1578

Inducibility of Atrial Fibrillation During AtrioventricularPacing with Varying Intervals: Role of Atrial

Electrophysiology and the Autonomic Nervous System

YI-JEN CHEN, M.D., CHING-TAJ TAI, M.D., CHUEN-WANG CHJOU, M.D.,ZU-CHIN WEN, M.D., PAUL CHAN, M.D., SHIH-HUANG LEE, M.D.,

and SHIH-ANN CHEN, M.D.

From the Division of Cardiology, Department of Medicine, National Yang-Ming University, School of Medicine, andVeterans General Hospital-Taipei and Kaohsiung, Taipei, Taiwan

Pacing and Atrial Electrophysiology. Introduction: Patients receiving VVI pacemakershave a higher incidence of paroxysmal atrial fibrillation (AF) than those receiving DDDpacemakers. However, the mechanism behind the difference is not clear. The purpose of thisstudy was to investigate whether atrial electrophysiology and the autonomic nervous system playa role in the occurrence of AF during AV pacing.

Methods and Results: The study population consisted of 28 patients who had (group I, n = 15)or did not have (group II, n = 13) AF induced by a single extrastimulus during pacing withdifferent AV intervals. Atrial pressure, atrial size, atrial effective refractory periods, and atrialdispersion were evaluated during pacing with different AV intervals. Twenty-four-hour heartrate variability and baroreflex sensitivity also were examined. Atrial pressure, atrial size,effective refractory periods in the right posterolateral atrium and distal coronary sinus, andatdal dispersion increased as the AV interval shortened from 160 to 0 msec. During AV pacing,group I patients had greater minimal (52 ± 17 vs 25 ± 7 msec; P < 0.005) and maximal (76 ±16 vs 36 ± 9 msec; P < 0.005) atrial dispersion than group II patients. The differences in atrialsize and atrial dispersion among different AV intervals were greater in patients with AF thanin those without AF. Baroreflex sensitivity (6.6 ± 1.7 vs 3.9 ± 1.0; P < 0.00005), but not heartrate variability, was higher in patients with AF than in those without AF.

Conclusion: Abnormal atrial electrophysiology and higher vagal reflex activity can playimportant roles in the genesis of AF in patients receiving pacemakers. (J Cardiovasc Electro-physiol, Vol. 10, pp. 1578-1585, December 1999}

atrial fibrillation, autonomic nervous system, atrioventricular pacing

Introduction

Atrial fibrillation (AF) is more likely todevelop in patients treated with VVI pacing

This study was supported in part by Grants NSC-87-2314-B-010-058,87-2314-B-010-059, 87-2314-B-075-070, and 87-2314-B-075-071from the National Science Council, and Grants VGHYM-S4-30,VGHYM-S4-31, and VGH-30 from Tzou's Foundation, Taipei, Tai-wan. YJ. Chen and P. Chan are from Cardiovascular Medicine, TaipeiMedical College and affiliated Wan-Fang Hospital.

Address for correspondence: Shih-Ann Chen, M.D., Division ofCardiology, Veterans General Hospital-Taipei, 201 Sec 2,Shih-Pai Road, Taipei, Taiwan. Fax: 886-2-28735656; E-mail:sachen@vghtpe.gov.tw

Manuscript received 27 April 1999; Accepted for publication 21July 1999.

than in patietits treated with DDD pacing,'"*but the mechanism behind this difference is notclear. Several studies showed that DDD pace-makers have distinct hemodynamic advantagesover VVI pacemakers due to synchronous AVpacing.^"^ Atrial enlargement and stretchingdue to atrial contraction against closed AVvalves or regurgitation during ventricular con-traction that occurs when AV synchrony is lostmay serve as a substrate for AF. Moreover, itis known that atrial effective refractory period(AERP) and dispersion of atrial refractorinesswere altered as atrial pressure and atrial sizeincreased.'''0 Therefore, it is possible thatmechanoelectrical feedback contributes to the

Chen, et al. Pacing and Atrial Electrophysiology 1579

higher incidence of AF in patients receivingVVI pacemakers.

