Clinical Exercise Testing, Idelle M. Weisman,R. Jorge Zeballos, vol 32

63 Wu MT, Chang JM, Chiang AA, Lu JY, Hsu 70 Landreneau RJ, Sugarbaker DJ, Mack MJ, Ha- 76 Olsen GN, Bolton JW, Weiman DS, Hornung HK, Hsu WH, Yang CF: Use of quantitative zelrigg SR, Luketich JD, Fetterman L, Liptay CA: Stair climbing as an exercise test to predict CT to predict postoperative lung function in MJ, Bartley S, Boley TM, Keenan RJ, Ferson the postoperative complications of lung resec- patients with lung cancer. Radiology 1994;191: PF, Weyant RJ, Naunheim KS: Wedge resec- tion. Two years’ experience. Chest 1991;99: 257–262. tion versus lobectomy for stage I (T1 N0 M0) 587–590. non-small-cell lung cancer. J Thorac Cardio- 64 Bolliger CT, Guckel C, Engel H, Stoehr S, Wys- vasc Surg 1997;113:691–698; discussion 698– 77 Bolton JW, Weiman DS, Haynes JL, Hornung er C, Habicht J, Jordan P, Tamm M, Soler M, 700. CA, Olsen GN, Almond CH: Stair climbing as Perruchoud AP: Predicted postoperative func- an indicator of pulmonary function. Chest tion after lung resection: Comparison of quan- 71 Ginsberg RJ, Rubinstein LV: Randomized 1987;92:783–788. titative CT, scintigraphy, and anatomical cal- trial of lobectomy versus limited resection for culations. AJRCCM 2000;161:A269. T1 N0 non-small cell lung cancer. Lung Cancer 78 Richter Larsen K, Svendsen UG, Milman N, Study Group. Ann Thorac Surg 1995;60:615– Brenoe J, Petersen BN: Exercise testing in the 65 Epstein SK, Faling LJ, Daly BD, Celli BR: Pre- 622. preoperative evaluation of patients with bron- dicting complications after pulmonary resec- chogenic carcinoma. Eur Respir J 1997;10: tion: Preoperative exercise testing vs a multi- 72 Bolliger CT: Pre-operative assessment of the 1559–1565. factorial cardiopulmonary risk index. Chest lung cancer patient. SAMJ 2001;91:120–123. 1993;104:694–700. 79 Ribas J, Diaz O, Barbera JA, Mateu M, Canalis 73 Van Nostrand D, Kjelsberg MO, Humphrey E, Jover L, Roca J, Rodriguez-Roisin R: Inva- 66 Melendez JA, Carlon VA: Cardiopulmonary EW: Preresectional evaluation of risk from sive exercise testing in the evaluation of pa- risk index does not predict complications after pneumonectomy. Surg Gynecol Obstet 1968; tients at high-risk for lung resection. Eur Respir thoracic surgery. Chest 1998;114:69–75. 127:306–312. J 1998;12:1429–1435. 67 Cooper JD, Trulock EP, Triantafillou AN, Pat- 74 Berggren H, Ekroth R, Malmberg R, Naucler J, 80 Wang JS, Abboud RT, Evans KG, Finley RJ, terson GA, Pohl MS, Deloney PA, Sundaresan William-Olsson G: Hospital mortality and Graham BL: Role of CO diffusing capacity dur- RS, Roper CL: Bilateral pneumectomy (vol- long-term survival in relation to preoperative ing exercise in the preoperative evaluation for ume reduction) for chronic obstructive pulmo- function in elderly patients with bronchogenic lung resection. Am J Respir Crit Care Med nary disease. J Thorac Cardiovasc Surg 1995; carcinoma. Ann Thorac Surg 1984;38:633– 2000;162:1435–1444. 109:106–116. 636. 81 Reilly JJ Jr: Evidence-based preoperative eval- 68 McKenna RJ Jr, Fischel RJ, Brenner M, Gelb 75 Bagg LR: The 12-min walking distance; its use uation of candidates for thoracotomy. Chest AF: Combined operations for lung volume re- in the pre-operative assessment of patients with 1999;116:474S-476S. duction surgery and lung cancer. Chest 1996; bronchial carcinoma before lung resection. 110:885–888. Respiration 1984;46:342–345. Andreas H. Diacon, MD Department of Internal Medicine 69 DeRose JJ Jr, Argenziano M, El-Amir N, Jellen Tygerberg Hospital, University of Stellenbosch PA, Gorenstein LA, Steinglass KM, Thoma- PO Box 19063, 7505 Tygerberg (RSA) show B, Ginsburg ME: Lung reduction opera- Tel. +27 21 938 9556, Fax +27 21 931 7442 tion and resection of pulmonary nodules in E-Mail [email protected] patients with severe emphysema. Ann Thorac Surg 1998;65:314–318. Preoperative Exercise Testing in Lung 241 Resection Candidates

Weisman IM, Zeballos RJ (eds): Clinical Exercise Testing. Prog Respir Res. Basel, Karger, 2002, vol 32, pp 242–253 OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO The Role of Cardiopulmonary Exercise Testing for Patients with Suspected Metabolic Myopathies and Other Neuromuscular Disorders Kevin R. Flahertya Idelle M. Weismanb R. Jorge Zeballosc Fernando J. Martineza aUniversity of Michigan Health System, Department of Internal Medicine, Division of Pulmonary and Critical Care Medicine, Ann Arbor, Mich.; bHuman Performance Laboratory, Department of Clinical Investigation, Pulmonary-Critical Care Service, William Beaumont Army Medical Center, El Paso, Tex., and Department of Medicine, Pulmonary-Critical Care Division, University of Texas Health Science Center at San Antonio, San Antonio, Tex.; cHuman Performance Laboratory, Department of Clinical Investigation, William Beaumont Army Medical Center, and Department of Anesthesiology, Texas Tech University Health Sciences Center, El Paso, Tex., USA Summary Neuromuscular diseases can be associated with an ab- normal muscle response to exercise. As such, these disor- Neuromuscular diseases (NMDs) can be associated with ders may present with weakness, myalgia, abnormal fa- significant exercise limitation. As cardiopulmonary exercise tigue, or exertional breathlessness [1, 2]. These disorders testing (CPET) examines the integrated response to exercise, it can be classified as those that are associated with impaired has been investigated in multiple NMDs, proving itself particu- muscle energy metabolism, those associated with impaired larly valuable in the evaluation of metabolic disorders, which muscle bulk, and those associated with impaired control of exhibit typical patterns of exercise response. Disorders of car- muscle contraction [2]. A general enumeration of these bohydrate metabolism usually demonstrate limitations of aero- disorders is presented in table 1. As reviewed elsewhere in bic capacity and abnormalities of glycogen utilization. Lipid dis- this book, exercise requires a close integration between orders seem to be associated with lesser aerobic limitation, skeletal muscle function, cardiovascular function with as- while disorders of mitochondrial function seem to be associat- sociated oxygen delivery, and ventilatory function with ed with the greatest limitation in aerobic capacity. Further- oxygenation of blood. As such, it is not surprising that more, a hyperventilatory and hypercirculatory pattern seems myopathic disorders can significantly impact exercise ca- to typify exercise response in these latter patients. As such, pacity. This chapter reviews the pathophysiologic implica- CPET can be used in the initial evaluation of patients with sus- tions of myopathic disorders on exercise capacity and pected carbohydrate or lipid metabolism as well as those with reviews clinical indications of exercise testing in the diag- defects of mitochondrial function. nosis and management of these disorders. Disorders of Muscle Energy Metabolism The opinions or assertions contained herein are the private views of Metabolic myopathies are disorders of muscle energy the authors and are not to be construed as official or as reflecting the metabolism. These disorders have been grouped into views of the Department of the Army or the Department of De- three broad categories: (1) defective carbohydrate utiliza- fense. tion, (2) abnormal lipid utilization, and (3) mitochondrial

Table1. General classification of muscle disorders ated with the accumulation of metabolic end products [5]. Figure 1 demonstrates the typical sequence of events dur- Disorders of muscle energy metabolism ing the conversion of glucose to pyruvate in the cyto- Disorders of carbohydrate metabolism plasm. Pyruvate is subsequently transferred into the mito- chondria where further conversion takes place to acetyl Myophosphorylase deficiency coenzyme A (CoA) by the action of pyruvate dehydroge- Phosphorylase kinase deficiency nase (PD) [4]. Acetyl CoA enters the tricarboxylic acid Phosphofructokinase deficiency cycle (TCA) and the reduced form of nicotinamide ade- Disorders of lipid metabolism nine dinucleotide (NADH) is generated. The enzymatic Fatty acid oxidation defects activity of PD is a key regulatory step; inhibition of PD Carnitine palmitoyltransferase deficiency results in accumulation of pyruvate, which is metabolized Disorders of mitochondrial function to lactate. Factors inhibiting PD activity include NADH, acetyl CoA, ATP and an anaerobic environment [4, 6, 7]. Disorders of muscle bulk Muscular dystrophies Free fatty acids (FFA) become an important source of Chronic motor neuropathies and neuropathies oxidative metabolism with prolonged exercise (130 min) Inflammatory disorders as well as during fasting. In this form of oxidation, FFA, particularly long chain FFA (LCFAs), are conjugated to Disorders of muscle activation carnitine-producing acylcarnitine which is transferred Focal or multifocal brain lesions into the mitochondria by carnitine palmitoyltransferase Neurodegenaritive disorders (CPT) [8]. As such, carnitine has two important functions, Psychiatric disorders facilitating transport of LCFAs into mitochondria and modulating the intramitochondrial coenzyme A (CoA)/ myopathies [3]. As such, abnormalities of muscle metabo- acyl-CoA ratio [9]. In the mitochondria, further conver- lism can affect the breakdown and utilization of carbohy- sion occurs to acetyl CoA and flavin adenine dinucleotide drates, lipid utilization, or mitochondrial oxidation [3]. A (FADH2) by acyl dehydrogenases. In the nutritionally basic understanding of substrate utilization at the level of replete state, acetyl CoA enters the TCA cycle while in the the skeletal muscle provides a useful framework to under- fasting state it is metabolized to ß-hydroxybutyrate. stand the varying exercise response noted in metabolic myopathies. Oxidative phosphorylation of ADP to ATP takes place within the inner mitochondrial membrane by five protein The force-generating elements of the muscle cell are complexes that utilize NADH from the TCA cycle and the myofibrils. Contraction occurs with the cyclical asso- FADH2 from FFA oxidation. These five enzyme com- ciation between the myosin thick filament and the actin plexes contain approximately 100 different protein sub- thin filament; each reversible cycle of association is asso- units [10] and include: complex I (NADH dehydroge- ciated with hydrolysis of a molecule of ATP [2]. As such, nase), complex II (succinate dehydrogenase), complex III the hydrolysis of ATP to ADP and inorganic phosphate (cytochrome c reductase), Complex IV (cytochrome c oxi- (Pi) provides the source of energy for contraction. Phos- dase) and complex V (ATP synthase). This respiratory phocreatine (PCr) buffers the ATP concentration during a chain produces ATP by creating a proton gradient across rapid increase in energy use by muscle; only after PCr is the inner mitochondrial membrane [10]. depleted does a significant portion of the energy for con- traction come from the hydrolysis of ATP. Unfortunately, It is apparent from this brief review that abnormalities the concentration of ATP in muscle is sufficient to pro- in any of these pathways can affect ATP generation and vide only a few seconds of maximal exercise [2]. Metabol- subsequent skeletal muscle function. For example, abnor- ic pathways within the cell that guarantee the supply of malities of PD function alter glucose metabolism leading ATP include glycolysis and fatty acid oxidation [4]. These to excess metabolism of lactate from pyruvate. Similarly, are illustrated in figure 1. abnormalities in ß-fatty acid oxidation compromise the use of FFA for ATP production. This will produce partic- Glycolysis is the principal pathway utilized in the ular functional abnormalities in the fasting state or after a nutritionally replete state. Glycogen stores are limited, meal high in fat content [4]. Most importantly, diseases however, and are depleted rapidly. Oxidative metabolism that affect oxidative phosphorylation can markedly im- of glycogen requires an adequate blood flow and oxygen pair ATP generation and skeletal muscle function during supply but produces a much higher yield of ATP than exercise. anaerobic glycolysis [2]. Anaerobic glycogenolysis, which can be used during periods of oxygen deficiency, is associ- Cardiopulmonary Exercise Testing 243 for Neuromuscular Disorders

