- Email: [email protected]
Improved Ventilatory Response to Exercise after Cardioversion of Chronic Atrial Fibrillation to Sinus Rhythm
Improved Ventilatory Response to Exercise after Cardioversion of Chronic Atrial Fibrillation to Sinus Rhythm* Torbjorn Lundstrom, M. D.; and Osten Karlsson, M. L
The purpose of this study was to assess hemodynamic and respiratory measures of submaximal and maximal exercise performance in patients with chronic atrial 6brillation, before and one month after cardioversion to sinus rhythm. Restoration of sinus rhythm (n= 16) produced signi6cant reductions in resting and exercise heart rates, 14 percent to 20 percent (p
and maximal tolerated work load (+ 6 percent; p
C
rate is adequately controlled.
hronic atrial fibrillation is a common arrhythmia l that carries important prognostic and therapeutic implications. Hemodynamic deterioration is likely to develop following the onset of atrial fibrillation (AF). 2 Ventricular filling may be diminished due to the rapid and irregular ventricular response. 3 Loss of atrial mechanical function further reduces ventricular filling and may lead to increased atrial pressure and congestion. 4 Only recently has the influence of cardioversion on maximal exercise capacity been studied. 5 ,6 A gradual return of atrial function after cardioversion is thought to explain delayed improvement in exercise performance. 5 •7 Findings in pacemaker-treated patients suggest, however, that the atrial contribution is of minor importance for cardiac output during heavy exercise. S Antiarrhythmic treatment instituted to prevent recurrence of AF, eg, disopyramide, may counteract hemodynamic improvement due to a negative inotropic effect. 9 The present study evaluates hemodynamic and respiratory gas exchange variables during exercise before and after cardioversion. The aim of the investigation was to determine whether restoration of sinus rhythm improves submaximal and maximal exercise performance in patients with AF in whom ventricular ·From the Department of Cardiology, Central Hospital, Skovde, Sweden. This study was supported by a grant from the Swedish Heart and Lung Foundation, Stockholm, Sweden. Manuscript received October 28, 1991; revision accepted March 13. Reprint requests: Dr. Lundstrom, Department of Cardiology, Central Hospital, S-541 85 Skovda, Sweckn
AlE ratio = ratio between atrial contribution and early passive filling in mitral blood Row; AF = atrial 6brillation; AFD = atrial 6brillation despite disopyramide; CaCO. arterial carbon dioxide content; C-vCO. = mixed venous carbon dioxide content; PVCO. partial pressure of carbon dioxide in mixed venous blood; SR sinus rhythm; SRD sinus rhythm. with disopyramide; Vco. carbon dioxide elimination; VENco. ratio between minute ventilation and carbon dioxide elimination
=
=
=
=
=
=
METHODS
Study Patients Thirty patients (20 men and 10 women), mean a~e 65 ± 9 years (range, 37 to 78 years), were included in the study f()lIowin~ informed consent. All had chronic AF with a known duration between one month and one year. Exercise toleran(.'e was limited by dyspnea or fatiW1e in all patients, none of them wvin~ a history of angina pectoris. All patients were in New York Heart Association functional class 1 and 2 before cardioversion. Patients re<..-eivin~ f,iblockade were excluded, as were patients with atrioventricular (AV) block 2 to 3 when in sinus rhythm (SR), ventricular arrhythmias, bronchopulmonary disease, or other serious illnesses likely to affect the evaluation of exercise performance. Five patients were excluded during the course of the study due to sudden death before cardioversion (n = 1), intolerability to disopyramide (n = 2), or ventricular arrhythmias (n = 2). The remaining 25 patients had all been therapeutically antk"Oa~lated f()r at least three weeks before cardioversion. Twenty-three of them were receiving digoxin 0.13 to 0.375 m~ daily. The di~()xin dosa~es were unchanged throughout the study. Six patients were treated with diuretics and one was treated with captopril. Seventeen were given verapamil 80 to 240 mg daily before cardioversion, which \\'as the only ventricular rate-regulating a~ent permitted besides di~()xin.
Our guidelines for the institution of pn)phylactic antiarrhythmic treatment after cardioversion have heen described previously. III After a first cardioversion, performed in 19 of the study patients, all medication was unchanged. The "SR group" (.·omprised nine of these patients who remained in SR. They were studied while receiving the same medication before and one month after cardioversion. Ten patients with relapse of AF proceeded to a se(.'()nd cardioversion. In these, and in seven other patients with recurrent AF after a previous cardioversion (including one patient from the SR group with relapse of AF after two months), disopyramide 250 mg twice daily (slow-release) was instituted the ni~t before CHEST I 102 I 4 I OCTOBER, 1992
1017
Table 1- Patient Characteristics in Different Groups Group Variable Sex, MIF Age, yr, mean±SD Diagnosis (No. of patients) Hypertension Valvular disease Idiopathic Miscellaneous Low ejection fraction «55%) Verapamil treatment Steady-state loads, W 40%
65%
SR (n=9)
SRD (n=7)
AFD (n=10)
613 64±11
413 67±5
6/4 63±8
3 2 3 3 2 5
3 1 3 1 0 6
6
1 2 1 0 7
49±21 82±35
49±14 81±23
58±19 96±31
cardioversion. Verapamil therapy was discontinued when disopyramide therapy was instituted. Thus, they were studied while receiving different medical regimens before and after cardioversion. The "SRD group" comprised seven patients receiving disopyramide who maintained SR after one month, and the uAFD group" comprised ten patients with relapse of AF despite therapy with disopyramide. Baseline characteristics for patients in different groups completing the study protocol are given in Table 1. Data were combined for all patients remaining in SR (n = 16) at the fOllowup test, but they were also analyzed separately for each group. The study protocol was approved by the local Ethics Committee on Human Research.