In addition, the autonomic nervous systemmay play a role in the genesis of AF in patientsreceiving VVI pacemakers. "-'^ Although previ-ous studies showed that patients receiving VVIpacemakers have higher sympathetic nerve activ-ity than those receiving DDD pacemakers,""'^the nature of the autonomic nervous system inpatients with or patients without AF is not clear.The purposes of this study were (1) to evaluatethe effects of pacing with different AV intervalson atrial pressure, atrial size, and atrial electro-physiologic characteristics; (2) to evaluate thedifferences between patients with and withoutinduced AF; and (3) to investigate the role of theautonomic nervous system in these patients.

Methods

Patient Charaeteristics

This study included 28 patients who under-went a clinically indicated electrophysiologicstudy. The patients were divided into two groups.In group I, there were 15 patients (8 men and 7women; mean age 58 ± 15 years) who hadinduced paroxysmal AF during the pacing study.Nine patients had symptomatic palpitations andsix had AV nodal reentrant tachycardia. In groupII, there were 13 patients (5 men and 8 women;mean age 61 ± 12 years) who did not have AFduring the pacing study. Ten patients had symp-tomatic palpitations and three had AV nodal re-entrant tachycardia. Four patients in group I andthree patients in group II had hypertension. Twopatients in group I had coronary artery disease.None of these patients had a history of AF.

Baseline Eleetrophysiologic Study

As described previously,"*'^ each patient un-derwent a baseline electrophysiologic studywhile in the fasting, unsedated state for at leastfive half-lives after discontinuation of antiar-rhythmic drugs. Informed consent of the baselineelectrophysiologic study was obtained from eachpatient. Four multipolar, closely spaced electrodecatheters (interelectrode space 2 mm; BostonScientific, Watertown, MA, USA) were intro-duced from the right and left femoral veins andplaced in the right atrial appendage, low rightposterolateral atrium, His-bundle area, and right

ventricle for programmed electrical stimulationand recording. One steerable decapolar catheter(2-5-2-5-2-5-2-5-2 configuration; Daig Corp.,Minnetonka, MN, USA) was introduced from theright internal jugular vein and placed in the cor-onary sinus. Intracardiac electrograms were dis-played simultaneously with ECG leads I, II, andVj on a multichannel oscilloscopic recorder(MIDAS system series 2000, PPG Instruments,Lenexa, KS, USA) and recorded on paper at aspeed of 100 to 150 mm/sec.

Study Protoeol

All patients underwent pacing (basic cyclelength 500 msec) with different AV intervals (0,60, 120, and 160 msec) in a random order. Aftercontinuous pacing during each AV interval for 5minutes, hemodynamic parameters (mean bloodpressure, right atrial pressure, and pulmonarycapillary wedge pressure), biatrial size, andAERPs were measured.

Measurement of hemodynamic parameters

Right atrial pressure and pulmonary capillarywedge pressure were measured using a fluid-filled luminal catheter. The patient's mean bloodpressure was measured using intra-arterial cath-eters from the femoral artery.

Measurement of AERP

AERPs in the right atrial appendage, low rightposterolateral atrium, and distal coronary sinuswere measured in randomized order during dif-ferent AV intervals. Measurement of AERPs wasobtained from an atrial extrastimulus (2-msecpulse duration, twice threshold) introduced fromearly diastole after every eighth atrial beat untilatrial capture occurred twice in succession (in2-msec increments). Measured AERPs were ac-ceptable for analysis only if these data wereidentical during two repeated measurements.AERP dispersion was defined as the maximaldifference of AERPs at the three stimulationsites. Atrial vulnerability was defined as the in-ducibility of AF sustained for > 30 seconds froma single atrial extrastimulus during measurementof AERP (as described earlier). Patients with AFsustained for > 10 minutes received electricalcardioversion to restore sinus rhythm. If AF wasinduced during the study protocol, repeated test-ing was performed 15 minutes after AF was

1580 Journal of Cardiovascular Electrophysiology Vol. 10, No. 12, December 1999

Spontaneously or electrically converted to sinusrhythm. To confirm a stable catheter, the atrialpacing threshold was remeasured before everymeasurement of AERP and after each electricalcardioversion.