244 Flaherty/Weisman/Zeballos/Martinez

Fig. 2. Maximal oxygen uptake in McArdle patients compared with creatinine kinase levels at rest which can rise dramatically healthy subjects, patients with carnitine palmityltransferase deficien- with attacks of myoglobinuria [3, 11]. cy, patients with myalgias, patients with muscular dystrophy and patients with electron transport defects [from 28, with permission]. The hallmark defect in muscle glycogenolysis is a fail- ure of lactate to rise during ischemic exercise. This rather Disorders of Carbohydrate Metabolism straightforward test involves inflating a blood pressure cuff 10 mm Hg above systolic pressure while the patient The glycogenoses are disorders of glycogen or glucose squeezes a hand ergometer. Serial blood levels of ammo- metabolism, and usually present with exertional muscle nia and lactate are measured; in normal subjects the fatigue, myalgia, contractures, or myoglobunuria [11]. ammonia and lactate levels increase by at least 3- to 4- Numerous disorders are recognized as recently reviewed fold- and return to normal levels within 10–15 min of by Amato [12]. As noted in figure 1, there are several completing exercise [16]. In patients with glycogen storage potential sites for abnormalities of carbohydrate metabo- myopathy, lactate levels do not rise and, in fact, lactate lism. levels may fall although ammonia levels do rise [5, 16]. The measurement of changes in ammonia levels serves as Muscle Phosphorylase Deficiency a control to evaluate the adequacy of patient effort [11]. In The prototypical disorder of muscle glycogenolysis is addition, changes in pH parallel lactate production, so a McArdle’s disease (type V glycogen storage disorder or lack of normal decline in pH is seen [5]. Given these phys- muscle phosphorylase deficiency) which is attributed to iologic abnormalities, it is expected that maximal exercise mutations in the muscle glycogen phosphorylase gene [13]. testing may provide important insights and add valuable The onset of symptoms typically occurs between ages 10 diagnostic information. and 30 years [11]. Although patients appear normal at rest, intense or isometric exercise provokes muscle cramps, The published literature suggests that patients with painful contractures with myoglobinuria, and occasionally muscle myophosphorylase deficiency exhibit a decreased renal failure [14]. Interestingly, many patients describe a maximal VO2 to approximately one-third to one-half of ‘second-wind’ phenomenon where initial submaximal ex- the value achieved by sedentary control subjects [3, 5, 17– ercise provokes fatigue and myalgia which is followed by 22]. This is illustrated in figure 2. The limitation in VO2 decreased limitation during subsequent exercise [11, 15]. seems to reflect substrate limitation and impaired oxida- Initial laboratory examination may demonstrate elevated tive phosphorylation [5, 23]. The latter is felt to reflect impaired substrate delivery to mitochondria [23]. The Fig. 1. Schematic illustration of the metabolic pathways (glycolysis, biochemical defect in this disorder highlights the impor- fatty acid oxidation, and oxidative phosphorylation) and selected tant role of glycogen as the fuel utilized to rapidly estab- associated disorders within a muscle cell. lish an oxidative steady state [5]. Its absence leads to marked fluctuation in exercise capacity according to the availability of alternative fuels. For example, infusions of glucose during exercise in patients with McArdle’s disease improves maximal VO2, the maximal a-v O2 difference and abolishes the excessive ATP degradation in working muscles [5, 24]. As a correlate, infusion of FFA, thereby increasing their availability during exercise, also leads to similar improvements during exercise [5]. The physiolog- ic correlate of these changes is the ‘second-wind’ phenom- enon. This refers to the characteristic increase in exercise capacity that is seen spontaneously in patients with McArdle’s disease during prolonged exercise [25]. Two distinct phases of exercise have been described during prolonged, submaximal exercise [15]. The first phase (‘ad- aptation phase’) is seen during the initial 15 min of exer- cise at a workload 30% of VO2max; the second phase (‘second wind’) followed during which patients were able to continue exercise without difficulty. The ‘second wind’ phase was characterized by an increase in cardiac output Cardiopulmonary Exercise Testing 245 for Neuromuscular Disorders

Fig. 3. Values of respiratory exchange ratio (RER) for 7 subjects with exercise-induced pain and contracture [11], although it is McArdle’s (solid symbols) and 7 control subjects (open symbols) at a rarer disorder [29]. In contrast to McArdle’s disease, rest, 50% peak VO2, 80% peak VO2, and after 3 min of recovery exercise intolerance occurs earlier and is more severe; a [from 17, with permission]. compensated hemolysis can also be seen [14]. Many of the responses to testing are similar to patients with myophos- and metabolic changes causing increases in blood-borne phorylase deficiency. As such, lactate does not rise (and glucose and FFA delivery to peripheral muscle. often falls) with ischemic arm testing [5]. Similarly, the normal rise in the lactate/pyruvate ratio is not seen. A unique manifestation of the biochemical defect in myophosphorylase function and the inability to utilize As expected, the functional consequences during maxi- glycogen during exercise is a failure of the respiratory mal exercise testing are similar to those seen in patients exchange ratio (RER) to rise above one despite maximal with myophosphorylase deficiency. In a study of 7 pa- exercise [17, 22]. This is reflected in figure 3 where the tients with phosphofructokinase deficiency (PFKD), 5 response to exercise in patients with McArdle’s disease is with selective deficiency of CPT (CPTD) and 6 healthy contrasted with normal controls [17]. This may serve as a controls, maximal VO2 and arteriovenous oxygen defi- diagnostic finding in patients examined for typical symp- ciency were markedly lower in those with PFKD (fig. 4) toms. In addition, the biochemical defect impairs the abil- [30]. Similarly, the rise in RER is attenuated [5]. Interest- ity of skeletal muscle to generate lactic acid during exer- ingly, peak exercise cardiac output and heart rate were cise [22]. In fact, muscle lactate can decrease during exer- similar among the three groups (fig. 4) [30]. These data cise in McArdle’s disease patients [26]. Despite the inabil- suggest that the subnormal maximal VO2 is primarily due ity to generate lactic acid, a ventilatory threshold has been to subnormal capacity for muscle O2 extraction [30]. The described by some [3, 19, 27], although not by others [17]. main difference between PFKD and myophosphorylase As such, the hyperventilation noted during exercise may deficiency is the inability of muscle deficient in PFK to be related to other factors including discomfort and alter- utilize glucose. As a result of this inability to utilize glu- nate metabolic signals [17]. In addition to hyperventila- cose, muscle function during exercise should be closely tion, McArdle’s disease patients demonstrate a hyperdy- tied to the ability to utilize FFA. This was confirmed by namic response during exercise. As such, an exaggerated Haller and Lewis [31] who studied 4 PFK-deficient pa- heart rate response and cardiac output response have tients during intravenous glucose infusion, overnight fast- been described [15, 28]. ing, and intravenous infusion of triglycerides and heparin. These interventions led to marked fluctuations in plasma FFA; lower FFA levels resulted in a lower maximal arte- riovenous difference and lower maximal VO2 [31]. This phenomenon has been termed the ‘out-of-wind’ phenom- enon to describe deterioration in exercise capacity after high carbohydrate ingestion in PFKD patients. Disorders of Lipid Metabolism Fatty acids present an important fuel for muscle func- tion, particularly with prolonged exertion and in the fast- ing state. The ß-oxidation of fatty acids takes place in the mitochondria; the translocation of these FFA requires carnitine and CPT [8]. As such, two defects can be associ- ated with impaired function of lipid metabolism during exercise, carnitine deficiency syndromes, and defects of CPT. Phosphofructokinase Deficiency Carnitine Deficiency Deficiency of phosphofructokinase, Tarui disease, Carnitine deficiency occurs when there is an insuffi- presents in a similar fashion to McArdle’s disease with cient amount of intracellular carnitine to accomplish the 246 Flaherty/Weisman/Zeballos/Martinez

Fig. 4. Mean O2 uptake (a), arteriovenous O2 difference (b), cardiac output (c), and heart rate (d) at rest ( g ) and during peak exercise ( X ) in patients with muscle phosphofructo- kinase deficiency (PFKD) and carnitine pal- mitoyltransferase deficiency (CPTD). * Sig- nificantly (p ! 0.001) lower peak exercise values for PKFD patients than for CPTD patients or healthy subjects [from 30, with permission]. primary functions of carnitine [9]. Carnitine deficiency normal lactate response. Maximal exercise testing has may occur as a primary deficiency although more fre- suggested that maximal VO2 is generally not impaired quently it is seen as a secondary deficiency [9, 32]. There (see fig. 2 and 4). Similarly, peak cardiac output and heart are numerous causes of secondary systemic carnitine defi- rate are similar to matched control subjects (see fig. 4) ciency including Acyl-CoA dehydrogenase deficiency, or- [30], as are other responses linked to oxidative phosphory- ganic acidemias, mitochondrial respiratory disorders and lation [3, 5, 30, 36, 37]. Interestingly, the RER has been numerous systemic disorders [9, 32]. In 22 of 77 patients reported to be higher than normal, corresponding to the with mitochondrial myopathies, abnormal carnitine dis- dependence on carbohydrate metabolism [5, 37, 38]. tribution was noted in muscle; this was equally seen in individuals with lipid storage myopathy (31.5%) and Disorders of Mitochondrial Function those with ragged red fibers (see below; 25.6%) [33]. Inter- estingly, this same group has reported improvement in The term mitochondrial myopathy refers to various these patients with L-carnitine supplementation [34]. syndromes with diverse pathologic, histochemical, and biochemical characteristics. These syndromes are often Carnitine Palmitoyltransferase Deficiency (CPTD) multisystemic with varying signs and symptoms affecting CPT exists as two genetically and catalytically distinct any organ system [39, 40]. The final pathogenesis of the enzymes (CPT I and CPT II) [35]. Inherited defects in the different syndromes is a decline in mitochondrial adeno- CTP II gene cause a spectrum of disorders with symptoms sine triphosphate (ATP)-generating capacity leading to a including attacks of myalgia, cramps, and muscle stiffness deficit in energy production [41]. Exercise intolerance is or tenderness [8]. As expected from the biologic defect, one manifestation described in some patients [3, 41]. As symptoms are most commonly precipitated by prolonged muscle requires oxidative phosphorylation for ATP pro- exertion. In addition, symptoms can be precipitated with duction, mitochondrial dysfunction can also produce fasting, cold exposure or a high fat intake [8]. In contrast muscular symptoms such as myalgia or weakness. Al- to the glycogenoses, ischemic exercise testing results in a Cardiopulmonary Exercise Testing 247 for Neuromuscular Disorders

