- Email: [email protected]
Pediatric exercise testing
Pediatric Exercise Testing Methodology,
Equipment, and Normal Values
GERALD
BARBER,
M.D.
De~~~~e~t of Pediatrics Nau York University School of Medicine New York, New York
Exercise testing is an important tool in the evaluation of cardiac and respiratory problems as well as physical working capacity. By taxing the cardiopulmonary system, exercise determines the patient‘s physiologic reserve, enabling the early detection of abnormalities not apparent at rest1 Testing quantifies a patient’s exercise ability in a way that questionnaires or other subjective estimates cannot .2,3 Exercise testing requires a thorough understanding of methodology, equipment, and normal values. This article will provide a basic review of these areas, but an understanding sufficient to perform exercise testing can only come with supervised experience in an exercise laboratory. Recently the American College of Physicians, the American College of Cardiology, and the American Heart Association published a task force statement on clinical competence in exercise testing.* They recommend that an individual perform at least 50 exercise tests in a supervised training setting and at least 25 per year thereafter to acquire and maintain minimum clinical competence. These recommendations, based mainly on testing in adults with coronary artery disease, may underestimate the requirements for clinical competence in pediatric exercise testing.
Address correspondence to Gerald Barber, M.D., Pediatric Cardiology, NYLJ School of Medicine, l-i-501, 560 First Avenue, New York, NY 10016-6402.
METHODOLOGY The protocol used for exercise testing depends largely on the purpose of the test. Further consideration must be paid to the age, size, and estimated exercise ability of the patient. One of the first issues to be decided is whether one wants to assess aerobic or anaerobic capability. Aerobic capability usually is measured by indices such as peak oxygen uptake and anaerobic threshold, whereas anaerobic capability usually is measured by indices such as peak power and fatigue rate. Anaerobic Capability Protocols to measure aerobic power (physical working capacity) are useful in assessing cardiac and pulmonary limitations as well as total exercise performance. These protocols can be divided by the exercise instrument (treadmill or cycle ergometer) and by the way in which workloads are changed (incremental or ramp protocols). The Bruce protocol is the principal treadmill protocol used today.5 It is an incremental protocol in which both speed and elevation are increased every 3 minutes until the individual can go no further (Table 1). As with all incremental protocols, it is limited in that measurements are made at only a certain number of preset workloads, independent of the subject’s exercise ability: therefore, the information obtained is inversely related to the patient’s degree of limitation. Also, in the Bruce protocol the work performed is dependent on the weight of the patient and the efficiency of gait. This is an Prog Pediatr Cardiof 1993; 2(2):4-10 Copyright 0 1993 by Andover Medical
Exercise Testing
TABLE
1. Bruce Exercise Protocol
5
TABLE 2. James Exercise Protocol
Stage
Speed (miles/h)
Elevation (%)
1 2 3 4 5 6 7
1.7 2.5 3.4 4.2 5.0 5.5 6.0
10 12 14 16 18 20 22
Body Surface Area
1.0-1.19
Stage
(kpm/min)
(kpm/min)
1 2 3 4 5 6 7
200 300 500 600 700 800 900
200 400 600 700 800 900 1000
200 500 800 1000 1200 1400 1600
m2
al.2
m2
l6.12kpmfmin = 1 W
advantage because the same protocol can be used over a wide range of patient sizes. It is also a disadvantage, however, in the direct comparison of exercise test results from different individuals. The Bruce protocol can be performed by small children whose legs may be too short to reach the pedals of a cycle ergometer and, because the upper torso is in constant motion, Bruce and other treadmill protocols are useful in the evaluation of motionsensitive, rate-responsive pacemakers. The James protocol6 is the major incremental cycle ergometer protocol in use today. Like the Bruce protocol, workloads are increased every 3 minutes (Table 2). Unlike the Bruce protocol, the workload depends on the resistance to ergometer pedaling and not on body weight. Because of this, the protocol is individualized according to body surface area. This size correction may not truly reflect the patient’s exercise ability, resulting in workloads that may be too strenuous for some children. Another limitation of the James protocol is that, despite patient size, the first workload is always 200 kilopond meters per minute (kpm/min). In my experience, children with significant limitations of exercise ability (e.g., many patients following Fontan operations) reach an anaerobic state before this workload. Furthermore, children who weigh less than 20 kg may have difficulty starting and maintaining the appropriate pedaling frequency at this initial workload. Similar to the Bruce protocol, data collected from patients performing a James protocol are limited to a small number of discrete workloads. With the availability of cycle ergometers with computerized controls and near-zero initial workloads, ramp protocols have become popular.7,8 In a ramp protocol, the workload starts at 0 watts
and is increased continuously at a gradual rate imperceptible to the patient. This allows for the assessment of physiologic responses over a wide range of workloads. Without prior knowledge of the child’s ability, the ramp protocol can encompass everything from low workloads necessary to assess patients with moderate or severe exercise intolerance to high workloads necessary for athletic children. Using a ramp protocol in over 700 patients in my laboratory,8 the mean exercise time was 9 minutes with 95% of the patients exercising between 3 l/2 and 14 l/2 minutes. Anaerobic
Capability
The Wingate protocol is the principal means of assessing anaerobic or peak power. It consists of pedaling at maximal speed for 30 seconds against a constant force. This force is individualized for the patient and is chosen to yield supramaximal power with noticeable fatigue.9*10 Peak power, mean power, and the rate of fatigue as a percentage of peak power are used to assess the results.
EQUIPMENT The first requirement for an exercise laboratory is an adequate facility. The American Heart Association Council on Cardiovascular Disease in the Young has published standards for exercise testing.” These state that an exercise laboratory should be at least 250 sq ft in size with a temperature of 22 oC and a relative humidity of no more than 50 % . In practice, 500 to 700 net sq ft is more practical in a modern exercise laboratory with multiple ergometers, analyzers, and computers.