Study Design The patients performed an initial, maximal, symptom-limited exercise test on a bicycle ergometer. Work loads corresponding to 40 percent and 65 percent of maximal were chosen for the subsequent tests (see Table 1). The latter load exceeded the gas exchange anaerobic threshold, determined as described by Wasserman, II by a mean of 5 W. Based on the ventricular response during this exercise test, the dose of verapamil was adjusted. The target was a maximal heart rate between 150 and 170 beats/min, which in this study was defined as an optimal ventricular rate regulation. Three hours after the maximal symptom-limited test, the patients were accustomed to the following test procedure. After at least 5 min supine rest, a bicycle exercise test was
commenced, starting with no load for the first 2 min. This load was then set at the 40 percent level for 6 to 8 min, after which it was increased to the 65 percent level for another 6 to 8 min. Respiratory gas exchange variables were continuously monitored (using the MGC 2001 system, Medical Graphics Corporation, St Paul, Minnesota). Mean values of oxygen uptake, carbon dioxide elimination, and minute ventilation over 15-s intervals were estimated. Following steady-state development, as visually appreciated by the appearance of a stable plateau in the continuously monitored plot of oxygen uptake, minute ventilation and carbon dioxide elimination for at least 30 s, two cardiac output estimations were performed on each work load. Ventricular rate was determined from 3O-s electrocardiogram (ECG) strips obtained immediately before cardiac output measurements. After 5 min on each load, systolic blood pressure was measured and perceived exertion was evaluated using the Borg 6- to 2o-point scale. 12 When steady-state measurements were concluded, work load was increased by 10 W every minute until symptom-limited maximum. The test protocol is summarized in Figure 1. This exercise test procedure was repeated the day before cardioversion and at a subsequent rehun visit a mean time of 33 days after cardioversion (range, 21 to 55 days). Tests before and after cardioversion were performed at the same time of the day. The patients had refrained from food or smoking for at least 3 h before each test. Cardiac output was estimated using a carbon dioxide rebreathing technique (indirect Fick method). With this method, cardiac output is calculated by the following equation: Cardiac output = Veo/(CvCO z - CaCOJ, where Veo z is the rate of carbon dioxide elimination, and CvC02 - CaC02 is the arteriovenous difference in carbon dioxide content. This difference is calculated from the partial pressures of carbon dioxide in mixed venous (PVCOJ and arterial {PaCOJ blood, respectively.13 PVC0 2 was estimated as described by Jones et all. from the equilibrium plateau obtained after rebreathing from a bag containing 11.3 percent carbon dioxide and 35 percent oxygen in nitrogen, the volume of which was fixed at 1.6 times the tidal volume of the preceding respirations. PaC02 was estimated from the expired endtidal partial pressure of carbon dioxide as gauged from the formula
DSA IrJ SAD IIID AFD
+30 +20
* *
+10
lSI
o
401
-10 -20
o
------'t
+
........--...a.-_ _~+----+- - t - t-t~' ACA
B
B
ACA
B
B
T
•
A
C
FIGURE 1. Exercise test protocol. Assessments: A = oxygen uptake, minute ventilation, respiratory exchange ratio, heart rate; B = cardiac output; C = systolic blood pressure, perceived exertion.
1018
40%
65%
FIGURE 2. Percentage of difference (mean ± SO and range) in cardiac output after cardioversion, during steady-state exercise on work loads corresponding to 40 percent and 65 percent of maximal exercise capacity. AFD = recurrent atrial fibrillation in disopyramide-treated patients: SR and SRn = sinus rhythm without or with disopyramide, respectively. Asterisk = p
Table 2-Hemodynamics at Hest, during Exercise Corresponding to 40 percent and 65 percent of Maximal Exercise Capacity, and at the Maximal Tolerated Uxu1 (Max) before (Atrial Fibrillation; AF) and after (Sinus Rhythm; SH) Cardioversion in Patients Maintaining Sinus Rhythm (n=16)* 40 percent
Supine Rest Mean±SD Heart rate, bpm Systolic B~ mm Hg Cardiac output, Umin Stroke volume, ml
AF SR
AF SR
AF
75± 13 64± 14 143± 19 153±22
Diff -lIt +10+
SR
AF SR
Mean±SD 124± 19 97± 17 176±32 189±37 1l.0±2.4 12.1 ±2.8 91±24 127±37
65 percent Diff -27t +13+ + l.lt +36t
Mean±SD 150±25 119± 17 191 ±36 203±37 14.2±3.1 15.4±4.2 95±18 132±39
Max Diff
-31t + 12t
Mean±SD 174±30 145± 18 204±36 218±38
DifT -29§ +14t
+ 1.2+ +37t
*BP = blood pressure; bpm = beats per minute; Diff = mean difference after vs bef(>re cardioversion. tp
The responses in heart rate and systolic blood pressure before and after cardioversion in patients maintaining SR are presented in Table 2. Heart rate was significantly lower, and systolic blood pressure was significantly higher, after cardioversion to SR, at rest and during exercise. The only significant change in the AFD group was an increase in resting heart rate from 74±9 to 89±21 beats/min (p<0.05).
group, as was an increase on the 65 percent load in the AFD group. Respiratory exchange ratios (carbon dioxide elimination/oxygen uptake, Table 3) were consistent with work above the anaerobic threshold on the 65 percent load and with maximal effort at peak exercise. No significant changes were noted after cardioversion. The ratio between minute ventilation and carbon dioxide elimination (VI-:- Vco2 ) was reduced during exercise in patients remaining in SR (no difference in percent change with or without disopyramide), but was unchanged in the AFD group (Fig 4). There was a significant relationship between change in vElVco2 and change in cardiac output in patients maintaining SR (r= -0.65; p
Cardiac Output and Stroke Volumes
Exercise Capacity
Estimations of cardiac output and stroke volumes during steady-state exercise before and after restitution of SR are outlined in Table 2. Changes in cardiac output in the three different groups are depicted in Figure 2. In the SR group, one patient with aneurysmatic atrial septum had an increase in exercise heart rate and a mean reduction in cardiac output by 7 percent after cardioversion. Overall, significant in-
The maximal tolerated load increased in sinus rhythm from 126 ± 45 to 132 ± 49 W ( + 6 ± 9 percent; p<0.05). The change after cardioversion was in the SR group + 10 ± 7 percent (p<0.05), in the SRD group + 1 ± 8 percent (NS), and in the AFD group - 8 ± 6 percent (p<0.05). Perceived exertion was similar before and after cardioversion in all three groups, averaging 11 ± 1, 15 ± 1, and 18 ± 1 at the 40 percent,
Statistics Data are presented as the mean value ± SD. Changes in variables before and after cardioversion were analyzed by the \Vilcoxon matched paired rank sum test (two-tailed). A p value <0.05 was considered significant. RESULTS
Ventricular Rate and Systolic Blood Pressure
CHEST I 102 I 4 I OCTOBER. 1992
1019
Table 3-Hespiratory Gas Exchange during Exercise before and after Cardioversion to Sinus Rhythm (n = 16)* 40 percent Mean±SD Oxygen uptake, mllkglmin
13.3± 1.4 13.3± 1.6 32.9±5.9 3O.4±5.1 O.94±O.06 0.94±0.03
AF SR AF SR AF SR
Minute ventilation, Umin Respiratory exchange ratio
Max
65 percent Mean±SD
Diff
17.8±2.6 18.1±2.9 49.3± 11.0 44.7±9.0 1.02±0.06 1.00±O.04
±O -2.5t ±O
o iff
Diff
~tean±SD
23.0±4.5 24.7±4.9 75.2±23.4 75.3±22.5 1.13 ± 0.09 1.12±O.O7
+0.3 -4.6:1: -0.02
1.7t +0.1 -0.01
*Explanations as in Table 2. tp
65 percent, and maximal exercise levels, respectively.