though these disorders were previously thought to be rare their patients with either mitochondrial myopathy or [39], a recent study estimated the prevalence of patients nonmitochondrial myopathy. They described a maximal with mitochondrial myopathy and unexplained exertion- inspiratory pressure (MIP) below the normal range in 0/ al limitation to be 8.5% [1]. 14 patients with mitochondrial myopathy and 2/6 pa- tients with nonmitochondrial myopathy. Similarly, a The diagnosis of a metabolic or mitochondrial myopa- maximum expiratory pressure (MEP) was below the nor- thy is complicated by the wide array of presenting signs mal range in only 4/14 of the patients with mitochondrial and symptoms that can affect any organ system and vary myopathy and 2/6 patients with nonmitochondrial myop- in severity between patients [39]. The exact criteria athy. required to make a diagnosis are also somewhat variable. A recent review recommended the following major diag- In the 28 patients described by Flaherty et al. [1], a nostic criteria: (1) abnormal histology (more than 2% rag- lower MIP (77 vs. 115% predicted; p = 0.01) and maximal ged-red fibers in a muscle biopsy; (2) abnormal enzymatic transdiaphragmatic pressure (Pdimax) (80 vs. 144 cm activity (more than 20% reduction compared to age- H2O; p = 0.0004) was seen in patients with mitochondrial matched controls), and (3) identification of genetic muta- myopathies compared to normal controls. No difference tions with undisputed pathogenicity [42]. A detailed dis- in transdiaphragmatic sniff pressure (Pdi sniff) was noted. cussion of the histologic features, enzymatic assays, and Furthermore, patients were divided into a group that genetic mutations is beyond the scope of this chapter but stopped exercise primarily due to dyspnea (n = 16) and a can be found in several recent excellent reviews [16, 43– group that stopped exercise primarily due to fatigue (n = 46]. Physiologic testing can potentially be utilized as a 11). There was a trend for patients that stopped due to screening tool for patients with suspected mitochondrial dyspnea to have lower respiratory muscle pressures as myopathy prior to performing a muscle biopsy. measured by MIP (64 vs. 95% predicted; p = 0.06), maxi- mal expiratory force (45 vs. 57% predicted; p = 0.22), Pdi Pulmonary Function Testing sniff (70 vs. 100 cm H2O; p = 0.01), and Pdimax (61 vs. Dandurand et al. [41] evaluated pulmonary function 115 cm H2O; p = 0.01). Patients that stopped exercise pri- and blood lactate levels in 13 patients with mitochondrial marily due to dyspnea also had a lower MVV (93 vs. myopathies, 7 patients with nonmetabolic myopathies, 128 l/min; p = 0.01) compared to patients that stopped and 12 healthy control subjects. In general, pulmonary exercise due to fatigue. function tests were normal and similar between disease groups although the mean functional residual capacity These studies suggest that respiratory muscle weakness was greater (119% predicted vs. 90% predicted; p = 0.006) may be present in patients with a mitochondrial myopa- for patients with mitochondrial myopathies compared to thy. Furthermore, respiratory muscle weakness may be nonmetabolic myopathy patients. Flaherty et al. [1] re- more prevalent in patients that manifest respiratory cently evaluated 28 patients with unexplained dyspnea symptoms compared to patients that complain primarily that were found to have mitochondrial myopathies and of fatigue. compared physiologic parameters to 11 healthy controls. No differences were noted in spirometry, lung volumes, Exercise Testing or gas transfer. However, the maximal voluntary ventila- Dandurand et al. [41] evaluated exercise performance tion was lower in patients as compared to controls (111 vs. and blood lactate levels in 13 patients with mitochon- 186 l/min; p ! 0.0001). These data suggest that mitochon- drial myopathies, 7 patients with nonmetabolic myopa- drial myopathies may be present in patients with unex- thies, and 12 health control subjects. Exercise results did plained dyspnea and normal resting pulmonary function. not differ between disease groups. Patients with mito- chondrial myopathy demonstrated a shorter duration of Respiratory Muscle Function exercise (6.9 vs. 11.6 min, p ! 0.001), lower percent pre- Several case reports have suggested that patients with dicted maximal work capacity (47 vs. 127, p ! 0.001), mitochondrial myopathy may manifest respiratory mus- lower percent predicted VO2max (61 vs. 115, p ! 0.001), cle weakness [47, 48]. These reports describe 3 patients lower VO2/kg at maximal workload (17 vs. 39, p ! (ages 27–70) that presented with acute respiratory failure 0.001), lower percent predicted VO2 at anaerobic thresh- requiring prolonged weaning from ventilatory support. old (AT) (35 vs. 71, p ! 0.001), lower percent predicted Interestingly, Dandurand et al. [41] did not report a sig- peak heart rate (85 vs. 101, p ! 0.001), and lower percent nificant prevalence of respiratory muscle weakness in predicted minute ventilation (52 vs. 76, p ! 0.01) [41]. Interestingly, 12/13 patients with mitochondrial myopa- 248 Flaherty/Weisman/Zeballos/Martinez

Fig. 5. Illustrative case of a 36-year-old fe- male with a 1-year history of progressive exertional dyspnea, previously a 100-mile/ week cyclist, with biopsy-proven mitochon- drial myopathy. A markedly reduced aerobic capacity (a), hypercirculatory response (b), hyperventilatory response (c, d) and an inde- terminate anaerobic threshold (e) are evi- dent [from 1, with permission]. thies exhibited abnormal heart rate responses compared 104 l/min; p ! 0.001), and elevated VE/VO2 (59 vs. 41; to 3/12 healthy controls and 3/7 nonmetabolic myopathy p = 0.02) and VE/VCO2 (55 vs. 42; p = 0.002) at peak patients (p = 0.02). Heart rate abnormalities included exercise. Importantly, individual responses were variable fixed elevations and an abnormally elevated heart rate for with 5 patients achieving a VO2max greater than 90% of a given work rate; a single patient had a heart rate below predicted. The response of a typical patient is illustrated predicted [41]. in figure 5. In the series of Flaherty et al. [1] mitochondrial myopa- Hooper et al. [49] also described the exercise response thy patients achieved a lower maximum VO2 (67 vs. for 3 patients with mitochondrial myopathy that present- 104% predicted; p ! 0.0001), lower AT (46 vs. 65% pre- ed with unexplained dyspnea. The exercise response was dicted VO2max; p = 0.03), elevated heart rate response variable with VO2max ranging from 22 to 84% predicted. (defined as change in heart rate/change in VO2, 91 vs. In response to steady-state exercise, all patients demon- 41; p = 0.01), lower maximal minute ventilation (71 vs. strated an early and rapid elevation of heart rate to greater Cardiopulmonary Exercise Testing 249 for Neuromuscular Disorders

Fig. 6. Relationship between heart rate and VO2 during progressive pyruvate level (normally !20) [54], and the subanaerobic incremental exercise testing in 3 patients with mitochondrial enzyme threshold exercise test (SATET) [55], may be useful in deficiency presenting as exercise limitation in normal-appearing screening patients for potential mitochondrial myopa- adults [from 49, with permission]. thies. The sensitivity and specificity of the SATET test was evaluated by Nashef and Lane [55] in 29 normal vol- than 80% of predicted maximum (fig. 6). No evidence of unteers and 6 patients with mitochondrial myopathy. The cardiac dysfunction was present despite detailed testing AT and predicted work rate were calculated [56] and sub- which included right heart catheterization during exercise jects exercised at 60 rpm for up to 15 min at 90% of their [49]. A single case report has confirmed an increased O2 predicted work rate for AT. Venous lactate levels were delivery but abnormal O2 extraction (markedly reduced collected pre-, post- and 30 min post-exercise. Only 2/29 C(a-v)O2) supporting abnormal skeletal muscle oxidative controls had a peak lactate of 15 mM; all patients had a metabolism [50]. An additional study, using noninvasive peak lactate level 15 mM. Using 5 mM as a cut-point, the estimates of cardiac output during maximal and submaxi- sensitivity of this test was 100% and the specificity was mal exercise, confirmed exercise intolerance related to 93% for identifying patients with mitochondrial myopa- impaired peripheral oxygen extraction [51]. It is evident thies [55]. that a hyperdynamic circulatory response appears to be one of the more consistent findings in patients with meta- Finsterer et al. [57] also evaluated the test characteris- bolic myopathies [3, 50]. This may relate to the regulatory tics for a lactate stress test for the diagnosis of mitochon- role that abnormalities in muscle oxidative metabolism drial myopathy. In this study, 31 healthy controls, 10 may play on the cardiovascular response as a reflection of patients with nonmitochondrial myopathy, and 26 pa- the normal coupling between O2 utilization and delivery tients with mitochondrial myopathy were evaluated. Sub- [51, 52]. Unfortunately, this hypercirculatory response is jects exercised on a cycle ergometer for 15 min at a con- not specific for mitochondrial myopathies; some investi- stant work rate of 30 W. Venous lactate was assayed at gators have described impaired muscle oxidative metabo- baseline; 5, 10, and 15 min after starting exercise; and lism in patients with exercise intolerance of unexplained 15 min after the completion of exercise. No control sub- origin [53]. jects and only 1 of the 10 nonmitochondrial myopathy patients experienced an increase in lactate level. This con- Deficiency of oxidative metabolism results in more trasted to 18 of the 31 patients with mitochondrial myop- glycolytic anaerobic activity leading to increased lactate athy that had an abnormal lactate stress test. This corre- production within the tissue [54]. Several investigators sponded to a sensitivity of 69% and a specificity of 90% have evaluated the ability of lactate levels, both at rest and for the diagnosis of a mitochondrial myopathy. These with exercise, to identify patients with mitochondrial studies demonstrate that an elevated lactate level at a low myopathies. It has also been suggested that the lactate to workload can be seen in patients with mitochondrial my- opathies. Although the sensitivity is not 100%, this find- ing may be useful to help screen patients for the presence of a mitochondrial myopathy prior to proceeding with a muscle biopsy and could help distinguish patients with a mitochondrial myopathy from patients with disorders of muscle glycogenolysis. A hyperventilatory response has frequently been noted in mitochondrial disorders (see fig. 5) [1]. The etiology is unknown, but has been postulated to occur in response to the excess of CO2 produced by the buffering of lactate [49]. An alternate hypothesis suggests that hyperventila- tion relates to an increase in respiratory drive originating in metabolically sensitive chemoreceptors localized in peripheral skeletal muscles, similar to the mechanism described to explain the hypercirculatory response [52]. It is evident from these data that patients with a mito- chondrial myopathy, on average, present with a reduced peak VO2, a hyperdynamic cardiac response, a normal to 250 Flaherty/Weisman/Zeballos/Martinez