Measurement of atrial size

Left and right atrial sizes were measured atend systole by two-dimensional echocardiogra-phy (SONOS 1000 ultrasound system; Hewlett-Packard, Andover, MA, USA) in the apical four-chamber view.'^

Baroreflex sensitivity study

Baroreflex sensitivity (BRS) was determinedusing intravenous phenylephrine before the pac-ing study only when patients had stable heart rateand blood pressure. Patients were placed in thesupine position and instructed to breathe regu-larly. Continuous arterial pressure and surfaceECG were recorded simultaneously at a paperspeed of 25 mm/sec. Blood pressure was in-creased by 15 to 30 mmHg by intravenous injec-tion of phenylephrine (2 to 4 /xg/kg bodyweight). The dose determined by this procedurewas repeated at least twice. Beat-by-beat changesin baseline systolic blood pressure (in mmHg)and in the lengths of the normal RR intervals (inmsec) were calculated from records. Each RRinterval was plotted against the preceding arterialpressure pulse, and linear regression analysis wasperformed. Only regression lines exhibiting asignificant correlation coefficient (P < 0.05)were accepted for analysis. The final slope rep-resents the mean value of three or more succes-sive determinations.

Heart rate variability study

The 24-hour ECG recordings were per-formed on the day before electrophysiologicstudy. Tape-recorded ECGs for each subjectwere digitally processed and annotated bymanual editing with a Holter Analysis System(Oxford Medilog Excel, Oxford Medical In-struments, Abington, Oxon, UK). After QRSconfiguration classification, the longest and theshortest RR intervals on the RR interval histo-gram, and the largest and the smallest RRratios on the RR ratio histogram, were manu-ally confirmed until no QRS complex was mis-labeled as either an artifact or an ectopic beat.

Time-domain variables considered in thisstudy were mean RR interval (the mean of allcoupling intervals between normal sinus beatsexpressed in msec), SDNN (the standard devi-ation of the mean RR interval expressed inmsec), rMSSD (the root mean square of dif-ferences of successive RR intervals), andPNN50 (the percentage of adjacent RR in-tervals that differed by > 50 msec). In ad-dition, three frequency-domain measurements,namely, low-frequency component (LF; 0.04to 0.15 Hz), high-frequency component (HF;0.15 to 0.40 Hz), and LF/HF ratio were com-puted.

Statistical Analysis

All data are expressed as mean ± SD. Com-parisons of parametric data measured during dif-ferent AV interval were analyzed by repeatedanalysis of variance. Differences of parametricdata between the two groups were analyzed byunpaired r-test. Chi-square test with Yates cor-rection or Fisher's exact test was used for cate-gorical data. P < 0.05 was considered statisti-cally significant.

Results

Effects of Different AV Intervals

Table 1 lists the effects of different AVintervals on the measured parameters in pa-tients who had induced AF (group I) duringthis study. As the AV intervals were decreasedfrom 160 to 0 msec, there was lower meanblood pressure, higher mean right atrial pres-sure, and higher pulmonary capillary wedgepressure. Atrial size also increased as the AVinterval decreased to 0 msec. AERP in theright atrial appendage was not changed duringdifferent AV intervals. However, as the AVintervals decreased from 120 to 0 msec, AERPin the right posterolateral atrium was increasedfrom 236 ± 20 msec to 258 ± 21 msec, andAERP in the distal coronary sinus was in-creased from 225 ± 25 msec to 246 ± 25msec. Moreover, AERP dispersion during pac-ing with an AV interval of 0 msec was greaterthan that during pacing with other AV inter-vals. Atrial vulnerability was higher duringpacing with an AV interval of 0 msec (100%)than that during pacing with AV intervals of 60