6
Progress in Pediatric Cardiology
Ergometers
Two kinds of ergometers are used in exercise testing, the treadmill and the cycle ergometer. Treadmills are easy to calibrate and they are appropriate for studying patients over a wide range of ages and sizes, because the work performed is related to body weight. Because more muscle groups are involved in the exercise, treadmills result in approximately 5% to 10% higher peak oxygen consumption. Treadmills are noisier than cycle ergometers and they require higher ceiling clearance and greater floor space. Treadmills also pose a slightly greater risk of patient injury from falls, and it is recommended that the area around them be padded. Treadmill workloads cannot be accurately measured because they depend on patient weight and efficiency of gait. This means that exercise efficiency cannot be calculated and this places a greater emphasis on measuring oxygen uptake to interpret the results. Finally, there is more upper torso motion with a treadmill than a cycle ergometer. This motion, although essential for studies of patients with motion-sensitive, rate-responsive pacemakers, creates difficulty with other studies such as exercise echocardiography or nuclear cardiography. There are several advantages to cycle ergometers. They are smaller and quieter than treadmills, and they can be used in either sitting or supine positions. They have a very fast response time to changing workloads. Workload depends on pedaling frequency and resistance to pedaling (torque); thus, it can be precisely determined and exercise efficiency can be calculated from this determination. Because the patient is supported-by the seat, there is less risk of patient injury. Physiologic measurements are made easier because of the stability of the upper torso. Patients with amputation of a lower limb can be tested using toe clips. There are several disadvantages to cycle ergometers. Handle bars, seats, and pedals must be adapted for a wide range of patient sizes. For greatest efficiency, the seat and pedals must be adjusted so that there is a 5% flexion at the knee when the leg is fully extended. Most cycle ergometers are not suitable for children younger than 8 years of age without special adaptations, and even with them, cycle ergometers rarely are suitable for children younger than 5 years of age. Due to the inertia of the
flywheels, most cycle ergometers have a minimum workload of 20 to 30 W. This load is too high for small children and for children with significant exercise limitations. The minimal workload of ergometers can be reduced to nearly 0 W by adding a motor to overcome flywheel inertia. However, this adds to the cost and complexity of the ergometer and currently is available only on one commercial system in the United States. Cycle ergometers come in two varieties. The simplest is the mechanically braked ergometer with a preset resistance to pedaling. To maintain a constant work rate, the individual must pedal at a constant frequency. Young children are often incapable of maintaining this constant pedaling frequency. In electronically braked cycle ergometers a feedback loop changes with the resistance to pedaling as pedaling frequency varies. These ergometers are more appropriate for young children because they maintain a constant work rate over a wide and variable range of pedaling frequencies. This feedback mechanism, however, adds to the cost of the ergometer and makes it more difficult to calibrate. Because neither a treadmill nor a cycle ergometer is appropriate for all exercise tests, an exercise laboratory ideally should have both instruments. Then the selection of the testing device can be based on specific patient characteristics. Respiratory
Gas Analysis
Assessments of oxygen uptake and carbon dioxide production require measurement of respiratory gas flow and concentration. Devices to measure instantaneous gas flow come in two basic types, pneumotachographs and turbine transducers. Pneumotachographs function by measuring the pressure difference across either a series of tubes or a wire mesh. To perform accurately, the flow through the device must be laminar; that is, the velocity of air flow must be low as compared with the cross-sectional area of the instrument. Furthermore, in the expiratory side of the circuit, the pneumotachograph should be heated to prevent condensation as expiratory gas cools. This heating increases the gas volume, which in the final analysis requires a volume correction. The viscosity of gas also affects the pressure-flow relationships. Oxygen is significantly more viscous than nitrogen. Ideally
Exercise
the flow meter should be calibrated with a gas similar in composition to the one to be analyzed, with subsequent viscosity corrections applied. Turbine transducers function like windmills. Air flow causes a turbine to rotate and the number of revolutions per unit of time is directly proportional to the flow rate. They have several advantages over pneumotachographs. Their response is linear over a wider range of flow rates, and they are not sensitive to changes in gas viscosity. They can function in a bi-directional mode allowing assessment of both inspiratory and expiratory flows with one flow meter. However, the turbine in the transducer is sensitive to water and saliva. Consequently, a saliva trap should be placed between the patient’s mouth and the transducer. Construction of the transducer is critical. If the turbine has a high sensitivity to low-flow states (i.e., a low resistance to spinning), inertia will cause it to continue spinning at the end of the respiratory phase. If resistance to spinning is increased to overcome this endrespiratory over-spin, the transducer will not accurately measure low-flow states. Therefore, before patient use, measurements of each turbine transducer should be validated across a range of physiologic flow states. Because flow meters can be sensitive to temperature, water vapor, and viscosity, as well as hose and connector geometry, they always should be calibrated before a test under the same conditions to be used during testing. Gas concentrations are usually measured by either discrete oxygen and carbon dioxide analyzers or mass spectrometry. The spectrometer operates with greater speed and accuracy and can measure gases other than oxygen and carbon dioxide, albeit at higher costs. With either unit, sampling can occur distally at a mixing chamber or proximally close to the patient’s mouth. Mixing chambers are used to provide average gas concentrations over the full expiratory cycle. Because of this averaging, mixing chambers do not require rapid analyzers, but they do require time for gases to reach a new equilibrium when expiratory gas concentration changes. This equilibrium period is related to the ratio of the volume of the mixing chamber and minute ventilation. If the mixing chamber is large compared with minute ventilation, the response time may be several minutes. If
Testing
7
the mixing chamber is small compared with minute ventilation, adequate mixing will not occur, and gas sampling will be inaccurate. Consequently, mixing chambers should not be used to track changes occurring more frequently than every 3 minutes. Although this may be adequate for assessing the patient at the end of each stage of a 3-minute incremental protocol, mixing chambers are not accurate for determining early responses to step-wise changes in workload or during continuously changing workloads such as with the ramp protocol. The main advantage of mixing chambers is simplicity. Because they generate a small amount of average data, strip-chart recorders can be used to record the measurements with rapid hand calculations of oxygen uptake and carbon dioxide production. Continuous gas sampling near the patient’s mouth allows for breath-by-breath calculation of oxygen uptake and carbon dioxide production. This requires very rapid analyzers, analog-todigital boards, and computers. Previous work by Wessel et a1.12 has shown that effective sampling rates must be at least 125 times per second for accurate breath-by-breath measurements. Because functional residual capacity and alveolar gas concentrations vary with each breath, breath-bybreath systems should correct for these changes.13 Secondary to the overhead of measuring both inspiratory and expiratory flows and gas concentrations at appropriate frequencies, in practice this is rarely done by commercially available metabolic carts. When a breath-by-breath system is appropriately constructed, it is invaluable for assessing the rapid metabolic changes that occur at the onset of a change in workload during incremental protocols and during nonsteady-state ramp protocols. Furthermore, this system enables accurate determinations of the anaerobic threshold by the V Slope technique. l4 When inappropriately constructed, individual breath data can still be averaged together to yield results similar to those obtained by a mixing chamber. In addition to speed and accuracy, as a gas analyzer, the mass spectrometer has significant advantages over discrete analyzers. One major advantage is the ability to determine cardiac output during exercise by analyzing gases such as acetylene and helium that are not affected by changes in metabolic