Echocardiography-Doppler There were no significant changes in cavity dimensions before and after cardioversion. A waves were identified in all patients with SR; the mean AlE ratio was 1.26±O.62. DISCUSSION
The overall results of this study clearly demonstrate benefits from cardioversion of AF to SR rhythm as regards hemodynamics and respiratory gas exchange during exercise.
Study Technique The noninvasive study technique offers obvious advantages in patients subject to anticoagulation, particularly when repeated observations are needed. Doppler mitral flow tracings (AlE ratios) suggested full recovery of left atrial contractility in patients with
6%
*
+30 +20 +10
*
*
sinus rhythm at the time of the follow-up test. Some concern has been raised as to the reproducibility and validity of the indirect Fick method for estimating cardiac output, 16,17 although excellent correlations with the invasive direct Fick method have been reported. 1H The present study confirms our previous experience with the carbon dioxide rebreathing method. 19 The equilibration technique yields highly reproducible estimates of cardiac output during exercise, fulfilling prerequisites for accurate serial comparisons.
Mechanis1ns of Henlodynamic Inlprovement The present results are consistent with previous invasive findings during exercise. 20-22 After cardioversion to SR, heart rate decreases, but cardiac output increases due to a proportionately larger increase in stroke volume. Theoretically, the necessary enhancement of ventricular filling and emptying may be the result of several contributing factors. The importance of atrial function is supported by a previous study> in which improved exercise capacity after cardioversion was related to resumption of atrial contractility. Considerable increases in resting and exercise cardiac output have been observed as early
45
o
o
oSR+SRD
40
-10
-
-20
-.-.-.-. ~~ ..... .:::..::: "
r-------1
35
30
•• AFD
'
~~
«(<«
-30
40%
65%
Max
FIGllRE 3. Percentage of difference in minute ventilation after cardioversion. Max = maximal tolerated exertion. Other abbreviations and explanations as in Figure 2.
1020
25
--I.----I'~""'__""""'-.........___L_"'L...-..&.............................£._.... ' --1.--1' LOAD
L..'
50
100
150
(watt)
FIGURE 4. Minute ventilation to carbon dioxide elimination ratio CVErVc0 2) during exercise before (squares) and after (circles) cardioversion in patients maintaining sinus rhythm (SR + SRD) or with relapse to atrial fibrillation (AFD). Two asterisks = p
as one to two days after cardioversion, however. 20-22 This increase should not be explained by return of atrial function alone, since at this time the atrial contribution to mitral flow may be modest, gradually increasing over the next month. 5 •7 During exercise, increased venous return and sympathetic stimulation will enhance early ventricular filling. The atrial contribution to mitral flow becomes less impressive during such conditions (Linde-Edelstarn et aI, to be published). Furthermore, in patients with rate-matched pacing, atrioventricular synchrony does not augment exercise cardiac output. 8 Ventricular irregularity has been shown to reduce resting cardiac output in dog experiments,23 but little is known of how exercise cardiac output is affected. Experimental studies in sedated animals support the hypothesis that ventricular filling is inadequate at excessively rapid heart rates. In that model there is a breaking point at approximately 180 beats/min, after which further increases in heart rate result in a decline in cardiac output. 3 Whether a similar breaking point exists during human exercise, and in that case at what heart rate cardiac output will level off, is unknown. In contrast to former studies, rate regulation with verapamil was attempted in patients with exceedingly rapid exercise heart rates in an effort to improve diastolic filling. Such treatment may improve exercise performance in chronic AF.24.25 Still, maximal heart rates were fairly high in our patients during AF, the presence of long R-R intervals at rest and/or side effects preventing increased verapamil dosage in some instances. It seems unlikely, however, that further reductions from the average of 124 (40 percent load) and 150 (65 percent load) beats/min seen during steady-state conditions would enhance cardiac output. This view is supported by findings during exercise in the AFD group. In those patients, the variability in cardiac output measurements was modest, despite a profound variability in heart rate response. Hence, intrinsic myocardial mechanisms seem to balance exercise cardiac output over a wide range of heart rates in AF. It is plausible that several mechanisms operate to improve exercise cardiac output after cardioversion. N either the present study nor previous investigations after cardioversion address the relative importance of different mechanisms. It would be of considerable interest to find an experimental model in which different factors could be modified separately during human exercise. Mechanisrns of Ventilatory Irnprovenl£nt
After reversion to SR, oxygen uptake was unaltered during submaximal work. The previously described increases in maximum oxygen uptake and exercise capacity were confirmed. 5 .6 The present study, how-
ever, is the first to demonstrate significant decreases in ventilation during submaximal exercise. Importantly, an increased maximal oxygen uptake was also achieved without a corresponding increase in ventilatory work. Thus, strong evidence points to an improved efficiency of ventilation, further supported by a favorable shift in the ventilation to carbon dioxide elimination ratio in sinus rhythm, but not in recurrent AF. Similar trends have been reported in a previous study. 6 A decrease in this ratio is considered to be the result of improved perfusion-ventilation matching in the lungs, with less dead space ventilation. 26 The significant relationship between the increase in cardiac output and the reduction of this ratio supports the role of cardiac output in inducing changes in dead space. When work load is increased during exercise, the anaerobic threshold is eventually reached. At this point, carbon dioxide elimination and ventilation start to increase disporportionately to oxygen uptake. 11 This is caused by an increased production ofcarbon dioxide as accumulating lactate is buffered by bicarbonate, and by the subsequent metabolic acidosis developing when buffering capacity is saturated. 27 Consequently, the decrease in ventilation during exercise above the anaerobic threshold may also reflect reduced lactate production. It is possible that in some patients, the anaerobic threshold was delayed in sinus rhythm to an extent that this point was not yet reached at the 65 percent load. Decreased blood lactate levels after cardioversion indicate favorable metabolic adaptations in skeletal muscles as a result of hemodynamic improvement. Such a response has been reported previously only in patients with severe rheumatic heart disease. 22 The improved ventilatory response to exercise was similar to that observed after exercise training in healthy individuals or in patients with congestive heart failure. 28.29 Influence of Disopyramide
Results were assessed in three different groups in an effort to evaluate possible differences in hemodynamic and ventilatory responses attributable to SR and to prophylactic antiarrhythmic treatment. It is obvious that differences in outcome should be interpreted with caution, since there were limited numbers of patients in each group. Dryness of the mouth, a common side effect of disopyramide, was found disturbing by many patients as ventilation approached maximum and it is possible that this symptom limited exercise capacity. From Figure 2 it appears that most of the increase in cardiac output occurred in patients taking disopyramide. If so, this suggests that cardiac output is positively influenced by the change from verapamil treatment before cardioversion to disopyramide after cardioversion. Most probably this is a chance phenomCHEST I 102 I 4 I OCTOBER, 1992
1021
enon, however. The nonsignificance of the change in the SR group was explained by the inclusion of one patient with aneurysmatic atrial septum, in whom hemodynamic changes after cardioversion were opposite the expected. Nevertheless, it is important that cardiac output measurements did not indicate any significant negative inotropic influence of disopyramide either in sinus rhythm or in atrial fibrillation. The decrease in minute ventilation was nonsignificant in the SRD group, and in the AFD group an increase in ventilation was observed on the 65 percent load. It is possible, therefore, that reversion to sinus rhythm and institution of disopyramide treatment affect ventilatory drive in opposite directions.