Chen, et al. Pacing and Atrial Electrophysiology 1581

TABLE 1Relationship Between Different AV Intervals, Hemodynamics, Atrial Size, and Electrophysiologic Characteristics in Group 1

AV Interval (msec)

SBP (mmHg)RAP (mmHg)PCWP (mmHg)RA area (cm^)LA area (cm^)ERP (msec)

RAARPLDCSDispersion

0

108 d10 d21 j

14.6 i15.2 d

188 i258 ;246:7 6 :

: 12: 4b 6t3.8t3.1

t 14t 21!:25t 16

60

118 ± 85 ± 3

15 ± 312.8 ± 3.313.7 ± 3,8

188 ± 15242 ± 17226 ± 2359 ± 12

120

120 ± 144 ± 3

13 ± 410.7 ± 2.911.1 ±3.7

186 ± 13236 ± 20225 ± 2552 ± 17

160

116±65 :

15 I11.8 :12.8 :

184:240:230:6 0 :

t 4t 4!:3.0t3.0

t 12t 15t 17t 18

P Value

<0.05<0.05<0.05<0.05<0.05

NS<0.01<0.01<0.01

DCS = distal coronary sinus; ERP = effective refractory period; LA = left atrium; PCWP = pulmonary capillary wedge pressure; RA = rightatrium; RAA = right atrial appendage; RAP = right atrial pressure; RPL = right posterolateral atrium; SBP = systolic blood pressure.P value refers to a comparison of the parameter at the four different AV intervals.

msec (40%), 120 msec (33%), and 160 msec(40%).

Table 2 lists the effects of altering AV inter-vals on the measured parameters in patients whodid not have AF (group II). Similar to the find-ings in group I, there was lower mean bloodpressure, higher mean right atrial pressure, andhigher pulmonary capillary wedge pressure asthe AV intervals were decreased from 160 to 0msec. Atrial size also increased as the AV inter-val decreased to 0 msec. AERP in the right atrialappendage was not changed during different AVintervals. However, as the AV intervals de-creased from 120 to 0 msec, AERP in the rightposterolateral atrium was increased from 228 ±21 msec to 246 ± 1 9 msec and AERP in thedistal coronary sinus was increased from 218 ±15 msec to 233 ± 14 msec. AERP dispersionduring pacing .with an AV interval of 0 msec wasgreater than that during pacing with other AVintervals.

Comparisons Between Groups I and II(Tables 1 and 2)

Hemodynamics

During pacing with different AV intervals, theminimal values of mean systolic blood pressure(108 ± 12 mmHg vs 106 ± 11 mmHg; P >0.05), mean right atrial pressure (4 ± 3 mmHg vs4 ± 2 mmHg; P > 0.05), and pulmonary capil-lary wedge pressure (13 ± 4 mmHg vs 15 ± 3mmHg; P > 0.05), and the maximal values ofmean systolic blood pressure (120 ± 14 mmHgvs 115 ± 11 mmHg; P > 0.05), mean right atrialpressure (10 ± 4 mmHg vs 8 ± 3 mmHg; P >0.05), and pulmonary capillary wedge pressure(21 ± 6 mmHg vs 19 ± 4 mmHg; P > 0.05)were similar between groups I and II. Addition-ally, among different AV intervals, the maximaldifferences of the mean systolic blood pressure(12 ± 8 mmHg vs 9 ± 8 mmHg; P > 0.05),

Relationship

SBP (mmHg)RAP (mmHg)PCWP (mmHg)RA area (cm'̂ )LA area (cm )̂ERP (msec)

RAARPLDCSDispersion

Between Different AV

0

106 ± 118 ± 3

19 ± 412.9 ± 2.513.1 ± 2.6

210 ± 24246 ± 19233 ± 1436 ± 9

TABLE 2Intervals, Herhodynamics, Atrial

AV Interval

60

112 + 96 ± 2

17 ± 311.6 ±2.312.1 ±2.4

202 ± 20232 ± 10224 ± 12

31 ± 8

Size, and Electrophysiologic

(msec)

120

115 ± 114 ± 2

15 ± 310.2 ± 2.410.5 ± 2.1

203 ± 20228 ± 21218 ± 1525 ± 7

Characteristics

160

110± 54 ± 3

15 ± 311.1 ±2.011.4 ± 2.0

204 ± 26236 ± 17228 ± 1132 ± 9

in Group II

P Value

<0.05<0.05<0.05<0.05<0.05

NS<0.01<0.01<0.05

Abbreviations as Table I.P value refers to a comparison of the parameter at the four different AV intervals.