stat~s.‘~ Another advantage is the ability to determine pulmonary diffusion capacity by using the heavy isotope of carbon monoxide.” ~atever system is used to measure gas flows and concentrations, the exercise laboratory will require assorted connecting valves, tubes, and hoses, These devices separate inspiratory and expiratory gas samples. They should be smooth walled and of sufficient size to prevent turbulence or other sources of interference with airflow, and they must have a low deadspace to prevent a signi~cant wasted respiratoryeffort from rebreathing exhaled gases. Electrocurdiogmphic Sysfems The exercise electrocardiographic (ECG) system should display at least three leads continuously: an anterior-posterior lead, a right-left lead, and a superior-inferior lead. it should also have the capability for intermittent display of a fuff 12-lead electrocardiogram and for full disclosure of the ECG data. This can be accomplished by a continuous ECG recording at a slow paper speed of ti mm/s or by storing all the ECG data in a computer format for subsequent display and printing. If the system is based on an oscilloscope display, there should be sufficient delay between the screen and the hard copy recorder to allow observed abno~ali~~es to be printed. Other Equipmenf A fully equipped exercise laboratory requires several other pieces of equipment. A pulse oximeter should be available to record arterial oxygen saturation at rest and ~ro~hout exercise. h is important, however, to consider that because they are motion sensitive, pulse oximeters may have difficulty tracking the signal at peak exercise. A good at rest pulse signal is essential if the oximeter is to function t~o~hout exercise. In my experience, an ear probe is best. Finger probes tend not to work when the subject grabs the support bar of a treadmill or the handle bars of a cycle ergometer, A means of recording blood pressure at rest and during exercise is essential. During exercise testing, a standard manometer and stethoscope are di~cult to use because of ambient laboratory noise, Computerized systems are motion sensitive and may malfunction either during exercise or with exerciseinduced arrhythmias. in my experience, a stetho-
phonic microphone taped over the brachial artery used with a standard mercury manometer works best. The laboratory should always have a variety of blood pressure cuffs of appropriate sizes for the entire range of patients to be tested, Although emergencies are rare in pediatric exercise testing laboratories, personnel and equipment should be available to handle any problem provoked by exercise. This equipment includes a cardiac defibrillator and a fully stocked crash cart. Furth~rmore~ the laboratory should have readily available medicine to treat exercise-induced asthma. A treatment table should be available for resuscitation or in case the patient feels dizzy and needs to lie down following exercise. The table is also used to record a standard supine ECG before exercise, A simple spirometer should be available to evaluate respiratory status before and after exercise. At a minimum, it should be able to measure flowvolume loops and maximal voluntary ventilation. Computers, printers, and plotters simplify data collection, analysis, and display. The computer should have a sufficient speed and hard disk space for appropriate data processing. In my experience, a 25-M&, 386 computer with at least 100 megabytes of hard disk storage is the mini~~rn required for continuous on-line breath-by-hre~th data analys& Similarly, the analog-t~~gital board used to process the data should have at least 12-bit accuracy and be fast enough to handle 8-ms sampling rates across the full range of recorded channels. The laboratory should have a barometer, humidistat, and thermometer to determine environmental conditions and devices to cahbrate or check the ECG systems ergometers, flow meters, and gas analyzers,
NORMAL VALUES Normal values for children have been published using the Bruce, James, and ramp protocols. 3ecause thee studies deal with a variety of measurements in children of different ages, sires, and gender, their data are only summarized here. Cummi~ published normal exercise times for 327 children, initially referred for evaluation of innocent heart murmurs. I7 These data are summarized in Table 3. Arenas supplemented these data with data from 100 normal lvfexican children with
Exercise Testing
9
TABLE 4. Ramp Protocol Normal Values*
TABLE 3. Bruce Protocol Normal Values*
Age (yrs)
Peak Oxygen Uptake (ml/kg/min)
Ventilatory Anaerobic Threshold (ml/kg/min)
d13 >13 Gil >ll
42 f 6 50 +- 8 38 + 7 34 + 4
26 + 27 + 23 + 19 k
Exercise Times (min)
(Y)
(mean f SD)
Females (mean f
4-5 6-7 8-9 10-12 13-15 16-18
10.4zk 1.9 11.8f 1.6 12.6k 2.3 12.7zk 1.9 14.1f 1.7 13.5f 1.4
9.5f 11.2f 11.8+ 12.3f 11.1+ 10.7f
Males
Age
SD)
1.8 1.5 1.6 1.4 1.3 1.4
Males Females
5 6 4 3
*FromCopperetal.7
*FromCummingeta1.17
a mean age of 10 years. l* The mean exercise test duration was 11.8 f 1.2 minutes in males and lo.7 -t 1.2 minutes in females, and the mean peak oxygen uptake was 45.2 + 4.9 ml/kg/min in males and 41.9 f 4.5 ml/kg/min in females. In 1980, James published normal data from his protocol derived from studies in 149 children with a mean age of 14 l/2 years and presented in 12 subgroups based on gender, body surface area, and height.6 Washington supplemented these normative data with measurements in 151 Denver children, 7 to 13 years of age, presented in six subgroups based on gender and body surface area.19 Calzolari subsequently published normative data for the James protocol in 102 Italian childrenZo and found no significant differences from the James data.6 In 1984, Cooper published results of the ramp protocol in 109 children.7 His data for peak oxygen uptake and anaerobic threshold are summarized in Table 4. These data are difficult to reproduce because the ramp slope was varied according to estimates of each child’s exercise capacity. In 1991, Tanner attempted to standardize the ramp protocol for children, using a constant work load of 0.25 W/kg/min.B Results were compared to tests of the James protocol in the same children. The two exercise tests were performed in a random order on separate days. Normal exercise times were 9.3 k 3.3 minutes with the ramp protocol, compared to 6.2
+ 2.9 minutes
with the James
protocol
(P <
.05). In her population of normal children, Tanner found no difference between the two protocols in measured peak oxygen uptake, peak carbon dioxide production, or peak cardiac index. However, a subsequent study in children with significant limi-
tations of exercise tolerance revealed significant differences between the two protocols, with the ramp protocol preferable.’ Normal values for the Wingate anaerobic test have been published by Inbar et a1.21 Anaerobic power corrected for body size is correlated with age until approximately the third decade of life. This is distinctly different from aerobic power corrected for size, which does not change significantly from early childhood to young adulthood.
