

cardiocardiac sympatho-vagal reflex in the cat. Circ Res 32: 215.1973

38. Randall DC, Hasson DM, Brady JV: Conditional cardiovascularresponse to shock before and after coronary arterial occlusion. AmJ Physiol 236: H273, 1979

39. Bishop VS, Kaspar RL, Barnes GE, Kardon MB: Left ventricularfunction during acute regional myocardial ischemia in the con-scious dog. J Appl Physiol 37: 785, 1974

40. Kirchheim HR: Systemic arterial baroreceptor reflexes. PhysiolRev 56: 100, 1976

41. Kolman BS, Verrier RL, Lown B: The effct of vagus nerve stimula-tion upon vulnerability of the canine ventricle. Role of sympathet-ic-parasympathetic interactions. Circulation 52: 578, 1975

42. Lown B, Verrier RL: Neural factors and sudden death. In NeuralMechanisms in Cardiac Arrhythmias, edited by Schwartz PJ,Brown AM, Malliani A, Zanchetti A. New York, Raven Press,1978, p 87

43. Verrier RL, Lown B: Vagal tone and ventricular vulnerabilityduring psychologic stress. Circulation 62 (suppl III): III-176, 1980

44. Scher AM, Ohm WW, Bumgarner K, Boynton R, Young AC:Sympathetic and parasympathetic control of heart rate in the dog,baboon and man. Fed Proc 31: 1219. 1972

880 CIRCULATION

by guest on July 4, 2018http://circ.ahajournals.org/

Dow

nloaded from

G E Billman, P J Schwartz and H L StoneBaroreceptor reflex control of heart rate: a predictor of sudden cardiac death.

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 1982 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/01.CIR.66.4.874

1982;66:874-880Circulation. 

http://circ.ahajournals.org/content/66/4/874.citationthe World Wide Web at:

The online version of this article, along with updated information and services, is located on

  http://circ.ahajournals.org//subscriptions/

is online at: Circulation Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer information about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. FurtherEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculationpublished in Requests for permissions to reproduce figures, tables, or portions of articles originallyPermissions:

by guest on July 4, 2018http://circ.ahajournals.org/

Dow

nloaded from