1582 Journal of Cardiovascular Electrophysiology Vol. 10, No. 12, December 1999

20

E 15o

10

20

I•c< 5

Group IGroup II

P < 0.01

T

Minimal Maximal Maximal A

Group IGroup II

P < 0.01

Minimal Maximal Maximal A

Figure 1. Minimal, maximal, and maximal difference f Aj ofthe left atrial area or right atrial area during AV pacingwith different intervals in groups I and II.

the right posterolateral atrium, and the minimal(225 ± 25 msec vs 218 ± 15 msec; P < 0.05)and maximal AERPs (246 ± 25 msec vs 233 ±14 msec; P < 0.05) in the distal coronary sinuswere similar between groups I and II. As the AVintervals were shortened from 160 to 0 msec, themaximal differences of the AERPs in the rightatrial appendage (4 ± 3 msec vs 8 ± 6 msec; P >0.05), right posterolateral atrium (22 ± 14 msecvs 18 ± 12 msec; P > 0.05), and distal coronarysinus (21 ± 18 msec vs 15 ± 16 msec; P > 0.05)were similar between groups I and II.

Both the minimal and maximal AERP disper-sions were greater in group I than in group II(Fig. 2). In addition, the maximal difference ofAERP dispersion was greater in group I than ingroup II.

Parameters of autonomic nervous system

BRS was significantly higher in group I thanin group II (6.6 ± 1.7 vs 3.9 ± 1.0; P < 0.00005;Fig. 3). Nevertheless, all parameters of 24-hourheart rate variability were similar betweengroups I and II (Table 3).

Discussion

mean right atrial pressure (6 ± 2 mmHg vs 4 ±1 mmHg; P > 0.05), and pulmonary capillarywedge pressure (8 ± 6 mmHg vs 4 ± 8 mmHg;P > 0.05) were similar between groups I and II.

Atrial size

During pacing with different AV intervals, theminimal and maximal atrial sizes were similarbetween groups I and II (Fig. 1). However, themaximal differences of right atrial size and leftatrial size were greater in group I than in groupII.

Atrial refractoriness and dispersion

During pacing with different AV intervals, theminimal (184 ± 12 msec vs 202 ± 20 msec; P <0.05) and maximal (188 ± 14 msec vs 210 ± 24msec; P < 0.05) AERPs in the right atrial ap-pendage were shorter in group I than in group II.However, both the minimal (236 ± 20 msec vs228 ± 21 msec; P < 0.05) and maximal AERPs(258 ± 21 msec vs 246 ± 19 msec; P < 0.05) in

Main Findings

This study demonstrated that AERP andAERP dispersion were changed as atrial pressurewas increased during pacing with different AVintervals. Patients with induced AF had greateratrial dispersion and greater increase of atrialsize during AV simultaneously pacing than those

100

a, 80

c.2 60CO

a.a:LU<

40

20

P < 0.005

Group 1Group II

Minimal Maximal Maximal A

Figure 2. Minimal, maximal, and maximal difference (A) ofatrial effective refractory period (AERP) dispersion duringAV pacing with different intervals in groups I and II.

Chen, et al. Pacing and Atrial Electrophysiology 1583

12

10

TO

g ^

P <0.00005

Yt

i

Figure 3. Plot of individual values of baroreflex sensitivity(BRS) in groups I and II.

without AF. Moreover, BRS, but not heart ratevariability, was higher in patients with AF thanthose without AF.