SUMMARY Exercise testing is a complicated discipline. It requires a thorough understanding of human physiology, multiple protocols, and a wide variety of equipment. As subsequent articles will demonstrate, when this understanding has been achieved, exercise physiology can contribute significantly to the management of many cardiovascular problems of children and adolescents.
REFERENCES Larsen RL, Barber G, Heise CT, August CS. Exercise assessment of cardiac function in children and young adults before and after bone marrow transplantation. Pediatrics. 1992;89:722-729. Murphy JK, Alpert BS, Christman JV, Willey ES. Physical fitness in children: a survey method based on parental report. Am 1 Public Health. 1988;78:
708-710. Barber G, Heise CT. Subjective estimates of exercise ability: comparison to objective measurements. Pediatr Exer Sci. 1991;3:327-332. Clinical competence in exercise testing. A statement for physicians from the ACP/ACC/AHA Task Force on Clinical Privileges in Cardiology. Circulation. 1990;82:1884-1888.
10
Progress in Pediatric
5. Bruce RA, McDonough JR. Stress testing in screening for cardiovascular disease. Bull NY Acad Med. 1969;45:1288-1305. 6. James FW, Kaplan S, Glueck C, Tsay JY, Knight MJS, Surwar CJ. Responses of normal children and young adults to controlled bicycle exercise. Circulation. 1980;61:902-912. 7. Cooper DM, Weiler Ravel D, Whipp BJ, Wasserman K. Aerobic parameters of exercise as a function of body size during growth in children. JAppl Physiol. 1984;56:628-634. 8. Tanner CS, Heise CT, Barber G. Correlation of the physiologic parameters of a continuous ramp versus an incremental James exercise protocol in normal children. Am J Cardiol. 1991;67:309-312. 9. Dotan R, Bar-Or 0. Load optimization for the Wingate anaerobic test. Eur J Appl Physiol. 1983;51: 409-417. 10. Bar-Or 0. The Wingate anaerobic test. An update on methodology, reliability and validity. Sports Med. 1987;4:381-394. 11. James FW, Blomqvist CG, Freed MD, et al: Standards for exercise testing in the pediatric age group. American Heart Association Council on Cardiovascular Disease in the Young. Ad hoc committee on exercise testing. Circulation. 1982;66:1377A1397A. 12.Wessel HU, Stout RL, Paul MH. Minicomputerbased system for breath-by-breath analysis of ventilation and pulmonary gas exchange. In: Ripley KL, Ostrow HG, editors. Computers in Cardiology; New York, IEEE Computer Society; 1978. p. 97104. 13. Wessel HU, Stout RL, Bastanier CK, Paul MH. Breath-by-breath variation of FRC: effect of VOz
Cardiology
and VC02 measured at the mouth. J Appl Physiol. 1979;46:1122-1126. 14. Beaver WL, Wasserman K, Whipp BJ. A new method for detecting anaerobic threshold by gas exchange. J Appl Physiol. 1986;60:2020-2027. 15. Triebwasse JH, Johnson RL, Burpo RP, et al. Noninvasive determination of cardiac output by a modified acetylene-helium rebreathing procedure utilizing mass spectrometer measurements. Aviat Space Environ Med. 1977;48:203-209. 16. Sackner MA, Greeneltch D, Heiman MS, Epstein S, Atkins N. Diffusion capacity, membrane diffusion capacity, capillary blood volume, pulmonary tissue volume and cardiac output measured by a rebreathing technique. Am Rev Respir Dis. 1975;111:157165. 17. Cumming GR, Everett D, Hastman L. Bruce treadmill test in children: normal values in a clinic population. Am J Cardiol. 1978;41:69-75. 18. Arenas Leon JL, Zajarias A, Femandez de la Vega P, Medrano G, Buendia A, Attie F. Response of normal children to the treadmill exercise test using the Bruce protocol. Arch lnst Cardiol Mex. 1985;55:227-233. 19. Washington RL, van Gundy JC, Cohen C, Sondheimer HM, Wolfe RR. Normal aerobic and anaerobic exercise data for North American school-age children. J Pediatr. 1988;112:223-233. 20. Calzolari A, Di Ciommo V, Drago F, et al. Cycloergometric exercise test in normal children: comparison of an Italian and a North American population. G Ital Cardiol. 1990;20:323-328. in 21. Inbar 0, Bar-Or 0. Anaerobic characteristics male children and adolescents. Med Sci Sports her. 1986;18:264-269.