Conclusions and Clinical Implications Cardioversion of AF to SR induced improvement in hemodynamics and in efficiency of ventilation. Less ventilatory work, with less dyspnea at a given work load, should contribute to symptomatic improvement after restoration of sinus rhythm. Similar hemodynamic responses during exercise with or without disopyramide prophylaxis indicated that the negative inotropic effects of this drug are of minor importance after cardioversion. ACKNOWLEDGMENT: We acknowledge Prof. Lars Ryden for valuable and constructive criticism of the manuscript.
REFERENCES 1 Ostrander LD, Brandt RL, Kjelsberg MO, Epstein FE. Electrocardiographic findings among the adult population of a total natural community, Tecumseh, Michigan. Circulation 1965; 31:888-98 2 Tischler MD, Lee TH, McAndrew KA, Sax PE, St John Sutton M, Lee RT. Clinical, echocardiographic and Doppler correlates ofclinical instability with onset ofatrial fibrillation. Am J Cardiol 1990; 66:721-24 3 Corday E, Irvine DW Hemodynamic disturbances resulting from cardiac arrhythmias. Philadelphia: WB Saunders Co, 1961:207-24 4 Skinner NS, Mitchell JH, Wallace AG, SarnoffSJ. Hemodynamic consequences of atrial fibrillation at constant ventricular rates. Am J Med 1964; 36:342-50 5 Lipkin D~ Frenneaux M, Stewart R, Joshi J, Lowe T, McKenna WJ. Delayed improvement in exercise capacity after cardioversion of atrial fibrillation to sinus rhythm. Br Heart J 1988; 59:572-77 6 AtwoodJE, Myers], Sullivan MA, ForbesS, Sandhu S, Callaham ~ et ale The effect of cardioversion on maximal exercise capacity in patients with chronic atrial fibrillation. Am Heart J 1989; 118:913-18 7 ManningWJ, Leeman DE, Gatch PJ, Come PC. Pulsed Doppler evaluation of atrial mechanical function after electrical cardioversion of atrial fibrillation. J Am Coli Cardiol 1989; 13:617-23 8 Kristensson B-E, Arnman K, Ryden L. The haemodynamic importance of atrioventricular synchrony and rate increase at rest and during exercise. Eur Heart J 1985; 6:773-78 9 Di Bianco R, Gottdiener JS, Singh SN, Fletcher RD. A review of the effects of disopyramide phosphate on left ventricular function and the peripheral circulation. Angiology 1987; 38:17483
1022
10 Lundstrom T, Ryden L. Chronic atrial fibrillation: long-term results of direct current conversion. Acta Med Scand 1988; 223:53-9 11 Wasserman K. The anaerobic threshold measurement to evaluate exercise performance. Am Rev Respir Dis 1984; 129(suppl): S35-S40 12 Borg GAY. Physical performance and perceived exertion (thesis). Lund, Sweden: Berlingska Boktryckeriet, 1962:39-41 13 McHardy GJR. The relationship between the differences in pressure and content of carbon dioxide in arterial and venous blood. Clin Sci 1967; 32:299-309 14 Jones NL, Campbell EJM, McHardy GJR, Higgs BE, Clode M. The estimation of car!x>n dioxide pressure of mixed venous blood during exercise. Clin Sci 1967; 32:311-27 15 Jones NL, Robertson DG, Kane JW Difference between endtidal and arterial Pco 2 in exercise. J Appl Physiol 1979; 47:95460 16 Marks C, Katch V; Rocchini A, Beekman R, Rosenthal A. Validity and reliability of cardiac output by CO 2 rebreathing. Sports Med 1985; 2:432-46 17 Russell AE, Smith SA, West MJ, Aylward PE, McRitchie RJ, Hassam RM, et a1. Automated non-invasive measurement of cardiac output by the carbon dioxide rebreathing method: comparisons with dye dilution and thermodilution. Br Heart J 1990; 63:195-99 18 Stewart RI, Lewis CM. The reliability of the carbon dioxiderebreathing, indirect Fick method of cardiac output determination in patients with pulmonary disease. Clin Sci 1983; 64:28993 19 Lundstrom T, Karlsson 6, Ryden L. Reproducibility of cardiac output estimations using the CO 2 rebreathing method [abstract]. Eur Heart J 1989; 10:203 20 Killip T, Baer RA. Hemodynamic effects after reversion from atrial fibrillation to sinus rhythm by precordial shock. J Clin Invest 1966; 45:658-71 21 Resnekov L. Haemodynamic studies hefore and after electrical conversion of atrial fibrillation and flutter to sinus rhythm. Br Heart J 1967; 29:700-08 22 Kaplan MA, Gray RE, Iseri LT. Metaholic and hemodynamic responses to exercise during atrial fibrillation and sinus rhythm. Am J Cardiol 1968; 22:543-49 23 Naito M, David D, Michelson EL, Schaffenburg M, Dreifus LS. The hemodynamic consequences of cardiac arrhythmias: evaluation of the relative roles of abnormal atrioventricular sequencing, irregularity of ventricular rhythm and atrial fibrillation in a canine model. Am Heart J 1983; 106:284-91 24 Lang R, Klein HO, Di Segni E, Gefen J, Sareli ~ Libhaher C, et a1. Verapamil improves exercise capacity in chronic atrial fibrillation: double-blind crossover study. Am Heart J 1983; 105:820-25 25 Lundstrom T, Ryden L. Ventricular rate control and exercise performance in chronic atrial fibrillation: effects of diltiazem and verapamil. J Am Coli Cardiol 1990; 16:86-90 26 Sullivan MJ, Cobb FR. The anaerobic threshold in chronic heart failure: relation to blood lactate, ventilatory basis, reproducibility and response to exercise training. Circulation 1990; 81(suppl 2):11-47 -II-58 27 Wasserman K, Beaver WL, Whipp BJ. Gas exchange theory and the lactic acidosis (anaerobic) threshold. Circulation 1990; 81(suppI2):11-14-11-30 28 Davis JA, Frank MH, Whipp BJ, Wasserman K. Anaerobic threshold alterations caused by endurance training in middleaged men. J Appl Physiol 1979; 46: 1039-46 29 Sullivan MJ, Hi~inootham MB, Cobb FR. Exercise training in patients with chronic heart failure delays ventilatory anaerobic threshold and improves submaximal exercise performance. Circulation 1989; 79:324-29 Improved Ventilatory Response to Exercise (Lundstrom, Karlsson)
The purpose of this study was to assess hemodynamic and respiratory measures of submaximal and maximal exercise performance in patients with chronic atrial 6brillation, before and one month after cardioversion to sinus rhythm. Restoration of sinus rhythm (n= 16) produced signi6cant reductions in resting and exercise heart rates, 14 percent to 20 percent (p
and maximal tolerated work load (+ 6 percent; p
C
rate is adequately controlled.