Possible Mechanisms of Development ofParoxysmal AE During Pacing

To evaluate the effects of DDD or VVI pace-maker on atdal electrophysiology and hemody-namics, this study investigated these parametersduring pacing with different AV intervals. Sev-eral studies demonstrated the hemodynamic ben-efit of DDD compared with VVI pacemakers.'-*With AV synchrony, DDD pacing is associatedwith increased cardiac output, increased systemicblood pressure, and decreased atrial pressure. Inaddition, different AV intervals were suggestedto play an important role in the hemodynamicand clinical status of DDD pacemakers.''-'"'-'^ Inthis study, we also demonstrated that pacing withdifferent AV intervals had different effects onhemodynamic performance and atrial size. Asthe AV interval was shortened, patients withinduced AF had greater increase of atrial sizethan those without AF. It is known that atrial sizeis an important determining factor for AF occur-rence.'^ The greater increase of atrial size duringAV simultaneous contraction may provide morewavelets for sustenance of AF in patients receiv-ing VVI pacemakers.

Mechanoelectrical feedback has been shownto play a role in the genesis of cardiac arrhyth-mias.20 Several investigators showed that atrial

refractoriness was increased as atrial pressureincreased.'0-2' Consistent with these findings, wealso demonstrated in this study that increase ofatrial pressure during pacing with different AVintervals increases AERPs in the right postero-lateral atrium and distal coronary sinus. How-ever, similar to our previous study'^ and that ofCalkins et al.,22 AERP in the right atrial append-age was not changed significantly during pacingwith different AV intervals. Although the mech-anism for these differences is not clear, it ispossible that different atrial tissue characteristicsor unequal atrial stretch play a role.'"

Patients with paroxysmal AF had been shownto have a wider atrial dispersion of refractorinessthan those without paroxysmal AF.'̂ -^^ In thisstudy, both the maximal and minimal atrial dis-persions during pacing with different AV inter-vals were greater in patients with induced AFthan in those without AF. Moreover, when theAV interval shortened to 0 msec, patients withinduced AF had greater increase of atrial disper-sion than those without AF. Because dispersionof refractoriness may create more heterogenousconduction, the greater increase of atrial disper-sion during AV simultaneous pacing may facil-itate induction and maintenance of AF, whichwould result in a higher incidence of AF.

Autonomic Nervous System

The autonomic nervous system was proposedto be involved in the genesis of AF, which hasbeen classified according to increased vagal toneor increased adrenergic tone.̂ '* It is unclear, how-ever, whether the nature of the autonomic ner-vous system in patients receiving pacemakerswith or without AF would be different. AlthoughVVI pacemakers have been shown to accentuate

Heart

Mean RR intervalSDNN (msec)rMSSD (msec)PNN50LF (msec^)HF (msec^)LF/HF

TABLE 3Rate Variability in

Group I

844 ± 128152 ± 13 •47 ± 3 26.9 ± 5.0549 ± 519241 ± 2292.3 ± 1.2

Groups I and II

Group II

922 ± 137123 ± 1236 ± 288.7 ± 7.2

695 ± 652353 ± 3201.9 ±0.8

P Value

NSNSNSNSNSNSNS

HF = high-frequency area of power spectral analysis; LF = low-frequency area of power spectral analysis; PNN50 = percentage ofadjacent RR intervals that differed by > 50 msec; rMSSD = rootmean square of differences of successive RR intervals; SDNN =standard deviation of NN intervals.