Equipment, and Normal Values
GERALD
BARBER,
M.D.
De~~~~e~t of Pediatrics Nau York University School of Medicine New York, New York
Exercise testing is an important tool in the evaluation of cardiac and respiratory problems as well as physical working capacity. By taxing the cardiopulmonary system, exercise determines the patient‘s physiologic reserve, enabling the early detection of abnormalities not apparent at rest1 Testing quantifies a patient’s exercise ability in a way that questionnaires or other subjective estimates cannot .2,3 Exercise testing requires a thorough understanding of methodology, equipment, and normal values. This article will provide a basic review of these areas, but an understanding sufficient to perform exercise testing can only come with supervised experience in an exercise laboratory. Recently the American College of Physicians, the American College of Cardiology, and the American Heart Association published a task force statement on clinical competence in exercise testing.* They recommend that an individual perform at least 50 exercise tests in a supervised training setting and at least 25 per year thereafter to acquire and maintain minimum clinical competence. These recommendations, based mainly on testing in adults with coronary artery disease, may underestimate the requirements for clinical competence in pediatric exercise testing.
Address correspondence to Gerald Barber, M.D., Pediatric Cardiology, NYLJ School of Medicine, l-i-501, 560 First Avenue, New York, NY 10016-6402.
METHODOLOGY The protocol used for exercise testing depends largely on the purpose of the test. Further consideration must be paid to the age, size, and estimated exercise ability of the patient. One of the first issues to be decided is whether one wants to assess aerobic or anaerobic capability. Aerobic capability usually is measured by indices such as peak oxygen uptake and anaerobic threshold, whereas anaerobic capability usually is measured by indices such as peak power and fatigue rate. Anaerobic Capability Protocols to measure aerobic power (physical working capacity) are useful in assessing cardiac and pulmonary limitations as well as total exercise performance. These protocols can be divided by the exercise instrument (treadmill or cycle ergometer) and by the way in which workloads are changed (incremental or ramp protocols). The Bruce protocol is the principal treadmill protocol used today.5 It is an incremental protocol in which both speed and elevation are increased every 3 minutes until the individual can go no further (Table 1). As with all incremental protocols, it is limited in that measurements are made at only a certain number of preset workloads, independent of the subject’s exercise ability: therefore, the information obtained is inversely related to the patient’s degree of limitation. Also, in the Bruce protocol the work performed is dependent on the weight of the patient and the efficiency of gait. This is an Prog Pediatr Cardiof 1993; 2(2):4-10 Copyright 0 1993 by Andover Medical
Exercise Testing
TABLE
1. Bruce Exercise Protocol
5
TABLE 2. James Exercise Protocol
Stage
Speed (miles/h)
Elevation (%)
1 2 3 4 5 6 7
1.7 2.5 3.4 4.2 5.0 5.5 6.0
10 12 14 16 18 20 22
Body Surface Area
1.0-1.19
Stage
(kpm/min)
(kpm/min)
1 2 3 4 5 6 7
200 300 500 600 700 800 900
200 400 600 700 800 900 1000
200 500 800 1000 1200 1400 1600
m2
al.2
m2
l6.12kpmfmin = 1 W
advantage because the same protocol can be used over a wide range of patient sizes. It is also a disadvantage, however, in the direct comparison of exercise test results from different individuals. The Bruce protocol can be performed by small children whose legs may be too short to reach the pedals of a cycle ergometer and, because the upper torso is in constant motion, Bruce and other treadmill protocols are useful in the evaluation of motionsensitive, rate-responsive pacemakers. The James protocol6 is the major incremental cycle ergometer protocol in use today. Like the Bruce protocol, workloads are increased every 3 minutes (Table 2). Unlike the Bruce protocol, the workload depends on the resistance to ergometer pedaling and not on body weight. Because of this, the protocol is individualized according to body surface area. This size correction may not truly reflect the patient’s exercise ability, resulting in workloads that may be too strenuous for some children. Another limitation of the James protocol is that, despite patient size, the first workload is always 200 kilopond meters per minute (kpm/min). In my experience, children with significant limitations of exercise ability (e.g., many patients following Fontan operations) reach an anaerobic state before this workload. Furthermore, children who weigh less than 20 kg may have difficulty starting and maintaining the appropriate pedaling frequency at this initial workload. Similar to the Bruce protocol, data collected from patients performing a James protocol are limited to a small number of discrete workloads. With the availability of cycle ergometers with computerized controls and near-zero initial workloads, ramp protocols have become popular.7,8 In a ramp protocol, the workload starts at 0 watts
and is increased continuously at a gradual rate imperceptible to the patient. This allows for the assessment of physiologic responses over a wide range of workloads. Without prior knowledge of the child’s ability, the ramp protocol can encompass everything from low workloads necessary to assess patients with moderate or severe exercise intolerance to high workloads necessary for athletic children. Using a ramp protocol in over 700 patients in my laboratory,8 the mean exercise time was 9 minutes with 95% of the patients exercising between 3 l/2 and 14 l/2 minutes. Anaerobic
Capability
The Wingate protocol is the principal means of assessing anaerobic or peak power. It consists of pedaling at maximal speed for 30 seconds against a constant force. This force is individualized for the patient and is chosen to yield supramaximal power with noticeable fatigue.9*10 Peak power, mean power, and the rate of fatigue as a percentage of peak power are used to assess the results.
EQUIPMENT The first requirement for an exercise laboratory is an adequate facility. The American Heart Association Council on Cardiovascular Disease in the Young has published standards for exercise testing.” These state that an exercise laboratory should be at least 250 sq ft in size with a temperature of 22 oC and a relative humidity of no more than 50 % . In practice, 500 to 700 net sq ft is more practical in a modern exercise laboratory with multiple ergometers, analyzers, and computers.
