Clin. Cardiol. 16, 133137 (1993) Limited Value of Anaerobic Threshold for Assessing Functional Capacity in Patients with Heart Failure KYOKO MYAC, M.D., HDJ3SUGU ASANO, M.D., SHNJ SHZAKA, M.D., TOMOK KAMEYAMA, M.D., SHCETAKE SASAYAMA, M.D. The Second Department of nternal Medicine, Toyama Medical and Pharmaceutical University, Toyama, Japan Summary: Exercise tolerance was assessed in 146 patients with cardiac dysfunction in terms of anaerobic threshold (ATge). Patients were divided into four classes according to the peak oxygen uptake: Class A (72 patients) exceeding 1000 ml/min; Class B (27 patients) 800999 ml/min; Class C (37 patients) 500799 ml/min, and Class D ( 10 patients) below 500 ml/min. An incidence of the ATge breakpoint was lower in patients of Class C (38%) than in those of either Class B (70%, p c 0.05) or Class A (87%, p c 0.05). The ATge could not be determined in any patients in Class D. The Vslope method improved the ability to determine ATge by 20%. n Classes C and D, ATge detection was precluded considerably by the fact that the initial workloads of exercise test involved oxygen uptake levels already close to or above the ATge. An oscillatory hyperventilation pattern was also significantly related to failure in defining ATge in Class C patients. Of the 5 1 patients whose ATge was undetermined, 9 had an atrial septal defect. n two of these, exerciseinduced righttoleft shunting led to progmsive arterial hypoxemia, and the consequent hyperventilation masked the appearance of ATge. Thus, ATge is virtually undetectable in patients with severe heart failure largely because of the early onset of anaerobic metabolism or abnormal ventilatory responses to exercise. Accordingly, the clinical application of ATge in the assessment of functional capacity would be limited to patients with mild to moderate heart failure. Key words: gas exchange anaerobic threshold, chronic heart failure, oscillatory ventilation, exercise tolerance Address for reprints: Hidetsugu Asanoi, M.D. The Second Department of nternal Medicine Toyama Medical and Pharmaceutical University 2630 Sugitani, Toyamashi Toyarna 93001, Japan Received: November 27, 1991 Accepted with revision: November 3, 1992 ntroduction Graded exercise testing is widely used to evaluate the exercise capacity of the patients with chronic heart failure. Because most of these patients are unable to achieve a plateau in oxygen uptake (maximal oxygen uptake) during exercise, many investigators have used the peak oxygen uptake to quantify exercise capacity. However, this measurement may be influenced not only by the patient s motivation but also by observer bias.2 Therefore, it has been proposed that determination of gas exchange anaerobic threshold (ATge) could provide a useful, ob jective, quantitative, as well as noninvasive index of functional capacity.3~~ ATge has also been used for gracllng the severity of heart failure and for evaluating the efficacy of drug thempy?s6 However, measurement of ATge can be precluded in some patients with heart failure. The purpose of this study was to examine the problems in detecting ATge as an index of functional capacity and to clarify the causes of unsuccessful determination of ATge in patients with heart failure. Materials and Methods Study Population Studies were performed in 146 patients with cardiac dysfunction (88 males and 58 females, aged 58 k 12 years) (Table ). Fortyseven patients were in New York Heart Association Class, 48 were in Class,and 5 1 were in Class m. They had terminated their exercise testing because of dyspnea or leg fatigue at a level of peak oxygen uptake less than 1500 dmin. The causes of cardiac dysfunction included valvular heart disease in 58 patients, old myocardial infarction in 37, dilated cardiomyopathy in 21, congenital heart disease in 17, and other miscellaneous causes such as hypertrophic cardiomyopathy and pericardial disease in 13. Patients with primary pulmonary disease, anemia, or angina pectoris, and those who had suffered a myocardial infarction within the past 3 months were excluded from the study. No patients were taking betablocking or other cardiosuppressive agents. Patients were divided into four classes according to their peak oxygen uptake: Class A (72 patients) reaching or exceeding lo00 mumin; Class B (27 patients) terminating their exercise at 800999 ml/min; Class C