low AT, and a higher minute ventilation per level of VO2. ilar levels in both myopathic patient groups. The authors It is likely that deconditioning plays some part in this suggested that the pattern was most consistent with de- abnormal response [41, 58, 59]. After 8 weeks of training, conditioning in both groups. A similar level of aerobic a 30% increase in aerobic capacity, reduction in blood lac- limitation has been documented in a separate study of 11 tic acid and improvement in ADP recovery has been patients with polymyositis [61]. Although 8 of 11 patients reported in patients with mitochondrial myopathy [59]. had normal pulmonary function, 9 of the patients exhib- This improvement is greater (30%) than seen in patients ited a decreased maximal VO2. The authors suggested with nonmetabolic myopathy (16%) and normal controls that occult pulmonary hypertension could account for (10%) after 8 weeks of exercise training [58]. Importantly, some of the observed limitation (7/11 exhibited echo- the aerobic capacity of these patients after training was cardiographic evidence of pulmonary vascular abnor- still reduced as compared to the sedentary normal control mality). The importance of deconditioning is support- group before training. ed by the improvement in aerobic capacity documented with short-term training in patients with mitochondrial Recent data have shed further light on the pathophysi- myopathies and patients with nonmetabolic myopathies ologic basis of symptomatic limitation in patients with (predominantly muscular dystrophies) [58]. After an 8- mitochondrial myopathies. Mitochondrial myopathy pa- week rehabilitation program, patients with nonmetabolic tients terminating exercise because of fatigue demonstrate myopathies improved submaximal workload achieved less impairment of respiratory muscle function than those (4.90–5.64 METs), heart rate during exercise (148–134 stopping exercise because of breathlessness [1]. In addi- beats/min), and lactate post-exercise (3.0–2.25 mmol/l) tion, increased breathlessness in these patients appears to [58]. relate to abnormal respiratory muscle recruitment during exercise. Several groups have reported similar data in patients with late sequelae of poliomyelitis. Stanghelle et al. [62] Disorders of Muscle Bulk examined 68 consecutive patients with post-polio syn- drome (31 exercised with arm ergometry and 37 with Given the requirement of preserved muscle function to bicycle ergometry). Although mean FVC was normal for maintain a normal exercise capacity, several groups have the patient group, a pronounced reduction in maximal examined the role of CPET in disorders of muscle bulk or VO2 was seen (32 with VO2max ! 60% predicted and 16 strength. Carroll et al. [60] examined exercise capacity in !50% predicted); this was particularly evident in females. 15 controls and 29 patients with a variety of neuromuscu- Fifteen patients were felt to have ‘pulmonary limitation’ lar diseases (10 metabolic myopathies, 8 muscular dystro- based on a maximal VE/MVV ratio above 70%. The phies and 11 miscellaneous disorders). Patients with neu- authors suggested a strong component of deconditioning. romuscular disease exhibited a decreased VO2max. Those Willen et al. [63] extended these findings by studying 32 with nonmetabolic myopathies achieved a maximal VO2 patients with post-polio syndrome. Decrements in maxi- of 19.2 ml/kg compared with 36.6 ml/kg in the control mal VO2 were seen among males (69% predicted) and subjects. Similarly, Dandurand et al. [41] studied 7 pa- females (79% predicted). Interestingly, AT was lower in tients with miscellaneous, nonmetabolic myopathies (in- males (34% maximal VO2) than females (47% maximal flammatory myopathies) noting a mean VO2max % pre- VO2). Strong correlations were noted between leg muscle dicted of 62.1% which was similar to that achieved by 15 strength and peak VO2 and workload. As in other non- patients with mitochondrial disease but significantly low- metabolic myopathies, patients with post-polio symptoms er than the 115.7% achieved by 12 healthy normals. have demonstrated improvement in aerobic capacity with Although the maximal achieved heart rate was similar in exercise training suggesting some component of poor con- mitochondrial and nonmetabolic myopathy patients, the ditioning [64]. Additional data have been reported in a latter group experienced less frequent abnormalities (3/7 group of 5 patients with significant respiratory muscle patients) than the former group (12/13 patients). Similar- involvement (mean VC 38% predicted) as a sequelae of ly, observed VE was similar between both groups of myo- polio [65]. All patients exhibited decreased exercise toler- pathic patients and a rapid, shallow breathing pattern was ance with a respiratory limitation in all but 1 of the frequently seen in both patient groups (11/13 mitochon- patients. Importantly, all patients demonstrated impaired drial myopathy patients and 7/7 nonmetabolic myopathy diaphragmatic function at rest and a decrease in peak gas- patients). Lastly, peak lactate and the AT occurred at sim- tric pressure during inspiration at peak exercise. Although a definitive conclusion cannot be reached, this pattern of Cardiopulmonary Exercise Testing 251 for Neuromuscular Disorders

breathing supports increased accessory muscle recruit- Conclusion ment during exercise. As such, nonmetabolic disorders are generally associated with impaired aerobic capacity, Given the importance of skeletal muscle to normal strongly suggestive of poor conditioning. With more se- exercise response, CPET is frequently abnormal in pa- vere disease, respiratory limitation is more likely to be tients with myopathies. Metabolic and nonmetabolic seen, particularly in the presence of respiratory muscle myopathies demonstrate stereotypical responses although involvement. the contribution of deconditioning confounds the inter- pretation of testing in both groups of disorders. References 1 Flaherty KR, Wald J, Weisman IM, Zeballos 14 DiMauro S, Tsujino S: Nonlysosomal glycoge- 26 Lofberg M, Lindholm H, Naveri H, Majander RJ, Schork MA, Blaivas M, Rubenfire M, Mar- noses; in Engel A, Franzini-Armstrong C (eds): A, Suomalainen A, Paetau A, Sovijarvi A, Har- tinez FJ: Unexplained exertional limitation. Myology. New York, McGraw-Hill, 1994, pp konen M, Somer H: ATP, phosphocreatine and Characterization of patients with a mitochon- 1554–1576. lactate in exercising muscle in mitochondrial drial myopathy. Am J Respir Crit Care Med disease and McArdle’s disease. Neuromusc 2001;164:425–432. 15 Braakhekke JP, De Bruin MI, Stegeman DF, Disord 2001;11:370–375. Wevers RA, Binkhorst RA, Joostein EMG: The 2 Taivassalo T, Reddy H, Matthews PM: Muscle second wind phenomenon in McArdle’s dis- 27 Hagberg JM, King DS, Rogers MA, Montain responses to exercise in health and disease. ease. Brain 1986;109:1087–1101. SJ, Jilka SM, Kohrt WM, Heller SL: Exercise Neurol Clin 2000;18:15–34. and recovery ventilatory and VO2 responses of 16 Pourmand R: Metabolic myopathies: a diag- patients with McArdle’s disease. J Appl Physiol 3 Elliot DL, Neil RM, Buist M, Goldberg L, Ken- nostic evaluation. Neurol Clin 2000;18:1–13. 1990;68:1393–1398. naway NG, Phil D, Powell BR, Kuehl KS: Met- abolic myopathies: Evaluation by graded exer- 17 Riley M, Nicholls DP, Nugent AM, Steele IC, 28 Lewis SF, Hall RG: The pathophysiology of cise testing. Medicine 1989;68:163–172. Bell N, Davies PM, Stanford CF, Patterson McArdle’s disease: clues to regulation in exer- VH: Respiratory gas exchange and metabolic cise and fatigue. J Appl Physiol 1986;61:391– 4 Clay AS, Behnia M, Brown KK: Mitochondrial responses during exercise in McArdle’s disease. 401. disease. A pulmonary and critical-care medi- J Appl Physiol 1993;75:745–754. cine perspective. Chest 2001;120:634–648. 29 Vora S, Durham S, deMartinville B, George 18 Haller RG, Lewis SF, Cook JD, Blomqvist CG: DL, Francke U: Assignment of the human gene 5 Haller RG, Bertocci LA: Exercise evaluation of Myophosphorylase deficiency impairs muscle for muscle-type phosphofructokinase (PFKM) metabolic myopathies; in Engel A, Franzini- oxidative metabolism. Ann Neurol 1985;17: to chromosome 1 (Region CEN-Q32) using so- Armstrong C (eds): Myology. Basic and Clini- 196–199. matic cell hybrids and monoclonal anti-M anti- cal, ed 2. New York, McGraw-Hill, 1994. body. Somat Cell Genet 1982;8:95–104. 19 Hagberg JM, Coyle EF, Carroll JE, Miller JM, 6 Taivassalo T, Matthews PM, DeStefano N, Sri- Martin WH, Brooke MH: Exercise hyperventi- 30 Lewis SF, Vora S, Haller RG: Abnormal oxida- pathi N, Genge A, Karpati G, Arnold DL: lation in patients with McArdle’s disease. J tive metabolism and O2 transport in muscle Combined aerobic training and dichloroace- Appl Physiol: Environ Exercise Physiol 1982; phosphofructokinase deficiency. J Appl Physi- tate improve exercise capacity and indices of 52:991–994. ol 1991;70:391–398. aerobic metabolism in muscle cytochrome oxi- dase deficiency. Neurology 1996;47:529–534. 20 Kissel JT, Beam W, Bresolin N, et al: Physio- 31 Haller RG, Lewis SF: Glucose-induced exer- logic assessment of phosphorylate mutase defi- tional fatigue in muscle phosphofructokinase 7 DeStefano N, Matthews PM, Ford B, Genge A, ciency. Neurology 1985;35:828–833. deficiency. N Engl J Med 1991;324:364–367. Karpati G, Arnold DL: Short-term dichloro- acetate treatment improves indices of aerobic 21 Andersen KI, Lund-Johansen P, Clausen G: 32 Infante JP, Huszagh VA: Secondary carnitine metabolism in patients with mitochondrial dis- Metabolic and circulatory responses to muscu- deficiency and impaired docosahexaenoic orders. Neurology 1995;45:1193–1198. lar exercise in a subject with glycogen storage (22:6n-3) acid synthesis: a common denomina- disease (McArdle’s disease). Scand J Clin Lab tor in the pathophysiology of diseases of oxida- 8 Zierz S: Carnitine palmitoyltransferase defi- Invest 1969;24:105–113. tive phosphorylation and ß-oxidation. FEBS ciency; in Engel A, Franzini-Armstrong C (eds): Letters 2000;468:1–5. Myology. Basic and Clinical, ed 2. New York, 22 Chaussain M, Camus F, Defoligny C, Eymard McGraw-Hill, 1994, pp 1577–1585. B, Fardeau M: Exercise intolerance in patients 33 Campos Y, Huertas R, Bautista J, Gutierrez G, with McArdle’s disease or mitochondrial myo- Segura D, Villanueva M, Cabello A, Alesso L, 9 Breningstall GN: Carnitine deficiency syn- pathies. Eur J Med 1992;1:457–463. Arenas J: Muscle carnitine deficiency and lipid dromes. Ped Neurol 1990;6:75–81. storage myopathy in patients with mitochon- 23 DeStefano N, Argov Z, Matthews PM, Karpati drial myopathy. Muscle Nerve 1993;16:778– 10 Larsson NG, Oldfors A: Mitochondrial myopa- G, Arnold DL: Impairment of muscle mito- 781. thies. Acta Physiol Scand 2001;171:385–393. chondrial oxidative metabolism in McArdle’s disease. Muscle Nerve 1996;19:764–769. 34 Campos Y, Huertas R, Lorenzo G, Bautista J, 11 Martin A, Haller RG, Barohn R: Metabolic Gutierrez G, Aparicio M, Alesso L, Arenas J: myopathies. Curr Opin Rheumatol 1994;6: 24 Mineo I, Kono N, Yamada Y, Hara N, Kiyoka- Plasma carnitine insufficiency and effective- 552–558. wa H, Hamaguchi T, Kawachi M, Yamasaki T, ness of L-carnitine therapy in patients with Nakajima H, Kuwajima M, et al: Glucose infu- mitochondrial myopathy. Muscle Nerve 1993; 12 Amato AA: Acid maltase deficiency and relat- sion abolishes the excessive ATP degradation 16:150–153. ed disorders. Neurol Clin 2000;18:151–163 in working muscles of a patient with McArdle’s disease. Muscle Nerve 1990;13:618–620. 35 Vladutiu GD: Biochemical and molecular cor- 13 Kubisch C, Wicklein EM, Jentsch TJ: Molecu- relations in carnitine palmitoyltransferase II lar diagnosis of McArdle disease: revised ge- 25 Pearson C, Rimer D, Mommaerts WFHM: A deficiency. Muscle Nerve 1999;22:949–951. nomic structure of the myophosphorylase gene metabolic myopathy due to absence of muscle and identification of a novel mutation. Hum phosphorylase. Am J Med 1961;30:502–517. Mut 1998;12:27–32. 252 Flaherty/Weisman/Zeballos/Martinez