6
Progress in Pediatric Cardiology
Ergometers
Two kinds of ergometers are used in exercise testing, the treadmill and the cycle ergometer. Treadmills are easy to calibrate and they are appropriate for studying patients over a wide range of ages and sizes, because the work performed is related to body weight. Because more muscle groups are involved in the exercise, treadmills result in approximately 5% to 10% higher peak oxygen consumption. Treadmills are noisier than cycle ergometers and they require higher ceiling clearance and greater floor space. Treadmills also pose a slightly greater risk of patient injury from falls, and it is recommended that the area around them be padded. Treadmill workloads cannot be accurately measured because they depend on patient weight and efficiency of gait. This means that exercise efficiency cannot be calculated and this places a greater emphasis on measuring oxygen uptake to interpret the results. Finally, there is more upper torso motion with a treadmill than a cycle ergometer. This motion, although essential for studies of patients with motion-sensitive, rate-responsive pacemakers, creates difficulty with other studies such as exercise echocardiography or nuclear cardiography. There are several advantages to cycle ergometers. They are smaller and quieter than treadmills, and they can be used in either sitting or supine positions. They have a very fast response time to changing workloads. Workload depends on pedaling frequency and resistance to pedaling (torque); thus, it can be precisely determined and exercise efficiency can be calculated from this determination. Because the patient is supported-by the seat, there is less risk of patient injury. Physiologic measurements are made easier because of the stability of the upper torso. Patients with amputation of a lower limb can be tested using toe clips. There are several disadvantages to cycle ergometers. Handle bars, seats, and pedals must be adapted for a wide range of patient sizes. For greatest efficiency, the seat and pedals must be adjusted so that there is a 5% flexion at the knee when the leg is fully extended. Most cycle ergometers are not suitable for children younger than 8 years of age without special adaptations, and even with them, cycle ergometers rarely are suitable for children younger than 5 years of age. Due to the inertia of the
flywheels, most cycle ergometers have a minimum workload of 20 to 30 W. This load is too high for small children and for children with significant exercise limitations. The minimal workload of ergometers can be reduced to nearly 0 W by adding a motor to overcome flywheel inertia. However, this adds to the cost and complexity of the ergometer and currently is available only on one commercial system in the United States. Cycle ergometers come in two varieties. The simplest is the mechanically braked ergometer with a preset resistance to pedaling. To maintain a constant work rate, the individual must pedal at a constant frequency. Young children are often incapable of maintaining this constant pedaling frequency. In electronically braked cycle ergometers a feedback loop changes with the resistance to pedaling as pedaling frequency varies. These ergometers are more appropriate for young children because they maintain a constant work rate over a wide and variable range of pedaling frequencies. This feedback mechanism, however, adds to the cost of the ergometer and makes it more difficult to calibrate. Because neither a treadmill nor a cycle ergometer is appropriate for all exercise tests, an exercise laboratory ideally should have both instruments. Then the selection of the testing device can be based on specific patient characteristics. Respiratory
Gas Analysis
Assessments of oxygen uptake and carbon dioxide production require measurement of respiratory gas flow and concentration. Devices to measure instantaneous gas flow come in two basic types, pneumotachographs and turbine transducers. Pneumotachographs function by measuring the pressure difference across either a series of tubes or a wire mesh. To perform accurately, the flow through the device must be laminar; that is, the velocity of air flow must be low as compared with the cross-sectional area of the instrument. Furthermore, in the expiratory side of the circuit, the pneumotachograph should be heated to prevent condensation as expiratory gas cools. This heating increases the gas volume, which in the final analysis requires a volume correction. The viscosity of gas also affects the pressure-flow relationships. Oxygen is significantly more viscous than nitrogen. Ideally
Exercise
the flow meter should be calibrated with a gas similar in composition to the one to be analyzed, with subsequent viscosity corrections applied. Turbine transducers function like windmills. Air flow causes a turbine to rotate and the number of revolutions per unit of time is directly proportional to the flow rate. They have several advantages over pneumotachographs. Their response is linear over a wider range of flow rates, and they are not sensitive to changes in gas viscosity. They can function in a bi-directional mode allowing assessment of both inspiratory and expiratory flows with one flow meter. However, the turbine in the transducer is sensitive to water and saliva. Consequently, a saliva trap should be placed between the patient’s mouth and the transducer. Construction of the transducer is critical. If the turbine has a high sensitivity to low-flow states (i.e., a low resistance to spinning), inertia will cause it to continue spinning at the end of the respiratory phase. If resistance to spinning is increased to overcome this endrespiratory over-spin, the transducer will not accurately measure low-flow states. Therefore, before patient use, measurements of each turbine transducer should be validated across a range of physiologic flow states. Because flow meters can be sensitive to temperature, water vapor, and viscosity, as well as hose and connector geometry, they always should be calibrated before a test under the same conditions to be used during testing. Gas concentrations are usually measured by either discrete oxygen and carbon dioxide analyzers or mass spectrometry. The spectrometer operates with greater speed and accuracy and can measure gases other than oxygen and carbon dioxide, albeit at higher costs. With either unit, sampling can occur distally at a mixing chamber or proximally close to the patient’s mouth. Mixing chambers are used to provide average gas concentrations over the full expiratory cycle. Because of this averaging, mixing chambers do not require rapid analyzers, but they do require time for gases to reach a new equilibrium when expiratory gas concentration changes. This equilibrium period is related to the ratio of the volume of the mixing chamber and minute ventilation. If the mixing chamber is large compared with minute ventilation, the response time may be several minutes. If
Testing
7
the mixing chamber is small compared with minute ventilation, adequate mixing will not occur, and gas sampling will be inaccurate. Consequently, mixing chambers should not be used to track changes occurring more frequently than every 3 minutes. Although this may be adequate for assessing the patient at the end of each stage of a 3-minute incremental protocol, mixing chambers are not accurate for determining early responses to step-wise changes in workload or during continuously changing workloads such as with the ramp protocol. The main advantage of mixing chambers is simplicity. Because they generate a small amount of average data, strip-chart recorders can be used to record the measurements with rapid hand calculations of oxygen uptake and carbon dioxide production. Continuous gas sampling near the patient’s mouth allows for breath-by-breath calculation of oxygen uptake and carbon dioxide production. This requires very rapid analyzers, analog-todigital boards, and computers. Previous work by Wessel et a1.12 has shown that effective sampling rates must be at least 125 times per second for accurate breath-by-breath measurements. Because functional residual capacity and alveolar gas concentrations vary with each breath, breath-bybreath systems should correct for these changes.13 Secondary to the overhead of measuring both inspiratory and expiratory flows and gas concentrations at appropriate frequencies, in practice this is rarely done by commercially available metabolic carts. When a breath-by-breath system is appropriately constructed, it is invaluable for assessing the rapid metabolic changes that occur at the onset of a change in workload during incremental protocols and during nonsteady-state ramp protocols. Furthermore, this system enables accurate determinations of the anaerobic threshold by the V Slope technique. l4 When inappropriately constructed, individual breath data can still be averaged together to yield results similar to those obtained by a mixing chamber. In addition to speed and accuracy, as a gas analyzer, the mass spectrometer has significant advantages over discrete analyzers. One major advantage is the ability to determine cardiac output during exercise by analyzing gases such as acetylene and helium that are not affected by changes in metabolic