hronic atrial fibrillation is a common arrhythmia l that carries important prognostic and therapeutic implications. Hemodynamic deterioration is likely to develop following the onset of atrial fibrillation (AF). 2 Ventricular filling may be diminished due to the rapid and irregular ventricular response. 3 Loss of atrial mechanical function further reduces ventricular filling and may lead to increased atrial pressure and congestion. 4 Only recently has the influence of cardioversion on maximal exercise capacity been studied. 5 ,6 A gradual return of atrial function after cardioversion is thought to explain delayed improvement in exercise performance. 5 •7 Findings in pacemaker-treated patients suggest, however, that the atrial contribution is of minor importance for cardiac output during heavy exercise. S Antiarrhythmic treatment instituted to prevent recurrence of AF, eg, disopyramide, may counteract hemodynamic improvement due to a negative inotropic effect. 9 The present study evaluates hemodynamic and respiratory gas exchange variables during exercise before and after cardioversion. The aim of the investigation was to determine whether restoration of sinus rhythm improves submaximal and maximal exercise performance in patients with AF in whom ventricular ·From the Department of Cardiology, Central Hospital, Skovde, Sweden. This study was supported by a grant from the Swedish Heart and Lung Foundation, Stockholm, Sweden. Manuscript received October 28, 1991; revision accepted March 13. Reprint requests: Dr. Lundstrom, Department of Cardiology, Central Hospital, S-541 85 Skovda, Sweckn
AlE ratio = ratio between atrial contribution and early passive filling in mitral blood Row; AF = atrial 6brillation; AFD = atrial 6brillation despite disopyramide; CaCO. arterial carbon dioxide content; C-vCO. = mixed venous carbon dioxide content; PVCO. partial pressure of carbon dioxide in mixed venous blood; SR sinus rhythm; SRD sinus rhythm. with disopyramide; Vco. carbon dioxide elimination; VENco. ratio between minute ventilation and carbon dioxide elimination
=
=
=
=
=
=
METHODS
Study Patients Thirty patients (20 men and 10 women), mean a~e 65 ± 9 years (range, 37 to 78 years), were included in the study f()lIowin~ informed consent. All had chronic AF with a known duration between one month and one year. Exercise toleran(.'e was limited by dyspnea or fatiW1e in all patients, none of them wvin~ a history of angina pectoris. All patients were in New York Heart Association functional class 1 and 2 before cardioversion. Patients re<..-eivin~ f,iblockade were excluded, as were patients with atrioventricular (AV) block 2 to 3 when in sinus rhythm (SR), ventricular arrhythmias, bronchopulmonary disease, or other serious illnesses likely to affect the evaluation of exercise performance. Five patients were excluded during the course of the study due to sudden death before cardioversion (n = 1), intolerability to disopyramide (n = 2), or ventricular arrhythmias (n = 2). The remaining 25 patients had all been therapeutically antk"Oa~lated f()r at least three weeks before cardioversion. Twenty-three of them were receiving digoxin 0.13 to 0.375 m~ daily. The di~()xin dosa~es were unchanged throughout the study. Six patients were treated with diuretics and one was treated with captopril. Seventeen were given verapamil 80 to 240 mg daily before cardioversion, which \\'as the only ventricular rate-regulating a~ent permitted besides di~()xin.
Our guidelines for the institution of pn)phylactic antiarrhythmic treatment after cardioversion have heen described previously. III After a first cardioversion, performed in 19 of the study patients, all medication was unchanged. The "SR group" (.·omprised nine of these patients who remained in SR. They were studied while receiving the same medication before and one month after cardioversion. Ten patients with relapse of AF proceeded to a se(.'()nd cardioversion. In these, and in seven other patients with recurrent AF after a previous cardioversion (including one patient from the SR group with relapse of AF after two months), disopyramide 250 mg twice daily (slow-release) was instituted the ni~t before CHEST I 102 I 4 I OCTOBER, 1992
1017
Table 1- Patient Characteristics in Different Groups Group Variable Sex, MIF Age, yr, mean±SD Diagnosis (No. of patients) Hypertension Valvular disease Idiopathic Miscellaneous Low ejection fraction «55%) Verapamil treatment Steady-state loads, W 40%
65%
SR (n=9)
SRD (n=7)
AFD (n=10)
613 64±11
413 67±5
6/4 63±8
3 2 3 3 2 5
3 1 3 1 0 6
6
1 2 1 0 7
49±21 82±35
49±14 81±23
58±19 96±31
cardioversion. Verapamil therapy was discontinued when disopyramide therapy was instituted. Thus, they were studied while receiving different medical regimens before and after cardioversion. The "SRD group" comprised seven patients receiving disopyramide who maintained SR after one month, and the uAFD group" comprised ten patients with relapse of AF despite therapy with disopyramide. Baseline characteristics for patients in different groups completing the study protocol are given in Table 1. Data were combined for all patients remaining in SR (n = 16) at the fOllowup test, but they were also analyzed separately for each group. The study protocol was approved by the local Ethics Committee on Human Research.