36 Haller RG, Lewis SF, Cook JD, Blomqvist CG: 48 O’Brien A, Blaivas M, Albers J, Wald J, Watts 58 Taivassalo T, DeStefano N, Chen J, Karpati G, Hyperkinetic circulation during exercise in C: A case of respiratory muscle weakness due to Arnold DL, Argov Z: Short-term aerobic train- neuromuscular disease. Neurology 1983;33: cytochrome c oxidase enzyme deficiency. Eur ing response in chronic myopathies. Muscle 1283–1287. Respir J 1998;12:742–744. Nerve 1999;22:1239–1243. 37 Carroll JE, Brooke MH, DeVivo DC, Kaiser 49 Hooper RG, Thomas AR, Kearl RA: Mito- 59 Taivassalo T, DeStefano N, Argov Z, Mat- KK, Hagberg JM: Biochemical and physiologic chondrial enzyme deficiency causing exercise thews PM, Chen J, Genge A, Karpati G, Ar- consequences of carnitine palmityltransferase limitation in normal-appearing adults. Chest nold DL: Effects of aerobic training in patients deficiency. Muscle Nerve 1978;1:103–110. 1995;107:317–322. with mitochondrial myopathies. Neurology 1998;50:1055–1060. 38 Layzer RB, Havel RJ, McIlroy MB: Partial 50 Haller RG, Henriksson KG, Jorfeldt L, Hult- deficiency of carnitine palmityltransferase: man E, Wibom R, Sahlin K, Areskog N, Gun- 60 Carroll JE, Hagberg JM, Brooke MH, Shumate Physiologic and biochemical consequences. der M, Ayyad K, Blomqvist CG, et al: Defi- JB: Bicycle ergometry and gas exchange mea- Neurology 1980;30:627–633. ciency of skeletal muscle succinate dehydroge- surements in neuromuscular diseases. Arch nase and aconitase. J Clin Invest 1991;88: Neurol 1979;36:457–461. 39 Chinnery PF, Turnbull DM: Mitochondrial 1197–1206. medicine. QJM 1997;90:657–667. 61 Hebert CA, Byrnes TJ, Baethge BA, Wolf RE, 51 Vissing J, Galbo H, Haller RG: Exercise fuel Kinasewitz GT: Exercise limitation in patients 40 Andreu AL, Hanna MG, Reichmann H, Bruno mobilization in mitochondrial myopathy: A with polymyositis. Chest 1990;98:352–357. C, Penn AS, Tanji K, Pallotti F, Iwata S, Bonil- metabolic dilemma. Ann Neurol 1996;40:655– la E, Lach B, et al: Exercise intolerance due to 662. 62 Stanghelle JK, Festvag L, Aksnes AK: Pulmo- mutations in the cytochrome b gene of mito- nary function and symptom-limited exercise chondrial DNA. N Engl J Med 1999;341:1037– 52 Haller RG, Lewis SF, Estabrook RW, DiMauro testing in subjects with late sequelae of polio- 1044. S, Servidei S, Foster DW: Exercise intolerance, myelitis. Scand J Rehab Med 1993;25:125– lactic acidosis, and abnormal cardiopulmonary 129. 41 Dandurand RJ, Matthews PM, Arnold DL, Ei- regulation in exercise associated with adult delman DH: Mitochondrial disease: Pulmo- skeletal muscle cytochrome c oxidase deficien- 63 Willen C, Cider A, Sunnerhagen KS: Physical nary function, exercise performance, and blood cy. J Clin Invest 1989;84:155–161. performance in individuals with late effects of lactate levels. Chest 1995;108:182–189. polio. Scand J Rehab Med 1999;31:244–249. 53 Argov Z, DeStefano N, Taivassalo T, Chen J, 42 Walker UA, Collins S, Byrne E: Respiratory Karpati G, Arnold DL: Abnormal oxidative 64 Halstead LS, Rossi CD: Post-polio syndrome: chain encephalomyopathies: A diagnostic clas- metabolism in exercise intolerance of undeter- results of a survey of 539 survivors. Orthope- sification. Eur Neurol 1996;36:260–267. mined origin. Neuromusc Disord 1997;7:99– dics 1985;8:845–850. 104. 43 Shoffner JM: Metabolic myopathies: Mito- 65 Weinberg J, Borg J, Bevegard S, Sinderby C: chondrial myopathy diagnosis. Neurol Clin 54 Argov Z: Functional evaluation techniques in Respiratory response to exercise in postpolio 2000;18:105–123. mitochondrial disorders. Eur Neurol 1998;39: patients with severe inspiratory muscle dys- 65–71. function. Arch Phys Med Rehabil 1999;80: 44 Vladutiu GD: Metabolic myopathies: The mo- 1095–1100. lecular diagnosis of metabolic myopathies. 55 Nashef L, Lane RJM: Screening for mitochon- Neurol Clin 2000;18:53–104. drial cytopathies: The sub-anaerobic threshold Fernando J. Martinez, MD exercise test (SATET). J Neurol Neurosurg Psy- 1500 E. Medical Center Drive 45 Zeviani M, Tiranti V, Piantadosi C: Reviews in chiatry 1989;52:1090–1094. 3916 Taubman Center molecular medicine: Reviews in mitochondrial Ann Arbor, MI 48109 (USA) disorders. Medicine 1998;77:59–72. 56 Reinhard U, Muller PH, Schmulling RM: De- Tel. +1 734 763 2540, Fax +1 734 936 5048 termination of anaerobic threshold by the ven- E-Mail [email protected] 46 Simon DK, Johns DR: Mitochondrial disor- tilation equivalent in normal individuals. Res- ders: Clinical and genetic features. Annu Rev piration 1979;38:36–42. Med 1999;50:111–127. 57 Finsterer J, Shorny S, Capek J, Cerny-Zachar- 47 Cros D, Palliyath S, DiMauro S, Ramirez C, ias C, Pelzl B, Messner R, Bittner RE, Mamoli Shamsnia M, Wizer B: Respiratory failure re- B: Lactate stress test in the diagnosis of mito- vealing mitochondrial myopathy in adults. chondrial myopathy. J Neurol Sci 1998;159: Chest 1992;101:824–828. 176–180. Cardiopulmonary Exercise Testing 253 for Neuromuscular Disorders

Weisman IM, Zeballos RJ (eds): Clinical Exercise Testing. Prog Respir Res. Basel, Karger, 2002, vol 32, pp 254–263 OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO Role of Cardiopulmonary Exercise Testing in Lung and Heart-Lung Transplantation Trevor J. Williams William R. Slater Lung Transplant Service Medical, Department of Respiratory Medicine, Alfred Hospital, Monash University, Prahran (Melbourne), Australia Summary geon/physiologist Vladimir Demikhov who, having re- turned from WWII, performed a remarkable series of Lung transplantation has evolved over the last 2 decades as experiments in the second half of the 1940s. After initially an important treatment option for severe pulmonary and pul- performing heterotopic heart transplantation he moved monary vascular disease. Initially, survival and duration of sur- on to perform heart-lung transplantation (HLTx) in over vival were the only outcome measures measured. Function out- 50 dogs [1], although the lungs demonstrated the capacity comes including cardiopulmonary exercise testing have only for gas exchange, within 1 week all dogs died because of more recently been reported. Exercise capacity seems to cor- progressive ventilatory failure. These experiments were relate very poorly with measured lung function, but early onset both instructive and remarkable. The recognition that lactic acidosis and premature exercise termination at a substan- complete denervation of both lungs seems to result in pro- tially reduced VO2 peak are seen almost universally. Cardiac gressive respiratory failure led Demikhov to single lung and ventilatory limitation are only sporadically seen. It appears transplantation with the first dog autotransplant surviv- that peripheral muscle dysfunction due to pre- and posttrans- ing 1 month. plant factors is the most frequent cause of exercise limitation. Although clear criteria for exercise testing of lung transplant In 1963, Hardy et al. [2] described the first attempt at recipients have yet to be developed, cardiopulmonary exercise human single lung transplantation (SLTx). Although the testing (CPET) will help clarify the degree and mechanism of man died 18 days later of renal failure, the ability for the exercise limitation, aiding decision making as to the timing of transplanted lung to contribute to gas exchange in the transplantation. Furthermore, posttransplantation the assess- recipient was clearly demonstrated. Through the 1960s ment of vocational and exercise capacity as well as detection of and 1970s, approximately 40 attempts at lung transplan- allograft dysfunction can be greatly aided by CPET. tation were made worldwide, the longest reported surviv- al in this era was a SLTx recipient with silicosis who sur- Introduction vived 10 months [3] spending all but a few weeks in hospi- tal during this time. The 1980s saw the first long-term Solid organ transplantation has been one of the great quality survival in lung transplant recipients. In 1982, the medical miracles of the 20th century. The notion of Stanford University Group led by Reitz et al. [4] reported replacing diseased organs dates back many centuries. The two long-term survivors, one having eventually survived first attempt at lung replacement is attributed to the sur- for 5 years, having received HLTx for pulmonary vascular disease. The following year, the Toronto Lung Transplant Group successfully performed SLTx resulting in almost a decade of quality survival in that recipient [5].

Double lung transplantation was first performed suc- diagnosis with patients with emphysema and cystic fibro- cessfully in 1985 as an ‘en bloc’ procedure [6] although the sis having better survival posttransplant than those with incidence of anastomotic complications was alarmingly pulmonary fibrosis or pulmonary hypertension. high. The procedure was subsequently modified in 1989 to the bilateral (sequential) lung transplant (BLTx) procedure Lung Function utilizing a clam shell incision and usually obviating the Towards the end of the 1980s and early 1990s trans- requirement for cardiopulmonary bypass [7]. Live donor plant groups around the world had sufficient long-term lung transplantation was first performed in the early 1990s survivors that outcomes from lung transplantation in- using the procedure developed by Starnes and co-workers cluding lung function could be reported. These generally at the University of Southern California. It has generally showed a dramatic effect on lung function by lung trans- been performed utilizing a single lobe from each of two plantation with HLTx and BLTx resulting in only mildly living donors which is transplanted ipsilaterally and ortho- restricted spirometry with only a slight reduction in dif- topically into a smaller single recipient [8, 9]. fusing capacity and a normal PA-a O2 gradient. Improve- ment in spirometry occurs during the first 2 years follow- Measurement of Outcomes in Lung ing bilateral lung transplantation unless complications Transplantation supervened [13, 14]. Spirometry following SLTx is almost always abnormal, Survival generally reflecting the underlying pretransplant patholo- In the first two decades of human lung transplantation gy. Patients receiving SLTx for obstructive lung disease from 1963, the length of survival from the point of generally have an FEV1 around 50–65% of predicted but engraftment was the principle and often only outcome with quite a broad range. Generally, for SLTx recipients variable measured. Patient survival immediately postop- diffusing capacity is significantly reduced and PA-a O2 eratively, followed by reports of increasing lengths of time gradient significantly widened. Spirometry improves little out of hospital became the key reported outcome vari- in SLTx recipients beyond 3 months posttransplantation ables. Earlier on, however, reference was made to physio- [13, 14]. Live donor lobar transplantation is typically logical effects including the ability of the transplanted associated with a mild restrictive ventilatory defect and lung to support gas exchange, and interestingly an under- diffusing capacity at the lower end of the normal range [8, standing of potential complications of two different lungs 9]. The spirometry and diffusing capacity reported gener- with quite different mechanical properties as occurs in ally reflect stable patients without significant complica- single lung transplantation for emphysema [10]. tion after lung transplantation. Many factors, however, By the end of the 1980s there had been a dramatic will impact on lung function including the presence of increase in the number of lung transplant procedures pleural complications, diaphragmatic injury or injury to occurring worldwide to the point that meaningful actuar- the lung allograft caused by rejection or allograft infec- ial assessments of survival could be made in large patients tion. The most feared complication in the long-term recip- populations. Two major registries have been reporting ient is the development of chronic rejection in the form of outcomes: the International Society of Heart and Lung bronchiolitis obliterans syndrome (BOS) manifesting as a Transplant Registry and the St Louis International Lung progressive obstructive ventilatory defect. Transplant Registry. Although there were differences in terms of completeness of data in each of the registries, by Functional Class 1995 the Registries were reporting 1- and 5-year survivals Those directly involved in lung transplant programs of 62–78% and 40–50% in lung transplantation and have little doubt of the substantial improvement in func- HLTx [11, 12]. It became evident that two important tional class of successfully transplanted individuals. Nev- trends in survival analysis were occurring. There had been ertheless, data reporting New York Heart Association a progressive improvement in survival virtually com- (NYHA) functional class show all patients are NYHA pletely attributable to improved 90-day survival, without Class III or IV pretransplantation with 70% of patients evidence of improvement in survival in long-term survi- improved to NYHA class 1 and 21% to NYHA class II vors who had already survived the initial posttransplant assessed 2–4 years posttransplant, i.e. 91% were at least period. The second important feature was that survival one NYHA class better [15]. did seem to vary considerably depending on pretransplant More recently, formal measures of quality of life have been reported with Gross et al. [16] showing a substantial Role of Cardiopulmonary Exercise Testing 255 in Lung and Heart-Lung Transplantation