stat~s.‘~ Another advantage is the ability to determine pulmonary diffusion capacity by using the heavy isotope of carbon monoxide.” ~atever system is used to measure gas flows and concentrations, the exercise laboratory will require assorted connecting valves, tubes, and hoses, These devices separate inspiratory and expiratory gas samples. They should be smooth walled and of sufficient size to prevent turbulence or other sources of interference with airflow, and they must have a low deadspace to prevent a signi~cant wasted respiratoryeffort from rebreathing exhaled gases. Electrocurdiogmphic Sysfems The exercise electrocardiographic (ECG) system should display at least three leads continuously: an anterior-posterior lead, a right-left lead, and a superior-inferior lead. it should also have the capability for intermittent display of a fuff 12-lead electrocardiogram and for full disclosure of the ECG data. This can be accomplished by a continuous ECG recording at a slow paper speed of ti mm/s or by storing all the ECG data in a computer format for subsequent display and printing. If the system is based on an oscilloscope display, there should be sufficient delay between the screen and the hard copy recorder to allow observed abno~ali~~es to be printed. Other Equipmenf A fully equipped exercise laboratory requires several other pieces of equipment. A pulse oximeter should be available to record arterial oxygen saturation at rest and ~ro~hout exercise. h is important, however, to consider that because they are motion sensitive, pulse oximeters may have difficulty tracking the signal at peak exercise. A good at rest pulse signal is essential if the oximeter is to function t~o~hout exercise. In my experience, an ear probe is best. Finger probes tend not to work when the subject grabs the support bar of a treadmill or the handle bars of a cycle ergometer, A means of recording blood pressure at rest and during exercise is essential. During exercise testing, a standard manometer and stethoscope are di~cult to use because of ambient laboratory noise, Computerized systems are motion sensitive and may malfunction either during exercise or with exerciseinduced arrhythmias. in my experience, a stetho-
phonic microphone taped over the brachial artery used with a standard mercury manometer works best. The laboratory should always have a variety of blood pressure cuffs of appropriate sizes for the entire range of patients to be tested, Although emergencies are rare in pediatric exercise testing laboratories, personnel and equipment should be available to handle any problem provoked by exercise. This equipment includes a cardiac defibrillator and a fully stocked crash cart. Furth~rmore~ the laboratory should have readily available medicine to treat exercise-induced asthma. A treatment table should be available for resuscitation or in case the patient feels dizzy and needs to lie down following exercise. The table is also used to record a standard supine ECG before exercise, A simple spirometer should be available to evaluate respiratory status before and after exercise. At a minimum, it should be able to measure flowvolume loops and maximal voluntary ventilation. Computers, printers, and plotters simplify data collection, analysis, and display. The computer should have a sufficient speed and hard disk space for appropriate data processing. In my experience, a 25-M&, 386 computer with at least 100 megabytes of hard disk storage is the mini~~rn required for continuous on-line breath-by-hre~th data analys& Similarly, the analog-t~~gital board used to process the data should have at least 12-bit accuracy and be fast enough to handle 8-ms sampling rates across the full range of recorded channels. The laboratory should have a barometer, humidistat, and thermometer to determine environmental conditions and devices to cahbrate or check the ECG systems ergometers, flow meters, and gas analyzers,
NORMAL VALUES Normal values for children have been published using the Bruce, James, and ramp protocols. 3ecause thee studies deal with a variety of measurements in children of different ages, sires, and gender, their data are only summarized here. Cummi~ published normal exercise times for 327 children, initially referred for evaluation of innocent heart murmurs. I7 These data are summarized in Table 3. Arenas supplemented these data with data from 100 normal lvfexican children with
Exercise Testing
9
TABLE 4. Ramp Protocol Normal Values*
TABLE 3. Bruce Protocol Normal Values*
Age (yrs)
Peak Oxygen Uptake (ml/kg/min)
Ventilatory Anaerobic Threshold (ml/kg/min)
d13 >13 Gil >ll
42 f 6 50 +- 8 38 + 7 34 + 4
26 + 27 + 23 + 19 k
Exercise Times (min)
(Y)
(mean f SD)
Females (mean f
4-5 6-7 8-9 10-12 13-15 16-18
10.4zk 1.9 11.8f 1.6 12.6k 2.3 12.7zk 1.9 14.1f 1.7 13.5f 1.4
9.5f 11.2f 11.8+ 12.3f 11.1+ 10.7f
Males
Age
SD)
1.8 1.5 1.6 1.4 1.3 1.4
Males Females
5 6 4 3
*FromCopperetal.7
*FromCummingeta1.17
a mean age of 10 years. l* The mean exercise test duration was 11.8 f 1.2 minutes in males and lo.7 -t 1.2 minutes in females, and the mean peak oxygen uptake was 45.2 + 4.9 ml/kg/min in males and 41.9 f 4.5 ml/kg/min in females. In 1980, James published normal data from his protocol derived from studies in 149 children with a mean age of 14 l/2 years and presented in 12 subgroups based on gender, body surface area, and height.6 Washington supplemented these normative data with measurements in 151 Denver children, 7 to 13 years of age, presented in six subgroups based on gender and body surface area.19 Calzolari subsequently published normative data for the James protocol in 102 Italian childrenZo and found no significant differences from the James data.6 In 1984, Cooper published results of the ramp protocol in 109 children.7 His data for peak oxygen uptake and anaerobic threshold are summarized in Table 4. These data are difficult to reproduce because the ramp slope was varied according to estimates of each child’s exercise capacity. In 1991, Tanner attempted to standardize the ramp protocol for children, using a constant work load of 0.25 W/kg/min.B Results were compared to tests of the James protocol in the same children. The two exercise tests were performed in a random order on separate days. Normal exercise times were 9.3 k 3.3 minutes with the ramp protocol, compared to 6.2
+ 2.9 minutes
with the James
protocol
(P <
.05). In her population of normal children, Tanner found no difference between the two protocols in measured peak oxygen uptake, peak carbon dioxide production, or peak cardiac index. However, a subsequent study in children with significant limi-
tations of exercise tolerance revealed significant differences between the two protocols, with the ramp protocol preferable.’ Normal values for the Wingate anaerobic test have been published by Inbar et a1.21 Anaerobic power corrected for body size is correlated with age until approximately the third decade of life. This is distinctly different from aerobic power corrected for size, which does not change significantly from early childhood to young adulthood.