Study Design The patients performed an initial, maximal, symptom-limited exercise test on a bicycle ergometer. Work loads corresponding to 40 percent and 65 percent of maximal were chosen for the subsequent tests (see Table 1). The latter load exceeded the gas exchange anaerobic threshold, determined as described by Wasserman, II by a mean of 5 W. Based on the ventricular response during this exercise test, the dose of verapamil was adjusted. The target was a maximal heart rate between 150 and 170 beats/min, which in this study was defined as an optimal ventricular rate regulation. Three hours after the maximal symptom-limited test, the patients were accustomed to the following test procedure. After at least 5 min supine rest, a bicycle exercise test was
commenced, starting with no load for the first 2 min. This load was then set at the 40 percent level for 6 to 8 min, after which it was increased to the 65 percent level for another 6 to 8 min. Respiratory gas exchange variables were continuously monitored (using the MGC 2001 system, Medical Graphics Corporation, St Paul, Minnesota). Mean values of oxygen uptake, carbon dioxide elimination, and minute ventilation over 15-s intervals were estimated. Following steady-state development, as visually appreciated by the appearance of a stable plateau in the continuously monitored plot of oxygen uptake, minute ventilation and carbon dioxide elimination for at least 30 s, two cardiac output estimations were performed on each work load. Ventricular rate was determined from 3O-s electrocardiogram (ECG) strips obtained immediately before cardiac output measurements. After 5 min on each load, systolic blood pressure was measured and perceived exertion was evaluated using the Borg 6- to 2o-point scale. 12 When steady-state measurements were concluded, work load was increased by 10 W every minute until symptom-limited maximum. The test protocol is summarized in Figure 1. This exercise test procedure was repeated the day before cardioversion and at a subsequent rehun visit a mean time of 33 days after cardioversion (range, 21 to 55 days). Tests before and after cardioversion were performed at the same time of the day. The patients had refrained from food or smoking for at least 3 h before each test. Cardiac output was estimated using a carbon dioxide rebreathing technique (indirect Fick method). With this method, cardiac output is calculated by the following equation: Cardiac output = Veo/(CvCO z - CaCOJ, where Veo z is the rate of carbon dioxide elimination, and CvC02 - CaC02 is the arteriovenous difference in carbon dioxide content. This difference is calculated from the partial pressures of carbon dioxide in mixed venous (PVCOJ and arterial {PaCOJ blood, respectively.13 PVC0 2 was estimated as described by Jones et all. from the equilibrium plateau obtained after rebreathing from a bag containing 11.3 percent carbon dioxide and 35 percent oxygen in nitrogen, the volume of which was fixed at 1.6 times the tidal volume of the preceding respirations. PaC02 was estimated from the expired endtidal partial pressure of carbon dioxide as gauged from the formula
DSA IrJ SAD IIID AFD
+30 +20
* *
+10
lSI
o
401
-10 -20
o
------'t
+
........--...a.-_ _~+----+- - t - t-t~' ACA
B
B
ACA
B
B
T
•
A
C
FIGURE 1. Exercise test protocol. Assessments: A = oxygen uptake, minute ventilation, respiratory exchange ratio, heart rate; B = cardiac output; C = systolic blood pressure, perceived exertion.
1018
40%
65%
FIGURE 2. Percentage of difference (mean ± SO and range) in cardiac output after cardioversion, during steady-state exercise on work loads corresponding to 40 percent and 65 percent of maximal exercise capacity. AFD = recurrent atrial fibrillation in disopyramide-treated patients: SR and SRn = sinus rhythm without or with disopyramide, respectively. Asterisk = p
Table 2-Hemodynamics at Hest, during Exercise Corresponding to 40 percent and 65 percent of Maximal Exercise Capacity, and at the Maximal Tolerated Uxu1 (Max) before (Atrial Fibrillation; AF) and after (Sinus Rhythm; SH) Cardioversion in Patients Maintaining Sinus Rhythm (n=16)* 40 percent
Supine Rest Mean±SD Heart rate, bpm Systolic B~ mm Hg Cardiac output, Umin Stroke volume, ml
AF SR
AF SR
AF
75± 13 64± 14 143± 19 153±22
Diff -lIt +10+
SR
AF SR
Mean±SD 124± 19 97± 17 176±32 189±37 1l.0±2.4 12.1 ±2.8 91±24 127±37
65 percent Diff -27t +13+ + l.lt +36t
Mean±SD 150±25 119± 17 191 ±36 203±37 14.2±3.1 15.4±4.2 95±18 132±39
Max Diff
-31t + 12t
Mean±SD 174±30 145± 18 204±36 218±38
DifT -29§ +14t
+ 1.2+ +37t
*BP = blood pressure; bpm = beats per minute; Diff = mean difference after vs bef(>re cardioversion. tp
The responses in heart rate and systolic blood pressure before and after cardioversion in patients maintaining SR are presented in Table 2. Heart rate was significantly lower, and systolic blood pressure was significantly higher, after cardioversion to SR, at rest and during exercise. The only significant change in the AFD group was an increase in resting heart rate from 74±9 to 89±21 beats/min (p<0.05).
group, as was an increase on the 65 percent load in the AFD group. Respiratory exchange ratios (carbon dioxide elimination/oxygen uptake, Table 3) were consistent with work above the anaerobic threshold on the 65 percent load and with maximal effort at peak exercise. No significant changes were noted after cardioversion. The ratio between minute ventilation and carbon dioxide elimination (VI-:- Vco2 ) was reduced during exercise in patients remaining in SR (no difference in percent change with or without disopyramide), but was unchanged in the AFD group (Fig 4). There was a significant relationship between change in vElVco2 and change in cardiac output in patients maintaining SR (r= -0.65; p
Cardiac Output and Stroke Volumes
Exercise Capacity
Estimations of cardiac output and stroke volumes during steady-state exercise before and after restitution of SR are outlined in Table 2. Changes in cardiac output in the three different groups are depicted in Figure 2. In the SR group, one patient with aneurysmatic atrial septum had an increase in exercise heart rate and a mean reduction in cardiac output by 7 percent after cardioversion. Overall, significant in-
The maximal tolerated load increased in sinus rhythm from 126 ± 45 to 132 ± 49 W ( + 6 ± 9 percent; p<0.05). The change after cardioversion was in the SR group + 10 ± 7 percent (p<0.05), in the SRD group + 1 ± 8 percent (NS), and in the AFD group - 8 ± 6 percent (p<0.05). Perceived exertion was similar before and after cardioversion in all three groups, averaging 11 ± 1, 15 ± 1, and 18 ± 1 at the 40 percent,
Statistics Data are presented as the mean value ± SD. Changes in variables before and after cardioversion were analyzed by the \Vilcoxon matched paired rank sum test (two-tailed). A p value <0.05 was considered significant. RESULTS
Ventricular Rate and Systolic Blood Pressure
CHEST I 102 I 4 I OCTOBER. 1992
1019
Table 3-Hespiratory Gas Exchange during Exercise before and after Cardioversion to Sinus Rhythm (n = 16)* 40 percent Mean±SD Oxygen uptake, mllkglmin
13.3± 1.4 13.3± 1.6 32.9±5.9 3O.4±5.1 O.94±O.06 0.94±0.03
AF SR AF SR AF SR
Minute ventilation, Umin Respiratory exchange ratio
Max
65 percent Mean±SD
Diff
17.8±2.6 18.1±2.9 49.3± 11.0 44.7±9.0 1.02±0.06 1.00±O.04
±O -2.5t ±O
o iff
Diff
~tean±SD
23.0±4.5 24.7±4.9 75.2±23.4 75.3±22.5 1.13 ± 0.09 1.12±O.O7
+0.3 -4.6:1: -0.02
1.7t +0.1 -0.01
*Explanations as in Table 2. tp
65 percent, and maximal exercise levels, respectively.
Echocardiography-Doppler There were no significant changes in cavity dimensions before and after cardioversion. A waves were identified in all patients with SR; the mean AlE ratio was 1.26±O.62. DISCUSSION
The overall results of this study clearly demonstrate benefits from cardioversion of AF to SR rhythm as regards hemodynamics and respiratory gas exchange during exercise.