34 Clin. Cardiol. Vol. 16, February 1993 TABLE Profile of patients evaluated A B C D Total Class n=72 n=27 n=37 n=lo n=146 M/F 54/18 17/10 17/20 0/10 88/58 Age(years) 54f10 58f9 61f13 72fY 58f12 Etiology VHD 24 9 19 6 58 OM1 17 10 9 37 DCM 9 5 4 3 21 CHD 5 0 17 Others 2 0 0 13 Classes: A: Peak V02 2 1000ml/min, B: PeakVO2 = 800999mV min, C: Peak302 = 5 ~7~m~min. D PeakVa c ~OOmVmin. p < 0.05 versus Classes A and B. Abbreviations: M = male, F = female, VHD = valvular heart disease, OM1 = old myocardial infarction, DCM = dilated cardiomyopathy, CHD = congenital heart disease. 50t vo, vco, mhin 1,50040 30 1,000 20 500 1 0 A AT J (37 patients) teminating their exercise at 500799 mumin; and Class D (10 patients) failing to reach 500 mumin. Mean age of Class D was significantly higher than that in the other classes. Before the study, each patient performed an incremental test at least once to become familiar with exercise testing. nformed consent was obtained from all patients. Exercise protocol and Data Analysis Patients were evaluated at least 2 h after a meal. They performed exercise while seated upright on an electronically bded cycle ergometer (Corival400, Lode, The Netherlands). nitially the patients performed 3 minunloaded cycling with the work rate increased by 3 to 15 W every min. The work rate increment was individualized depending on the patient's exercise capacity. Heart rate and arrhythmia were monitored and the blood pressure was measured by the cuff method at min intervals throughout testing. Breathbybreath measurement of oxygen uptake, carbon dioxide production, ventilatory volume, and endtidal oxygen concentration (ET02) and endtidal carbon dioxide concentration (ETCO2) were performed using a Minato AE280 (Japan) metabolic measurement cart equipped with an oxygen and carbon dioxide analyzer (Fig. 1 ). Respiratory flow was measured by the thermal dissipate technique. Two experienced reviewers independently judged whether ATge could be reliably detected using the following two criteria:', (1) a systematic increase in the ventilatory equivalent for oxygen up take without an increase in the ventilatory equivalent for carbon dioxide production, and (2) a systematic increase in ET02 without a concomitant decrease in ETCO2. When one or both reviewers failed to determine the ATge, the value was regarded as undetermined. We also determined the ATge by the Vslope method described by Beaver et al." The breathbybreath system was validated by testing healthy male subjects at workloads ranging from 25 to 200 W on a bicycle ergometer or on a treadmill exercise using the Bruce Rest Unloaded '2030 40 50 6070 80 90 100 110 1 w Detection of anaerobic threshold by cardiopulmonary exer FG. 1 cise testing in a patient with cardiac dysfunction (Class A). The anaerobic threshold (AT) is discerned where minute ventilation (VE) and carbon dioxide production (VCOz) begin to increase nonlinearly despite a linear increase in oxygen uptake (Va) and where the endtidal oxygen concentration (ET02) begins to increase without a concomitant decmse in endtidal carbon dioxide concentration (ETCO?). At the anaerobic threshold, ventilatory equivalent for oxygen uptake (VWVa) begins to increase without a simultaneous increase in ventilatory equivalent for carbon dioxide production (VWVC02). protocol. During each selected workload, expired air was collected for 3 min in a Douglas bag, and the oxygen uptake, carbon dioxide production, and ventilatory volume were determined by the Douglas bag method. The gas composition was determined using a zirconium oxygen sensor and an infrared carbon dioxide analyzer. A computerized breathbybreath calculation was made simultaneously with the Douglas bag col
K. Miyagi et al.: Anaerobic threshold in heart failure 135 TABLE n Anaerobic threshold in four classes Class Leg fatiguddyspnea ~eakvio2 (dmin) (dminfl<g) Peak R Conventional method Vslope method ATge (dmin) (mvmin/kg) ATgeReak Vo2 (%) A n=72 47/25 1275 f2196 22.2 f 3.36 1.13 f 0. 146 62 (87%)b 63 (88%)b 801 f 134 14.0 f 2.5 63f 10 B n=27 17/10 881 16.8 f 2. ",b 1.12 f 0.1 16 19 (70%)b 22 (81%o)b 622 f 76" 1 1.9 f 2.2" 71f8 C n=37 12/25 65 1 f 77"~~ 13.6f 2.Wnb 1.09 f 0.19 14 (38%y 21 (57%y 5 13 f 75" 10.7 f 1.6" 75f5 D n= 10 18 420 f 58" 9.4f 1.1" 1.03 f 0. 14" 0 (O%P 2 (20%y Total n= 146 78/68 986 f 346 18.2 jz 5.1 1.11 f0.14 95 (65%) 108 (74%) 723 f 162 13.1 f 2.6 66f 10 ~ Anaerobic threshold (ATge) was detected by both conventional method and Vslope method. The Vslope method improved the incidence of ATge detection. Peak R = respiratory exchange ratio at peak exercise. A, B, C, and D are the same as in Table. p < 0.05 versus Class A, p c 0.05 versus Class D. lection. t was found that measurements of the computerized system were within f 2% of the Douglas bag method. nterobserver variation for ATge assessment in our laboratory was 24 f41 dmin. Data were expressed as mean f standard deviation. The statistical significance of the incidence of ATge detection among the three classes was analyzed by the chisquare test. A level of p c 0.05 was accepted as statistically significant. Results Table 11 lists the data of ATge determination in the 4 classes. All patients terminated their exercise because of leg fatigue (78 patients) or