SUMMARY Exercise testing is a complicated discipline. It requires a thorough understanding of human physiology, multiple protocols, and a wide variety of equipment. As subsequent articles will demonstrate, when this understanding has been achieved, exercise physiology can contribute significantly to the management of many cardiovascular problems of children and adolescents.
REFERENCES Larsen RL, Barber G, Heise CT, August CS. Exercise assessment of cardiac function in children and young adults before and after bone marrow transplantation. Pediatrics. 1992;89:722-729. Murphy JK, Alpert BS, Christman JV, Willey ES. Physical fitness in children: a survey method based on parental report. Am 1 Public Health. 1988;78:
708-710. Barber G, Heise CT. Subjective estimates of exercise ability: comparison to objective measurements. Pediatr Exer Sci. 1991;3:327-332. Clinical competence in exercise testing. A statement for physicians from the ACP/ACC/AHA Task Force on Clinical Privileges in Cardiology. Circulation. 1990;82:1884-1888.
10
Progress in Pediatric
5. Bruce RA, McDonough JR. Stress testing in screening for cardiovascular disease. Bull NY Acad Med. 1969;45:1288-1305. 6. James FW, Kaplan S, Glueck C, Tsay JY, Knight MJS, Surwar CJ. Responses of normal children and young adults to controlled bicycle exercise. Circulation. 1980;61:902-912. 7. Cooper DM, Weiler Ravel D, Whipp BJ, Wasserman K. Aerobic parameters of exercise as a function of body size during growth in children. JAppl Physiol. 1984;56:628-634. 8. Tanner CS, Heise CT, Barber G. Correlation of the physiologic parameters of a continuous ramp versus an incremental James exercise protocol in normal children. Am J Cardiol. 1991;67:309-312. 9. Dotan R, Bar-Or 0. Load optimization for the Wingate anaerobic test. Eur J Appl Physiol. 1983;51: 409-417. 10. Bar-Or 0. The Wingate anaerobic test. An update on methodology, reliability and validity. Sports Med. 1987;4:381-394. 11. James FW, Blomqvist CG, Freed MD, et al: Standards for exercise testing in the pediatric age group. American Heart Association Council on Cardiovascular Disease in the Young. Ad hoc committee on exercise testing. Circulation. 1982;66:1377A1397A. 12.Wessel HU, Stout RL, Paul MH. Minicomputerbased system for breath-by-breath analysis of ventilation and pulmonary gas exchange. In: Ripley KL, Ostrow HG, editors. Computers in Cardiology; New York, IEEE Computer Society; 1978. p. 97104. 13. Wessel HU, Stout RL, Bastanier CK, Paul MH. Breath-by-breath variation of FRC: effect of VOz
Cardiology
and VC02 measured at the mouth. J Appl Physiol. 1979;46:1122-1126. 14. Beaver WL, Wasserman K, Whipp BJ. A new method for detecting anaerobic threshold by gas exchange. J Appl Physiol. 1986;60:2020-2027. 15. Triebwasse JH, Johnson RL, Burpo RP, et al. Noninvasive determination of cardiac output by a modified acetylene-helium rebreathing procedure utilizing mass spectrometer measurements. Aviat Space Environ Med. 1977;48:203-209. 16. Sackner MA, Greeneltch D, Heiman MS, Epstein S, Atkins N. Diffusion capacity, membrane diffusion capacity, capillary blood volume, pulmonary tissue volume and cardiac output measured by a rebreathing technique. Am Rev Respir Dis. 1975;111:157165. 17. Cumming GR, Everett D, Hastman L. Bruce treadmill test in children: normal values in a clinic population. Am J Cardiol. 1978;41:69-75. 18. Arenas Leon JL, Zajarias A, Femandez de la Vega P, Medrano G, Buendia A, Attie F. Response of normal children to the treadmill exercise test using the Bruce protocol. Arch lnst Cardiol Mex. 1985;55:227-233. 19. Washington RL, van Gundy JC, Cohen C, Sondheimer HM, Wolfe RR. Normal aerobic and anaerobic exercise data for North American school-age children. J Pediatr. 1988;112:223-233. 20. Calzolari A, Di Ciommo V, Drago F, et al. Cycloergometric exercise test in normal children: comparison of an Italian and a North American population. G Ital Cardiol. 1990;20:323-328. in 21. Inbar 0, Bar-Or 0. Anaerobic characteristics male children and adolescents. Med Sci Sports her. 1986;18:264-269.