Study Technique The noninvasive study technique offers obvious advantages in patients subject to anticoagulation, particularly when repeated observations are needed. Doppler mitral flow tracings (AlE ratios) suggested full recovery of left atrial contractility in patients with
6%
*
+30 +20 +10
*
*
sinus rhythm at the time of the follow-up test. Some concern has been raised as to the reproducibility and validity of the indirect Fick method for estimating cardiac output, 16,17 although excellent correlations with the invasive direct Fick method have been reported. 1H The present study confirms our previous experience with the carbon dioxide rebreathing method. 19 The equilibration technique yields highly reproducible estimates of cardiac output during exercise, fulfilling prerequisites for accurate serial comparisons.
Mechanis1ns of Henlodynamic Inlprovement The present results are consistent with previous invasive findings during exercise. 20-22 After cardioversion to SR, heart rate decreases, but cardiac output increases due to a proportionately larger increase in stroke volume. Theoretically, the necessary enhancement of ventricular filling and emptying may be the result of several contributing factors. The importance of atrial function is supported by a previous study> in which improved exercise capacity after cardioversion was related to resumption of atrial contractility. Considerable increases in resting and exercise cardiac output have been observed as early
45
o
o
oSR+SRD
40
-10
-
-20
-.-.-.-. ~~ ..... .:::..::: "
r-------1
35
30
•• AFD
'
~~
«(<«
-30
40%
65%
Max
FIGllRE 3. Percentage of difference in minute ventilation after cardioversion. Max = maximal tolerated exertion. Other abbreviations and explanations as in Figure 2.
1020
25
--I.----I'~""'__""""'-.........___L_"'L...-..&.............................£._.... ' --1.--1' LOAD
L..'
50
100
150
(watt)
FIGURE 4. Minute ventilation to carbon dioxide elimination ratio CVErVc0 2) during exercise before (squares) and after (circles) cardioversion in patients maintaining sinus rhythm (SR + SRD) or with relapse to atrial fibrillation (AFD). Two asterisks = p
as one to two days after cardioversion, however. 20-22 This increase should not be explained by return of atrial function alone, since at this time the atrial contribution to mitral flow may be modest, gradually increasing over the next month. 5 •7 During exercise, increased venous return and sympathetic stimulation will enhance early ventricular filling. The atrial contribution to mitral flow becomes less impressive during such conditions (Linde-Edelstarn et aI, to be published). Furthermore, in patients with rate-matched pacing, atrioventricular synchrony does not augment exercise cardiac output. 8 Ventricular irregularity has been shown to reduce resting cardiac output in dog experiments,23 but little is known of how exercise cardiac output is affected. Experimental studies in sedated animals support the hypothesis that ventricular filling is inadequate at excessively rapid heart rates. In that model there is a breaking point at approximately 180 beats/min, after which further increases in heart rate result in a decline in cardiac output. 3 Whether a similar breaking point exists during human exercise, and in that case at what heart rate cardiac output will level off, is unknown. In contrast to former studies, rate regulation with verapamil was attempted in patients with exceedingly rapid exercise heart rates in an effort to improve diastolic filling. Such treatment may improve exercise performance in chronic AF.24.25 Still, maximal heart rates were fairly high in our patients during AF, the presence of long R-R intervals at rest and/or side effects preventing increased verapamil dosage in some instances. It seems unlikely, however, that further reductions from the average of 124 (40 percent load) and 150 (65 percent load) beats/min seen during steady-state conditions would enhance cardiac output. This view is supported by findings during exercise in the AFD group. In those patients, the variability in cardiac output measurements was modest, despite a profound variability in heart rate response. Hence, intrinsic myocardial mechanisms seem to balance exercise cardiac output over a wide range of heart rates in AF. It is plausible that several mechanisms operate to improve exercise cardiac output after cardioversion. N either the present study nor previous investigations after cardioversion address the relative importance of different mechanisms. It would be of considerable interest to find an experimental model in which different factors could be modified separately during human exercise. Mechanisrns of Ventilatory Irnprovenl£nt
After reversion to SR, oxygen uptake was unaltered during submaximal work. The previously described increases in maximum oxygen uptake and exercise capacity were confirmed. 5 .6 The present study, how-
ever, is the first to demonstrate significant decreases in ventilation during submaximal exercise. Importantly, an increased maximal oxygen uptake was also achieved without a corresponding increase in ventilatory work. Thus, strong evidence points to an improved efficiency of ventilation, further supported by a favorable shift in the ventilation to carbon dioxide elimination ratio in sinus rhythm, but not in recurrent AF. Similar trends have been reported in a previous study. 6 A decrease in this ratio is considered to be the result of improved perfusion-ventilation matching in the lungs, with less dead space ventilation. 26 The significant relationship between the increase in cardiac output and the reduction of this ratio supports the role of cardiac output in inducing changes in dead space. When work load is increased during exercise, the anaerobic threshold is eventually reached. At this point, carbon dioxide elimination and ventilation start to increase disporportionately to oxygen uptake. 11 This is caused by an increased production ofcarbon dioxide as accumulating lactate is buffered by bicarbonate, and by the subsequent metabolic acidosis developing when buffering capacity is saturated. 27 Consequently, the decrease in ventilation during exercise above the anaerobic threshold may also reflect reduced lactate production. It is possible that in some patients, the anaerobic threshold was delayed in sinus rhythm to an extent that this point was not yet reached at the 65 percent load. Decreased blood lactate levels after cardioversion indicate favorable metabolic adaptations in skeletal muscles as a result of hemodynamic improvement. Such a response has been reported previously only in patients with severe rheumatic heart disease. 22 The improved ventilatory response to exercise was similar to that observed after exercise training in healthy individuals or in patients with congestive heart failure. 28.29 Influence of Disopyramide
Results were assessed in three different groups in an effort to evaluate possible differences in hemodynamic and ventilatory responses attributable to SR and to prophylactic antiarrhythmic treatment. It is obvious that differences in outcome should be interpreted with caution, since there were limited numbers of patients in each group. Dryness of the mouth, a common side effect of disopyramide, was found disturbing by many patients as ventilation approached maximum and it is possible that this symptom limited exercise capacity. From Figure 2 it appears that most of the increase in cardiac output occurred in patients taking disopyramide. If so, this suggests that cardiac output is positively influenced by the change from verapamil treatment before cardioversion to disopyramide after cardioversion. Most probably this is a chance phenomCHEST I 102 I 4 I OCTOBER, 1992
1021
enon, however. The nonsignificance of the change in the SR group was explained by the inclusion of one patient with aneurysmatic atrial septum, in whom hemodynamic changes after cardioversion were opposite the expected. Nevertheless, it is important that cardiac output measurements did not indicate any significant negative inotropic influence of disopyramide either in sinus rhythm or in atrial fibrillation. The decrease in minute ventilation was nonsignificant in the SRD group, and in the AFD group an increase in ventilation was observed on the 65 percent load. It is possible, therefore, that reversion to sinus rhythm and institution of disopyramide treatment affect ventilatory drive in opposite directions.