1584 Journal of Cardiovascular Electrophysiology Vol. 10, No. 12, December 1999

sympathetic tone, which may contribute to thehigher incidence of AF observed in these pa-tients,"-'3 the role of the parasympathetic ner-vous system in the genesis of AF in patientsreceiving pacemakers is unknown. In this study,patients with induced AF had greater BRS thanthose without AF. It is known that BRS is used toassess the vagal reflex of the heart. These find-ings suggest that enhanced vagal tone plays arole in AF occurrence in patients receiving pace-makers. Previous studies showed that increasingatrial pressure activates cardiac mechanorecep-tors and leads to enhancement of vagal activ-ity 25.26 Therefore, it is possible that patients withhigher vagal reflex activity have greater vagaltone during simultaneous AV pacing with theincrease of atrial pressure. Furthermore, acetyl-choline or vagal stimulation could facilitate in-duction and prolong the duration of paroxysmalAF; the increased vagal reflex may raise theoccurrence and maintenance of paroxysmal/^p 27.28 However, in this study, there was similarheart rate variability (including time-domainmeasures and power spectral measures) betweenpatients with and patients without AF. Analysisof heart rate variability has been used as a non-invasive index of parasympathetic tone and sym-pathovagal balance.̂ -̂̂ o The main difference be-tween heart rate variability and BRS is that theformer represents primary tonic vagal activity,whereas BRS represents the capability of theautonomic nervous system to respond to a stim-ulus. Therefore, the results of the present studyindicate that enhanced vagal tone in an instantmay be of particular importance in promoting theoccurrence of AF in these patients.

Study Limitations

The major limitation of this study is that weonly measured changes of atrial effective refrac-toriness at three different sites. One could arguethat the whole atrial dispersion may differ fromthe result of this study. Furthermore, withoutnerve recordings, we could not evaluate vagalactivity directly during the pacing study. Becauseof the short study period, our results may notfully represent the long-term effects of differentAV intervals pacing.

Conclusion

This study demonstrated the effects of pacingwith different AV intervals on the occurrence of

AF. Our results suggest that abnormal electro-physiologic characteristics and enhanced vagalreflex activity play an important role in the oc-currence of AF in patients receiving VVI pace-makers.

References

1. Feuer J, Shandling A, Messenger J, et al; Influence ofcardiac pacing mode on the long-term development ofatrial fibrillation. Am J Cardiol 1989;64; 1376-1379.

2. Markewitz A, Schad N, Hemmer W, et al: What is themost appropriate stimulation mode in patients withsinus node dysfunction? PACE 1986;9;1115-1120.

3. Snoeck J, Decoster H, Verherstraeten M, et al; Evolu-tion of P wave characteristics after pacemaker implan-tation. PACE 1990;13;2091-2095.

4. Hesselson AB, Parsonnet V, Bernstein AD, et al: Del-eterious effects of long-term single-chamber ventricularpacing in'patients with sick sinus syndrome; The hiddenbenefits of dual-chamber pacing. J Am Coll Cardiol1992; 19; 1542-1549

5. Rediker D, Eagle K, Homma S, et al: Clinical andhemodynamic comparison of VVI versus DDD pacingin patients with DDD pacemakers. Am J Cardiol 1988;61:323-329.

6. Videen JS, Huang S, Bazgan I, et al: Hemodynamiccomparison of ventricular pacing, atrioventricular se-quential pacing, and atrial synchronous ventricular pac-ing using radionuclide ventriculography. Am J Cardiol1986;57:1305-1308.

7. Janosik DL, Labovitz AJ; Basic physiology of cardiacpacing. In Ellenbogen KA, Kay GN, Wilkoff BL, eds:Clinical Cardiac Pacing. WB Saunders, Philadelphia,1995, pp. 367-398.

8. Fromer M, Kappenberger L, Babotai I: Subjective andobjective response to single- versus dual-chamber pac-ing. J Electrophysiol 1987,1:343-349.

9. Klein LS, Miles WM, Zipes DP: Effect of atrioventric-ular interval during pacing or reciprocating tachycardiaon atrial size, pressure, and refractory period. Circula-tion 1990;82:60-68.

10. Satoh T, Zipes DP: Unequal atrial stretch in dogs in-creases dispersion of refractoriness conducive to devel-oping atrial fibrillation. J Cardiovasc Electrophysiol1996;7:833-842.

11. Theodorakis G, Kremastinos D, Livanis MME, et al:c-AMP and ANP levels in VVI and DDD pacing withdifferent AV delays during daily activity and exercise.PACE 1990; 13:1773-1778.