Table1. Exercise responses for single (SLTx), bilateral (BLTx) and heart-lung (HLTx) transplant recipient expressed as mean reported in each study [as referenced] SLTx BLTx HLTx Six-minute walk test 480–670 [51, 52] 600–700 [51, 52] NR 6 MWT, m 30–45 [24, 53–55] 30–46 [24, 53, 54] 30 [54] Incremental CPETx 41–58 [19, 20, 24, 54, 55] 40–59 [19, 20, 24, 54] 38–52 [19, 22, 54] Peak work rate, % predicted legs 1 breathing [14] legs 1 breathing [14] legs 1 breathing [14] Peak VO2, % predicted 69–81 [19, 20, 24, 54, 55] 67–90 [19, 20, 24, 54] 59–83 [19, 22, 54] Symptoms (Borg scale) 60–65 [20, 54] 59–70 [20, 54] 63–64 [22, 54] Peak HR, % predicted 47–63 [19, 20, 24, 55, 56] 33–53 [19, 20, 24, 56] 39 [19] Peak O2 pulse, % predicted 94–97/90–94 [19, 20, 24] 95–97/94–97 [19, 20, 24] NR/97 [19] VEmax/MVV, % 25–41 [54, 55] 31 [54] 18–33 [22, 54] SaO2% rest/peak exercise increased [54, 55] increased [54] increased [22, 54] PAO2 – PaO2 (max ex), mm Hg 31–41 [19, 24] 42–43 [19, 24] 34 [19] VD/VT AT as % of predicted VO2max NR = Not reported. positive impact of lung transplantation on the dimensions Progressive incremental cardiopulmonary exercise of physical functions, health perception, social function testing (CPET) with measurement of ventilation, VO2 and role function using the MOS20 health profile. These and VCO2 have been rarely reported in those awaiting benefits appear to be maintained until BOS supervenes. lung transplantation. Theodore et al. [93] reported results Generally, however, return to work results have been dis- of 10 patients with severe pulmonary vascular disease appointing with Craven et al. [17] reporting only 38% of having a group mean VO2 peak of 9.4 ml/kg/min (24% of recipients returning to paid employment following lung predicted) using a treadmill exercise protocol. This sub- transplantation. stantial degree of exercise limitation has also been con- firmed by two other reports [13, 18] using incremental Six-Minute Walk Test bicycle ergometer protocols. Nevertheless, these were gen- The results of the six-minute walk test have been quite erally patients well enough to participate in CPET proto- extensively reported as an outcome measure following cols without requiring supplemental oxygen and if any- lung transplantation. Generally, patients range from 200 thing may overestimate the maximum exercise capacity to 400 m in 6 min pretransplant, improving to 500–700 m of the group. posttransplant, maintaining this benefit unless serious complications supervene [14]. Due to its simplicity, this A number of authors have now reported a significant test has developed wide currency in reporting outcome reduction in VO2 peak using CPET protocols measured from lung transplantation. approximately 3 months posttransplant (usually following formal rehabilitation) [18–24]. The mean VO2 peak re- Incremental Cardiopulmonary Exercise Testing ported ranged from 14 to 18 ml/kg/min in these studies (a (Table 1) range of 45–52% predicted). Similar results have also Treadmill exercise studies of utilizing the Bruce Proto- been reported in pediatric lung transplant recipients [25]. col have been reported occasionally. Williams et al. [13] A similar degree of exercise limitation seems to occur (de- reported that no patient prior to transplantation com- spite differences in the magnitude of lung function im- pleted Bruce Stage I despite the use of supplemental oxy- provement) in HLTx, BLTx and SLTx recipients [23, 24]. gen. Overall, almost 75% of all recipients achieved Bruce There does not appear to be any systematic impact of pre- Stage II or better posttransplant. The outcome in double- transplant diagnosis on posttransplant exercise capacity lung transplant recipients appeared slightly better than [21]. Stiebellehner et al. [26] have demonstrated that lung single-lung transplant recipients (stage 2.7 B 0.3 vs. transplant recipients are capable of demonstrating signifi- 2.0 B 0.25; mean B SEM). cant (albeit small) responses to supervised training pro- grams. Nevertheless, beyond 3 months, lung function gen- 256 Williams/Slater

erally improves slightly, but exercise capacity seems to Exercise-Induced Hypoxemia improve very little, with no authors reporting a patient Generally, desaturation is seen only in SLTx at peak group returning to normal exercise capacity. exercise [23, 24]. HLTx or BLTx recipients who are free of complications do not desaturate at peak exercise. Any Exercise Responses and Exercise Limitation after desaturation in HLTx or BLTx recipients at peak exercise Lung Transplantation indicates graft dysfunction. In single lung transplant re- cipients, mild desaturation does not indicate allograft dys- Although there are significant differences in terms of function. exercise responses comparing HLTx, BLTx, and SLTx recipients, a number of common features are to be found Ventilation on incremental exercise testing. Resting heart rate is gen- HLTx and BLTx recipients (unless with significant erally high, the anaerobic threshold occurs very early in allograft dysfunction) do not approach ventilatory limita- exercise and patients generally report leg tiredness as the tion at the point of exercise termination. Martinez et al. predominant symptom at exercise termination [23, 24]. [34] studied 7 single lung transplant recipients for ob- Although many factors may impact on oxygen delivery to structive lung disease and showed expiratory flow limita- exercising muscle (e.g. cardiac response, oxygenation of tion utilizing exercise flow-volume studies. Generally, mixed venous blood by the lung, haemoglobin concentra- single lung transplant recipients approach or reach venti- tion), the unifying factor appears to be an abnormality of latory limitation at the point of exercise cessation. oxygen uptake or utilization by peripheral muscles. Exercising Skeletal Muscle Dysfunction Cardiac Response There is now mounting evidence that exercise limita- In HLTx recipients, cardiac denervation is present tion following lung transplantation is due primarily to and, there is at least a theoretical concern of the possibility abnormalities of the exercising skeletal muscles. A num- that abnormal cardiac responses result in exercise limita- ber of abnormalities have now been reported and a com- tion. The denervated heart response to exercise is abnor- bination of pre- and posttransplant factors seems likely. mal with a high resting heart rate, and initial increases in The reduced skeletal muscle oxidative capacity is proba- cardiac output being entirely due to contractility [27]. A bly due to a combination of a number of factors. combination of a low cardiac output at peak exercise and The quadriceps femoris muscle mass is substantially high lactate initially lead to speculation as to the possibili- reduced in severe COPD [35], and this seems to persist ty to cardiac limitation in patients with a transplanted posttransplant. A number of factors may be responsible heart [28]. Kavanagh et al. [29] noted that these cardiac including nutrition [36], the effects of drugs particularly outputs were not measured at the same absolute workload corticosteroids [37], disuse/deconditioning, and endo- when compared to controls, and that arterio-venous oxy- crine effects of severe respiratory disease (e.g. low testos- gen difference was normal suggesting cardiac output and terone levels in males with severe COPD) [38]. oxygen delivery was normal at matched work rate. Subse- Using 31P magnetic resonance spectroscopy, Evans et quently, Jensen et al. [30] using acetylene-breathing meth- al. [39] demonstrated a lower resting skeletal muscle pH ods confirmed a normal relationship between cardiac out- with an earlier and more rapid fall of pH on exercise, sug- put and oxygen consumption in cardiac transplant recipi- gesting early anaerobic metabolism comparing lung trans- ents. The cardiac allograft is susceptible to early ischemic plantation recipients and matched controls. A subsequent injury, rejection or transplant coronary disease which study using near infrared spectroscopy [40] showed that need to be considered in HLTx as cardiac causes of pre- myoglobin oxygen saturation is normal despite very early mature exercise termination. onset of anaerobic metabolism strongly, suggesting that Right-ventricular impairment may contribute to exer- oxygen utilisation rather than delivery was limiting. cise limitation particularly in those with high pulmonary A recently reported study of quadriceps femoris biop- vascular resistance receiving SLTx [31–33]. Nevertheless, sies in stable lung transplant recipients [41] has shown a it appears that cardiac limitation is a rare and sporadic number of abnormalities including a marked reduction in event among lung transplant recipients and does not gen- type 1 fiber proportion as has previously been reported in erally explain the almost universal reduction in exercise patients with severe emphysema [42, 43]. The reduction capacity. in type 1 fiber proportion is very significant (25 vs. 56%) comparing lung transplant recipients to normal sedentary Role of Cardiopulmonary Exercise Testing 257 in Lung and Heart-Lung Transplantation

controls. Marked reduction in the mitochondrial enzymes the anaerobic threshold can be determined. Frequently, glutamate dehydrogenase, citrate synthase and oxygluter- severe exercise-induced hypoxemia leads to premature ate dehydrogenase, as well as 3-hydroxacyl CoA hydro- termination of the CPET. Attempting to supplement oxy- genase (HADH) was also demonstrated. Furthermore, gen during the CPET adds significantly to the complexity marked reduction in mitochondrial ATP production to an of testing and is often quite impractical for some laborato- array of substrates was seen. At rest, muscle fibers had ries. A second important technical point is to make sure higher lactate concentrations and IMP content consistent increments are appropriate – typically 5–10 W/min is with a high reliance on anaerobic metabolism. Whether suitable – as the rate of incremental increase in workload these changes represent a primary muscle defect or simply may impact on final measured Wpeak and VO2peak [49]. the effects of prolonged reduction in activity is unclear. It Generally, CPET is performed to confirm that the degree seems likely that the calcineurin antagonist cyclosporin A of exercise limitation reported by the patient is matched is responsible in part for the marked reduction in mito- by objective measurement. In the illustrative case 1, lung chondrial oxidative capacity as has been demonstrated transplantation when initially referred was likely to have both in vivo and in vitro in rats [44, 45]. Nevertheless, been quite premature, despite the patient’s complaints of they probably do explain the very early rise in plasma lac- severe exercise-induced breathlessness. Indeed at that tate concentration and premature exercise termination in time lung transplantation would likely worsen exercise this group of patients. capacity. Poor exercise capacity relative to measured lung function [50] may define a group at increased risk of dying Other abnormalities in the skeletal muscle of lung on the waiting list, as reported in cystic fibrosis patients. transplant have been suggested. Although potassium ho- meostasis is abnormal [46], no abnormality of Na/K After lung transplantation, a CPET after full rehabili- pump density or activity or Na/K ATPase enzyme activi- tation is very useful in allowing the physician to advise as ty is seen [47]. Sarcoplasmic reticulum function (Ca to exercise capacity and capacity to perform certain voca- release, uptake on Ca ATPase activity) seems reduced tions. The lung function test correlates very poorly with when account is taken of the decreased type I fiber pro- exercise capacity. Again, an appropriately low work rate portion [48]. increment (10–15 W/min) is required to give optimal test duration. This early posttransplant exercise test also pro- In summary, the common factor limiting exercise in vides a useful baseline – subsequent slight changes with stable fully rehabilitated patients following lung trans- respect to exercise-induced desaturation or reduced venti- plantation is leg fatigue. SLTx recipients approach or latory reserve may reflect early dysfunction of the pulmo- reach ventilatory limitation (with mild desaturation) at nary allograft. the point of exercise termination. Stable HLTx and BLTx recipients do not have evidence of ventilatory limitation To help in the understanding of the indications, con- nor desaturation and generally appear limited by their duct, features and interpretation of CPET in the lung exercising muscles. Nevertheless, the unifying factor transplant recipient, 4 illustrative cases are presented. across all lung transplant recipients appears to be a peripheral defect of oxygen utilization is the quadriceps Illustrative Cases femoris muscles. Conduct of Cardiopulmonary Exercise Tests in Case 1: Right SLTx for Idiopathic Pulmonary Fibrosis Lung Transplant Recipients A 49-year-old male presenting with exercise-induced dyspnea. A high resolution CT scan shows basal fibrosis CPET presents significant challenges in patients with with minimal ground glass opacity. Video-assisted thora- severe lung disease both before and after lung transplanta- coscopic lung biopsy confirms a usual interstitial pneu- tion. The exercise responses are almost always abnormal monitis pathological picture. He was commenced on and the focus is generally on assessing the contribution of prednisolone 25 mg daily p.o. and azathioprine 25 mg ventilatory, cardiac and peripheral factors to exercise lim- daily p.o. but his condition slowly deteriorated. In view of itation. symptoms, unfavorable pathology and failure to improve on therapy he was referred for lung transplant assess- Pretransplant CPET is infrequently performed be- ment. cause exercise is inevitably severely limited due to the CPET 1: Two Years Prior to Lung Transplantation. An underlying disease. Exercise is usually terminated before incremental CPET was performed because the degree of 258 Williams/Slater