Conclusions and Clinical Implications Cardioversion of AF to SR induced improvement in hemodynamics and in efficiency of ventilation. Less ventilatory work, with less dyspnea at a given work load, should contribute to symptomatic improvement after restoration of sinus rhythm. Similar hemodynamic responses during exercise with or without disopyramide prophylaxis indicated that the negative inotropic effects of this drug are of minor importance after cardioversion. ACKNOWLEDGMENT: We acknowledge Prof. Lars Ryden for valuable and constructive criticism of the manuscript.
REFERENCES 1 Ostrander LD, Brandt RL, Kjelsberg MO, Epstein FE. Electrocardiographic findings among the adult population of a total natural community, Tecumseh, Michigan. Circulation 1965; 31:888-98 2 Tischler MD, Lee TH, McAndrew KA, Sax PE, St John Sutton M, Lee RT. Clinical, echocardiographic and Doppler correlates ofclinical instability with onset ofatrial fibrillation. Am J Cardiol 1990; 66:721-24 3 Corday E, Irvine DW Hemodynamic disturbances resulting from cardiac arrhythmias. Philadelphia: WB Saunders Co, 1961:207-24 4 Skinner NS, Mitchell JH, Wallace AG, SarnoffSJ. Hemodynamic consequences of atrial fibrillation at constant ventricular rates. Am J Med 1964; 36:342-50 5 Lipkin D~ Frenneaux M, Stewart R, Joshi J, Lowe T, McKenna WJ. Delayed improvement in exercise capacity after cardioversion of atrial fibrillation to sinus rhythm. Br Heart J 1988; 59:572-77 6 AtwoodJE, Myers], Sullivan MA, ForbesS, Sandhu S, Callaham ~ et ale The effect of cardioversion on maximal exercise capacity in patients with chronic atrial fibrillation. Am Heart J 1989; 118:913-18 7 ManningWJ, Leeman DE, Gatch PJ, Come PC. Pulsed Doppler evaluation of atrial mechanical function after electrical cardioversion of atrial fibrillation. J Am Coli Cardiol 1989; 13:617-23 8 Kristensson B-E, Arnman K, Ryden L. The haemodynamic importance of atrioventricular synchrony and rate increase at rest and during exercise. Eur Heart J 1985; 6:773-78 9 Di Bianco R, Gottdiener JS, Singh SN, Fletcher RD. A review of the effects of disopyramide phosphate on left ventricular function and the peripheral circulation. Angiology 1987; 38:17483
1022
10 Lundstrom T, Ryden L. Chronic atrial fibrillation: long-term results of direct current conversion. Acta Med Scand 1988; 223:53-9 11 Wasserman K. The anaerobic threshold measurement to evaluate exercise performance. Am Rev Respir Dis 1984; 129(suppl): S35-S40 12 Borg GAY. Physical performance and perceived exertion (thesis). Lund, Sweden: Berlingska Boktryckeriet, 1962:39-41 13 McHardy GJR. The relationship between the differences in pressure and content of carbon dioxide in arterial and venous blood. Clin Sci 1967; 32:299-309 14 Jones NL, Campbell EJM, McHardy GJR, Higgs BE, Clode M. The estimation of car!x>n dioxide pressure of mixed venous blood during exercise. Clin Sci 1967; 32:311-27 15 Jones NL, Robertson DG, Kane JW Difference between endtidal and arterial Pco 2 in exercise. J Appl Physiol 1979; 47:95460 16 Marks C, Katch V; Rocchini A, Beekman R, Rosenthal A. Validity and reliability of cardiac output by CO 2 rebreathing. Sports Med 1985; 2:432-46 17 Russell AE, Smith SA, West MJ, Aylward PE, McRitchie RJ, Hassam RM, et a1. Automated non-invasive measurement of cardiac output by the carbon dioxide rebreathing method: comparisons with dye dilution and thermodilution. Br Heart J 1990; 63:195-99 18 Stewart RI, Lewis CM. The reliability of the carbon dioxiderebreathing, indirect Fick method of cardiac output determination in patients with pulmonary disease. Clin Sci 1983; 64:28993 19 Lundstrom T, Karlsson 6, Ryden L. Reproducibility of cardiac output estimations using the CO 2 rebreathing method [abstract]. Eur Heart J 1989; 10:203 20 Killip T, Baer RA. Hemodynamic effects after reversion from atrial fibrillation to sinus rhythm by precordial shock. J Clin Invest 1966; 45:658-71 21 Resnekov L. Haemodynamic studies hefore and after electrical conversion of atrial fibrillation and flutter to sinus rhythm. Br Heart J 1967; 29:700-08 22 Kaplan MA, Gray RE, Iseri LT. Metaholic and hemodynamic responses to exercise during atrial fibrillation and sinus rhythm. Am J Cardiol 1968; 22:543-49 23 Naito M, David D, Michelson EL, Schaffenburg M, Dreifus LS. The hemodynamic consequences of cardiac arrhythmias: evaluation of the relative roles of abnormal atrioventricular sequencing, irregularity of ventricular rhythm and atrial fibrillation in a canine model. Am Heart J 1983; 106:284-91 24 Lang R, Klein HO, Di Segni E, Gefen J, Sareli ~ Libhaher C, et a1. Verapamil improves exercise capacity in chronic atrial fibrillation: double-blind crossover study. Am Heart J 1983; 105:820-25 25 Lundstrom T, Ryden L. Ventricular rate control and exercise performance in chronic atrial fibrillation: effects of diltiazem and verapamil. J Am Coli Cardiol 1990; 16:86-90 26 Sullivan MJ, Cobb FR. The anaerobic threshold in chronic heart failure: relation to blood lactate, ventilatory basis, reproducibility and response to exercise training. Circulation 1990; 81(suppl 2):11-47 -II-58 27 Wasserman K, Beaver WL, Whipp BJ. Gas exchange theory and the lactic acidosis (anaerobic) threshold. Circulation 1990; 81(suppI2):11-14-11-30 28 Davis JA, Frank MH, Whipp BJ, Wasserman K. Anaerobic threshold alterations caused by endurance training in middleaged men. J Appl Physiol 1979; 46: 1039-46 29 Sullivan MJ, Hi~inootham MB, Cobb FR. Exercise training in patients with chronic heart failure delays ventilatory anaerobic threshold and improves submaximal exercise performance. Circulation 1989; 79:324-29 Improved Ventilatory Response to Exercise (Lundstrom, Karlsson)