12. Linde-Edelstam C, Hjemdahl P, Pehrsson SK, et al: IsDDD pacing superior to VVIR? A study on cardiacsympathetic nerve activity and myocardial oxygen con-sumption at rest and during exercise. PACE 1992;15:425-434.

13. Taylor JA, Morillo CA, Eckberg DL, et al: Highersympathetic nerve activity during ventricular (VVI)

Chen, et al. Pacing and Atrial Electrophysiology 1585

than during dual-chamber (DDD) pacing. J Am CollCardiol 1996;28; 1753-1758.

14. Chen SA, Chiang CE, Yang CJ, et al; Accessory path-way and atrioventricular node reentrant tachycardia inelderly patients; Clinical features, electrophysiologiccharacteristics and results of radiofrequency ablation.J Am Coll Cardiol 1994;23;702-708.

15. Chen YJ, Chen SA, Tai CT, et al; Role of atrial elec-trophysiology and autonomic nervous system in pa-tients with supraventricular tachycardia and paroxys-mal atrial fibrillation. J Am Coll Cardiol 1998;32;732-738.

16. ManolisBommer W, Weinert L, Neumann A, et al;Determination of right atrial and right ventricular sizeby two-dimensional echocardiography. Circulation1979;60;91-100.

17. Guardigli G, Ansani L, Percoco GF, et al; AV delayoptimization and management of DDD paced patientswith dilated cardiomyopathy. PACE 1994;17;1984-1988.

18. Brecker SJD, Xiao HB, Sparrow J, et al; Effects of dualchamber pacing with short atrioventricular delay indilated cardiomyopathy. Lancet 1992;340;1308-1312.

19. Allessie MA, Brugada RJ, Smeets JLRM, et al; Patho-physiology of atrial fibrillation. In Zipes DP, Jalife J,eds; Cardiac Electrophysiology; From Cell to Bedside.WB Saunders, Philadelphia, 1990, pp. 548-559.

20. Nazir SA, Lab MJ; Mecbanoelectric feedback in theatrium of the isolated guinea-pig heart. Cardiovasc Res1996;32;112-119.

21. Klein LS, Miles WM, Zipes DP; Effect of atrioventric-ular interval during pacing or reciprocating tachycardia

on atrial size, pressure, and refractory period. Circula-tion 1990;82;60-68.

22. Calkins H, El-Atassi R, Leon A, et al; Effect of theatrioventricular relationship on atrial refractoriness inhumans. PACE 1992;15;771-778.

23. Misier AR, Opthof T, vanHemel NM, et al; Increaseddispersion of refractoriness in patients with idiopathicparoxysmal atrial fibrillation. J Am Coll Cardiol 1992;19;1531-1535.

24. Coumel P; Neural aspects of paroxysmal atrial fibrilla-tion. In Ealk RH, Podrid PJ, eds; Atrial Fibrillation;Mechanisms and Management. Raven Press, New York1992, pp. 109-125.

25. Thoren P; Role of cardiac vagal C-fibers in cardiovas-cular control. Rev Physiol Biochem Pharmacol 1979;86; 1-94.

26. Hainsworth R; Reflex from the heart. Physiol Rev1991;71;617-658.

27. Zipes DP, Mihalick MJ, Robbins GT; Effects of selec-tive vagal and stellate ganglion stimulation on atrialrefractoriness. Cardiovasc Res 1974;8;647-655.

28. Wang J, Bourne GW, Wang Z, et al; Comparativemechanism of antiarrhythmic drug action in experimen-tal atrial fibrillation; Importance of use-dependent ef-fects of refractoriness. Circulation 1993;88; 1030-1044.

29. Pagani M, Lombardi F, Guzzetti S, et al; Power spectralanalysis of heart rate and arterial pressure variabilitiesas a marker of sympathovagal interaction in man andconscious dog. Circ Res 1986;59;178-193.

30. Malliani A, Pagani M, Lombardi F, et al; Cardiovascu-lar neural regulation explored in the frequency domain.Circulation 1991;84;482-492.