Table 2. Case 1: right SLTx for idiopathic pulmonary fibrosis CPET 1 CPET 2 CPET 3 CPET 4 Predicted 2 years 2 months 3 months 12 months values pre-SLTx pre-SLTx post-SLTx post-SLTx FEV1.0, liters 3.75 2.40 3.32 2.60 (BOS1) 3.58 VC, liters 4.70 2.78 4.23 3.78 4.85 FER, % 80 86 78 69 175 DLCO, ml/min/mm Hg 16.9 10.2 20.2 15.5 28.3 Wmax, watts 180 115 132 132 158 HR, /min 66 74 96 79 Rest 180 123 155 141 178 Peak VO2/HR, ml/beat 4.2 6.6 4.9 5.3 Rest 15.9 10.8 8.7 11.0 Peak 129 71 96 103 VE peak, liters/min (91) (84) (116) (% of predicted MVV) 28.6 12.5 10.4 14.5 33.1 VO2 at AT, ml/kg/min (90) (41) (34) (48) ! 34 (% of predicted VO2 max at AT) 37.1 17.5 16.4 20.2 VO2peak, ml/kg/min 33 46 48 47 VE/VO2 pre-AT SaO2, % 96 97 100 99 Rest 86 78 97 93 Peak Borg 9 7 8 8 Breathlessness 5 5 5 3 Legs A 49-year-old male who received a right SLTx for idiopathic pulmonary fibrosis. CPET was performed both pre- and posttransplantation. exercise limitation seemed out of keeping with the pa- work and VO2. Breathlessness was the predominant tient’s measured lung function, which showed normal spi- symptom at exercise termination. Heart rate was not lim- rometry with a moderately reduced diffusion capacity. An iting, but minute ventilation approached limiting levels, incremental bicycle ergometer exercise test with 20-W/ with marked desaturation occurring progressively min increments was performed to volitional exertion. throughout the exercise test. Ventilation was clearly ex- Breathlessness was the predominant symptom at exercise cessive below an early anaerobic threshold. The conclu- termination. Above normal work-rate and VO2 was sion drawn was that progression of disease was clearly evi- achieved. Heart rate was limiting and ventilation ap- dent and transplantation was indicated on prognostic proached limiting levels. Mild desaturation was seen pro- grounds, without concern that lung transplantation would gressively through the exercise test. The exercise response negatively impact on exercise capacity. was abnormal but immediate listing for transplantation was deferred because this was likely to result in substan- CPET 3: Three Months after SLTx. At 3 months after tial worsening of exercise capacity. right SLTx, spirometry shows a mild restrictive ventila- tion defect with a mildly impaired diffusion capacity. A CPET 2: Two Months Prior to SLTx. A steady deterio- repeat incremental bicycle ergometer exercise test was ration had occurred over the ensuing 18 months and he performed. Work rate was mildly reduced, but VO2 sub- was actively listed for SLTx 3 months prior to this exer- stantially reduced (53% predicted) at peak exercise. Heart cise test. Spirometry now showed a moderately restrictive rate was not limiting but ventilation exceeded predicted ventilatory defect with marked reduction in diffusion maximum. Very mild desaturation was noted. Excessive capacity. A repeat incremental exercise test with 16.5-W/ ventilation was seen below a very early anaerobic thresh- min increments was performed, demonstrating a low peak old. Anaerobic threshold was even earlier than 2 months Role of Cardiopulmonary Exercise Testing 259 in Lung and Heart-Lung Transplantation

prior to transplantation. Although successfully trans- Table 3. Case 2: SLTx for obstructive lung disease planted, only a slight impact had been made on maximum exercise capacity. Measured Predicted CPET 4: Twelve Months after SLTx. Exercise capacity FEV1.0, liters 1.76 2.43 was reassessed 12 months after right single lung transplan- VC, liters 2.22 3.17 tation. He had developed significant chronic rejection FER, % 79 1 75 (BOS class 1 – BOS 1). Spirometry shows mild mixed DLCO, ml/min/mm Hg 21.1 22.6 obstructive and restrictive ventilatory defect with moder- Wmax, Watts 82.5 112 ately reduced diffusion capacity. A repeat incremental HR, /min exercise test showed mildly reduced work and moderately 82 176 reduced VO2 (61% predicted) at peak exercise. Heart rate Rest 148 was not limiting although ventilation approached limiting Peak 7.6 levels, with modest desaturation noted. Excessive ventila- VO2/HR 5.3 tion was seen prior to a slightly early anaerobic threshold. Rest 62 29.0 The development of BOS 1 had not impacted on his exer- Peak (100) ! 34 cise capacity and a significant improvement in anaerobic VEPeak, /min threshold and VO2 peak was seen from 3 to 12 months (% MVV) 9.3 posttransplant. Interestingly, the VO2 peak at 12 months VO2 at AT, ml/kg/min (40) was only 54% of the VO2 peak 2 years prior to transplan- (% of predicted VO2 max) 16.0 tation, with a substantially earlier anaerobic threshold. VO2peak, ml/kg/min 35 VE/VO2 pre-AT Comment. Transplantation was indicated on prognos- SaO2, % 99 tic grounds and did slightly improve exercise capacity Rest 98 from that measured 2 months pretransplant. Most strik- Peak ing was the low VO2 peak and early anaerobic threshold Borg 6 when compared to those measured 2 years prior to trans- Breathlessness 5 plantation (table 2). Legs A 51-year-old female 3 months after left SLTx for severe obstruc- tive lung disease. Illustrative Case 2 SLTx for Obstructive Lung Disease native lung. The main findings on exercise are typical of History. A 51-year-old lady who presents with a long SLTx, that is an early anaerobic threshold and exercise history of chronic asthma, and who developed severe termination at or near MVV (table 3). fixed airflow obstruction despite appropriate therapy. She received a left SLTx and made an uneventful postopera- Illustrative Case 3 BLTx for Cystic Fibrosis tive recovery, completing 2 months of supervised rehabil- History. A 23-year-old lady presenting with a steady itation. deterioration in lung function and recurrent admissions CPET at Three Months Posttransplant. Spirometry due to exacerbations of cystic fibrosis. She received BLTx shows a mild-moderate restrictive ventilatory defect, with and had an uneventful postoperative course. She com- evidence of minimal airflow obstruction on tests of small menced supervised rehabilitation 3 days per week at 2 airway function. Diffusion capacity is normal. An incre- weeks posttransplantation, continuing to 3 months post- mental bicycle ergometer exercise test was performed transplantation. (16.5-W/min increments) to volitional exhaustion. A CPET at Three Months after BLTx. Spirometry and mildly reduced work rate and moderately reduced VO2 diffusion capacity are normal. An incremental bicycle peak were seen. Exercise termination was predominantly exercise test using 10-W/min increments to volitional due to leg tiredness. Heart rate was not limiting but venti- exhaustion. Leg tiredness was the predominant symptom lation reached 100% predicted at exercise termination, at exercise termination. A low peak work and peak VO2 although no significant desaturation occurred. Excessive (54% predicted) were achieved. Neither cardiac nor respi- ventilation was seen below a very early anaerobic thresh- ratory function appeared limiting. Anaerobic threshold old. was reached very early in exercise, although below this Comment. The spirometry is somewhat atypical as the ventilation appeared normal in relation to VO2. predominant abnormality is a restrictive ventilatory de- fect despite the substantial airflow obstruction in the 260 Williams/Slater

Table 4. Case 3: BLTx for cystic fibrosis Table 5. Case 4: HLTx for VSD/pulmonary hypertension Measured Predicted Measured Predicted FEV1.0, liters 3.36 3.31 FEV1.0, liters 4.28 4.53 VC, liters 3.45 3.78 VC, liters 5.17 5.39 FER, % 97 1 75 FER, % 83 1 75 DLCO, ml/min/mm Hg 22.3 26.5 DLCO, ml/min/mm Hg 23.7 33.4 Wmax, Watts 100 158 Wmax, Watts 116 247 HR, /min HR, /min 77 195 100 196 Rest 162 Rest 130 Peak 118 Peak 150 VO2/HR, ml/beat 5.2 VO2/HR, ml/beat 5.9 Rest 6.8 39.0 Rest 11.2 49.0 Peak 50 ! 34 Peak 66 ! 34 VEPeak, liters/min 14.7 VEPeak, liters/min 16.4 VO2 at AT, ml/kg/min (37) VO2 at AT, ml/kg/min (30) (% of predicted VO2 max) 21.0 (% of predicted VO2 max) 25.0 VO2peak, ml/kg/min 30 VO2peak, ml/kg/min 32 VE/VO2 pre-AT VE/VO2 pre-AT SaO2, % 96 SaO2, % 97 Rest 97 Rest 97 Peak Peak Borg 5 Borg 3 Breathlessness 7 Breathlessness 7 Legs Legs A 23-year-old female, 3 months after BLTx for cystic fibrosis. A 21-year-old man, fully recovered from a HLTx for ventricular septal defect and pulmonary hypertension tested 18 months post- transplant. Comment. This is a typical response to incremental Neither ventilatory limitation nor desaturation were evi- exercise in a BLTx recipient. Exercise terminated well dent at peak exercise. Ventilation was not excessive in short of cardiac or ventilatory limitation due to leg tired- relation to VO2 below a very early anaerobic threshold. ness. The anaerobic threshold occurred very early in exer- cise (table 4). Comment. Despite a functionally excellent result from HLTx, substantially reduced maximum exercise capacity, Illustrative Case 4 HLTx for VSD/Pulmonary presumed due to impaired oxygen uptake or utilization, is Hypertension evident. A high heart rate at rest and blunted response to History. A 21-year-old male with severe pulmonary exercise is seen due to cardiac denervation (table 5). hypertension secondary to a ventricular septal defect who was listed for transplantation due to progressive right ven- Conclusion tricular failure. He received a HLTx and recovered rapid- ly and uneventfully. A CPET was performed 18 months Lung transplantation has evolved to the point where posttransplant having returned to full time employment survival is expected and we are now focusing on the quali- as a professional fisherman 12 months previously. ty of outcome. Substantial exercise limitation is present CPET at 18 Months after HLTx. Spirometry is normal on CPET both before and after lung transplantation. The and diffusion capacity mildly reduced. An incremental mechanism of exercise limitation changes from ventilato- bicycle ergometer exercise test was performed with 16.5- ry limitation pretransplantation to peripheral limitation W/min increments. Leg tiredness was the predominant posttransplantation. CPET will help clarify the degree symptom at exercise termination. A high resting heart and mechanism of exercise limitation where a discrepan- rate with blunted chronotropic response to exercise is cy between lung function and exercise capacity is sus- seen, but no evidence of cardiac limitation was evident. pected or expected. Role of Cardiopulmonary Exercise Testing 261 in Lung and Heart-Lung Transplantation