dyspnea (68 patients). n contrast to Classes A and B, patients in Classes C and D complained of dyspnea more than of leg fatigue at peak exercise. Mean respiratory exchange ratio at the symptomatic maximal exercise was greater than 1.O in all classes. However, the ratio in Class D was significantly lower than that in Classes A and B. Measurement of ATge was achieved in 65% of all patients, 87% in Class A, 70% in Class B, 38% in Class C, and none in Class D. Thus, the determination of ATge became more difticult with a decrease in exercise tolerance. The mean value of ATge was 801 f 134 mvmin in Class A, 622 f 76 dmin in Class B, and 5 13 f 75 dmin in Class C. The ATge detection was improved by use of the Vslope method, to 88% in Class A, 8 1 % in Class B, 57% in Class C, and 20% in Class D. n 20 patients in Classes C and D, ATge could not be determined because of small changes in respiratory gas responses and a premature increase in ventilatory volume and ETOZ at the beginning of incremental exercise (Table ). An oscillatory hyperventilation pattern, which was apparent at low level exercise but diminished during strenuous exercise, was observed in nine patients in Class C (Fig. 2). This peculiar respiratory pattern completely obscured the breakpoint of ATge. rregular respiration commonly observed in all classes was another im@ TABLE threshold Causes of unsuccessful determination of anaerobic A B C D Total Class n=72 n=27 n=37 n=lo n=146 Number 10 8 23 10 51 Causes rregular respiration 5 7 4" 3 19 Low exercise tolerance 0 0 10 10 20 Oscillatory ventilation 0 0 9 0 P Atrialseptaldefect 4 1 4 0 9 mer 1 0 0 0 1 Number = number of patients in whom conventional method could not define ATge. A, B, C, and D are the same as in Table. these patients also had low exercise tolerance. ment to the detection of ATge. An additional nine patients in whom ATge was undetermined had an atrial septa1 defect. The dye dilution technique was carried out to monitor intracardiac shunt during exercise. n two of these patients, the shunt was exclusively directed from left to right, but inmmental exercise provoked a righttoleft shunt with a pmpssive fall in arterial partial oxygen tension (Fig. 3). n this patient, hyperventilation originating from hypoxemia masked the ATge point completely. Discussion Patients with severe congestive heart failure have a limited exercise capacity and were rapidly exhausted during a graded exercise because of leg fatigue or dyspnea. Although there was a little Merence in symptoms at peak exercise between Classes A and B and Classes C and D, almost all patients achieved respiratory gas exchange ratio near or above 1.O where oxygen
136 Clin. Cardiol. Vol. 16, February 1993 Respiratory flow vopvcop ml/min 1,000 r i / w Rest 1 Unwed 15 18 21 24 27 30 FG. 2 Respiratory gas responses to incremental exercise in a Class D patient with dilated cardiomyopathy. The anaerobic threshold cannot be reliably defined in this case because of a periodically oscillatingl%,voz, and VCO?. supply is inadequate and anaerobic metabolism contributes to the energy source. Though ATge was reliably identified in most patients with peak oxygen uptake exceeding 800 mumin (Classes A and B), it was not possible to determine ATge in more than half the patients with a moderately severe or severe impairment of functional capacity (Classes C and D). There are several explanations for difficulties in the determination of ATge in patients with low exercise tolerance. The most likely mechanism is that in these patients the initial workloads of exercise test involved oxygen uptake levels already close to or above ATge. Several investigators'. 4* lo have shown that the lower the peak oxygen uptake, the lower ATge in patients with chronic heart failure. Generally, peak oxygen uptake measured fmm respiratory gas exchange serves a noninvasive measure of pump reserve that can be used to predict the heart's maximal cardiac output.' Therefore, the low peak oxygen uptake indicates a depressed response of cardiac output to exercise, resulting in early onset of anaembic metabolism. 40 L 1 0 Rest d 20 30 40 50 60 70 FG. 3 Respiratory gas responses to incremental exercise in a Class A patient with atrial septa1 defect. As shown by the dye dilution curves, a righttoleft shunt is provoked during exercise, causing a progressive fall in arterial oxygen partial pressure (Pa&). The anaerobic threshold point is masked by hyperventilation caused by hypoxemia. Second, in patients with limited exercise capacity (Class C or D), the small number of samples and minimal changes in ventilatory responses obscured the appearance of ATge. To resolve this problem, Beaver ef al? introduced a vslope method for detecting ATge using computerized regression analysis of the carbon dioxide production and oxygen uptake relation. With this method they showed a small interobserver variation in detecting ATge. n the present study, the Vslope method improved the incidence of ATge detection but only by 20% or less. Third, the exercise protocol might influence the detection of ATge. A short test (less than 8 min) with large increments could produce the early onset of anaerobic metabolism, resulting in a vague threshold phenomenon. Buchfuhrer ef d." recom
K. Miyagi et al.: Anaerobic threshold in heart failure 137 mended an 8 to 12 min test in each subject. Although we individualized the increments in work rate according to the exercise capacity of each patient, some of those patients in Class D could not achieve 8 min of exercise. Finally, some abnormal respiratory responses could mask the point of ATge. An oscillatory ventilation pattern during exercise is a peculiar impedment to the determination of ATge. This abnormal ventilation predominated during low level exercise and diminished with vigorous exercise. While the mechanism for this is not fully understood, delay in chemoreceptor signaling or a transient ventilatiodperfusion mismatch has been proposed as a potential cause. Recently, Kremser etal.12 showed that patients who showed oscillatory ventilation had a low exercise tolerance. The present results also demonstrated that patients with severe heart failure developed oscillatory ventilation more frequently. n some patients with an atrial septal defect, exercise induced a transient righttoleft shunt which progressively decreased the arterial oxygen partial pressure as the work rate increased. The hypoxemia that develops during exercise stimulates breathing intensively and therefore blurs the changes in gas exchange caused by anaerobic metabolism. An irregular breathing due to unfamiliarity with cardiopulmonary testing can be a cause of error in measurement of ATge even in patients without heart failure or in patients with mild heart failure. Although all patients employed in the present study had performed the exercise test at least once prior to the study, 12% of the patients in Classes A and B and 15% of those in Classes C and D showed irregular respiration. Conclusion We demonstrated that ATge was not detected with reliability in most patients with severe chronic heart failure. Failure to determine ATge was caused largely by the early onset of anaerobic metabolism or by abnormal ventilatory responses to graded exercise in these patients. Although Nge provides an objective and noninvasive quantification of cardiovascular impairment in metabolic terms, its clinical use would be limited to patients with mild to moderate heart failure. References. Weber KT, Kinasewitz GT, Janicki JS, Fishman AP Oxygen utilization and ventilation during exercise in patients with chronic cardiac failure. Circulation 65, 12131223 (1982) 2. LeJemtel TH, Mancini D, Gumbardo D, Chadwick B: Pitfalls and limitations of maximal oxygen uptake as an index of cardiovascular functional capacity in patients with chronic heart failure. Heart Failure May/June, 112124 (1985) 3. Wasserman K, Mcnroy MB: Detecting the threshold of anaerobic metabolism in cardiac patients during exercise. Am J Cardiol 14, 844852 (1964) 4. Matsumura N, Nishijima H, Kojima S, Hashimoto F, Minami M, Yasuda H: Determination of anaerobic threshold for assessment of functional state in patients with chronic heart failure.. Citrulafion 68, 360367 (1983) 5. Sullivan M, Atwood JE, Myers J, Feuer J, Hall P,Kellerrnan B, Forbes S, Fmlicher V: ncreased exercise capacity after digoxin administration in patients with heart failure. JAm Coll Cardiol 13, 113&1143(1989) 6. toh H, Taniguchi K, Koike A, Doi M: Evaluation of severity of heart failure. using ventilatory gas analysis. Circulation 8l(suppl t), 3137(1990) 7. Davis JA, Vodak P, Whore JH, Vodak J, Kurtz P Anaerobic threshold and maximal aerobic power for three modes of exercise. JAppl fhysiol41.544550 (1976) 8. Wasserman K, Whipp BJ, Koyal SN, Beaver WL: Anaerobic threshold and respiratory gas exchange during exercise. JAppl Physiol35, 236243 (1973) 9. Beaver WL, Wasserman K, Whipp BJ: A new method for detecting anaerobic threshold by gas exchange. JAppl fhysiol 60,20202027 ( 986) 10. Metra M, Raddino R, Cas LD, Visioli 0 Assessment of peak oxygen consumption, lactate and ventilatory thresholds and correlation with resting and exercise hemodynamic data in chronic heart failure. Am J Cardiol65,11271133 (1990) 1. Buchfuhrer MJ, Hansen JE, Robinson TE, Sue DY, Wasserman K, Whipp BJ: Optimizing the exercise protocol for cardiopulmonary assessment. JAppZfhysiol55,155&1564 (1983) 12. Kremser CB, OToole MF, Leff AR: Oscillatory hyperventilation in severe congestive heart failure secondary to idiopathic dilated cardiomyopathy or to ischemic cardiomyopathy. Am J Cardiol59, 900905 ( 1987)