References 1 Shumacker HBJ: A surgeon to remember: 17 Craven JL, Bright J, Dear CL: Psychiatric, psy- 31 Carere R, Patterson GA, Liu P, Williams T, Notes about Vladimir Demikhov. Ann Thorac chosocial, and rehabilitative aspects of lung Maurer J, Grossman R: Right and left ventricu- Surg 1994;58:1196–1198. transplantation. Clin Chest Med 1990;11:247– lar performance after single and double lung 257. transplantation. The Toronto Lung Transplant 2 Hardy JD, Webb WR, Dalton ML, Walker GR: Group. J Thorac Cardiovasc Surg 1991;102: Lung homotransplantation in man. JAMA 18 Ross DJ, Waters PF, Mohsenifar Z, Belman 115–122. 1963;186:1065–1074. MJ, Kass RM, Koerner SK: Hemodynamic responses to exercise after lung transplantation. 32 Globits S, Burghuber OC, Koller J, Schenk P, 3 Derom F, Barbier F, Ringoir S, Versieck J, Rol- Chest 1993;103:46–53. Frank H, Grimm M, et al: Effect of lung trans- ly G, Berzsenyi G, et al: Ten-month survival plantation on right and left ventricular volumes after lung homotransplantation in man. J Tho- 19 Levy RD, Ernst P, Levine SM, Shennib H, and function measured by magnetic resonance rac Cardiovasc Surg 1971;61:835–846. Anzueto A, Bryan CL, et al: Exercise perfor- imaging. Am J Respir Crit Care Med 1994;149: mance after lung transplantation. J Heart Lung 1000–1004. 4 Reitz BA, Wallwork JL, Hunt SA, Pennock JL, Transplant 1993;12:27–33. Billingham ME, Oyer PE, et al: Heart-lung 33 Katz WE, Gasior TA, Quinlan JJ, Lazar JM, transplantation: Successful therapy for patients 20 Miyoshi S, Trulock EP, Schaefers HJ, Hsieh Firestone L, Griffith BP, et al: Immediate ef- with pulmonary vascular disease. N Engl J Med CM, Patterson GA, Cooper JD: Cardiopulmo- fects of lung transplantation on right ventricu- 1982;306:557–564. nary exercise testing after single and double lar morphology and function in patients with lung transplantation. Chest 1990;97:1130– variable degrees of pulmonary hypertension. J 5 Unilateral lung transplantation for pulmonary 1136. Am Coll Cardiol 1996;27:384–391. fibrosis. Toronto Lung Transplant Group. N Engl J Med 1986;314:1140–1145. 21 Orens JB, Becker FS, Lynch JP, III, Christen- 34 Martinez FJ, Orens JB, Whyte RI, Graf L, sen PJ, Deeb GM, Martinez FJ: Cardiopulmo- Becker FS, Lynch JP III: Lung mechanics and 6 Patterson GA, Cooper JD, Goldman B, Weisel nary exercise testing following allogeneic lung dyspnea after lung transplantation for chronic RD, Pearson FG, Waters PF, et al: Technique transplantation for different underlying disease airflow obstruction. Am J Respir Crit Care of successful clinical double-lung transplanta- states. Chest 1995;107:144–149. Med 1996;153:1536–1543. tion. Ann Thorac Surg 1988;45:626–633. 22 Theodore J, Morris AJ, Burke CM, Glanville 35 Bernard S, Whittom F, Leblanc P, Jobin J, Bel- 7 Pasque MK, Cooper JD, Kaiser LR, Haydock AR, VanKessel A, Baldwin JC, et al: Cardio- leau R, Berube C, et al: Aerobic and strength DA, Triantafillou A, Trulock EP: Improved pulmonary function at maximum tolerable training in patients with chronic obstructive technique for bilateral lung transplantation: constant work rate exercise following human pulmonary disease. Am J Respir Crit Care Med Rationale and initial clinical experience. Ann heart-lung transplantation. Chest 1987;92: 1999;159:896–901. Thorac Surg 1990;49:785–791. 433–439. 36 Kelsen SG, Ference M, Kapoor S: Effects of 8 Barr ML, Schenkel FA, Cohen RG, Chan KM, 23 Williams TJ, Howard M, Roget J, Esmore DS: prolonged undernutrition on structure and Barbers RG, Marboe CC, et al: Living-related Exercise capacity after combined heart-lung function of the diaphragm. J Appl Physiol lobar transplantation: Recipient outcome and transplantation. Transplant Proc 1992;24: 1985;58:1354–1359. early rejection patterns. Transplant Proc 1995; 2018. 27:1995–1996. 37 Decramer M, Lacquet LM, Fagard R, Rogiers 24 Williams TJ, Patterson GA, McClean PA, Za- P: Corticosteroids contribute to muscle weak- 9 Starnes VA, Barr ML, Cohen RG, Hagen JA, mel N, Maurer JR: Maximal exercise testing in ness in chronic airflow obstruction. Am J Res- Wells WJ, Horn MV, et al: Living-donor lobar single and double lung transplant recipients. pir Crit Care Med 1994;150:11–16. lung transplantation experience: Intermediate Am Rev Respir Dis 1992;145:101–105. results. J Thorac Cardiovasc Surg 1996;112: 38 Kamischke A, Kemper DE, Castel MA, Luthke 1284–1290. 25 Nixon PA, Fricker FJ, Noyes BE, Webber SA, M, Rolf C, Behre HM, et al: Testosterone levels Orenstein DM, Armitage JM: Exercise testing in men with chronic obstructive pulmonary 10 Stevens PM, Johnson PC, Bell RL, Beall AC Jr, in pediatric heart, heart-lung, and lung trans- disease with or without glucocorticoid therapy. Jenkins DE: Regional ventilation and perfu- plant recipients. Chest 1995;107:1328–1335. Eur Respir J 1998;11:41–45. sion after lung transplantation in patients with emphysema. N Engl J Med 1970;282:245– 26 Stiebellehner L, Quittan M, End A, Wieseltha- 39 Evans AB, Al Himyary AJ, Hrovat MI, Pap- 249. ler G, Klepetko W, Haber P, et al: Aerobic pagianopoulos P, Wain JC, Ginns LC, et al: endurance training program improves exercise Abnormal skeletal muscle oxidative capacity 11 St.Louis International Lung Transplant Regis- performance in lung transplant recipients. after lung transplantation by 31P-MRS. Am J try Report: St. Louis, Washington University Chest 1998;113:906–912. Respir Crit Care Med 1997;155:615–621. School of Medicine, 1996. 27 Pope SE, Stinson EB, Daughters GT, Schroe- 40 Tirdel GB, Girgis R, Fishman RS, Theodore J: 12 Hosenpud JD, Novick RJ, Breen TJ, Keck B, der JS, Ingels NB Jr, Alderman EL: Exercise Metabolic myopathy as a cause of the exercise Daily P: The Registry of the International Soci- response of the denervated heart in long-term limitation in lung transplant recipients. J Heart ety for Heart and Lung Transplantation: cardiac transplant recipients. Am J Cardiol Lung Transplant 1998;17:1231–1237. Twelfth official report – 1995. J Heart Lung 1980;46:213–218. Transplant 1995;14:805–815. 41 Wang XN, Williams TJ, McKenna MJ, Li JL, 28 Savin WM, Haskell WL, Schroeder JS, Stinson Fraser SF, Side EA, et al: Skeletal muscle oxi- 13 Williams TJ, Grossman RF, Maurer JR: Long- EB: Cardiorespiratory responses of cardiac dative capacity, fiber type, and metabolites af- term functional follow-up of lung transplant transplant patients to graded, symptom-limited ter lung transplantation. Am J Respir Crit Care recipients. Clin Chest Med 1990;11:347–358. exercise. Circulation 1980;62:55–60. Med 1999;160:57–63. 14 Williams TJ, Snell GI: Early and long-term 29 Kavanagh T, Yacoub MH, Mertens DJ, Ken- 42 Jakobsson P, Jorfeldt L, Brundin A: Skeletal functional outcomes in unilateral, bilateral, nedy J, Campbell RB, Sawyer P: Cardiorespira- muscle metabolites and fibre types in patients and living-related transplant recipients. Clin tory responses to exercise training after ortho- with advanced chronic obstructive pulmonary Chest Med 1997;18:245–257. topic cardiac transplantation. Circulation disease (COPD), with and without chronic re- 1988;77:162–171. spiratory failure. Eur Respir J 1990;3:192– 15 Pasque MK, Trulock EP, Cooper JD, Trianta- 196. fillou AN, Huddleston CB, Rosenbloom M, et 30 Jensen RL, Yanowitz FG, Crapo RO: Exercise al: Single lung transplantation for pulmonary hemodynamics and oxygen delivery measure- 43 Jobin J, Maltais F, Doyon JF, Leblanc P, hypertension: Single institution experience in ments using rebreathing techniques in heart Simard PM, Simard AA, et al: Chronic obstruc- 34 patients. Circulation 1995;92:2252–2258. transplant patients. Am J Cardiol 1991;68: tive pulmonary disease: Capillarity and fiber- 129–133. type characteristics of skeletal muscle. J Car- 16 Gross CR, Savik K, Bolman RM III, Hertz MI: diopulm Rehabil 1998;18:432–437. Long-term health status and quality of life out- comes of lung transplant recipients. Chest 1995;108:1587–1593. 262 Williams/Slater

44 Hokanson JF, Mercier JG, Brooks GA: Cyclo- 49 Debigare R, Maltais F, Mallet M, Casaburi R, 54 Schwaiblmair M, Reichenspurner H, Muller C, sporine A decreases rat skeletal muscle mito- Leblanc P: Influence of work rate incremental Briegel J, Furst H, Groh J, et al: Cardiopulmo- chondrial respiration in vitro. Am J Respir Crit rate on the exercise responses in patients with nary exercise testing before and after lung and Care Med 1995;151:1848–1851. COPD. Med Sci Sports Exerc 2000;32:1365– heart-lung transplantation. Am J Respir Crit 1368. Care Med 1999;159:1277–1283. 45 Mercier JG, Hokanson JF, Brooks GA: Effects of cyclosporine A on skeletal muscle mitochon- 50 Nixon PA, Orenstein DM, Kelsey SF, Doer- 55 Systrom DM, Pappagianopoulos P, Fishman drial respiration and endurance time in rats. shuk CF: The prognostic value of exercise test- RS, Wain JC, Ginns LC: Determinants of ab- Am J Respir Crit Care Med 1995;151:1532– ing in patients with cystic fibrosis. N Engl J normal maximum oxygen uptake after lung 1536. Med 1992;327:1785–1788. transplantation for chronic obstructive pulmo- nary disease. J Heart Lung Transplant 1998;17: 46 Hall MJ, Snell GI, Side EA, Esmore DS, Wal- 51 Cooper JD, Patterson GA, Trulock EP: Results 1220–1230. ters EH, Williams TJ: Exercise, potassium, and of single and bilateral lung transplantation in muscle deconditioning post-thoracic organ 131 consecutive recipients. Washington Uni- 56 Lands LC, Smountas AA, Mesiano G, Brosseau transplantation. J Appl Physiol 1994;77:2784– versity Lung Transplant Group. J Thorac Car- L, Shennib H, Charbonneau M, et al: Maximal 2790. diovasc Surg 1994;107:460–470. exercise capacity and peripheral skeletal mus- cle function following lung transplantation. J 47 Williams TJ, Fraser SF, McKenna MJ, Li JL, 52 Grossman RF, Frost A, Zamel N, Patterson Heart Lung Transplant 1999;18:113–120. Wang XN, Carey MF, et al: Skeletal muscle GA, Cooper JD, Myron PR, et al: Results of sodium-potassium pump activity is normal single-lung transplantation for bilateral pulmo- Dr. Trevor J. Williams post lung transplant. Am J Respir Crit Care nary fibrosis. The Toronto Lung Transplant Department of Respiratory Medicine Med 1996;153:A828. Group. N Engl J Med 1990;322:727–733. Alfred Hospital Prahran (Melbourne) 3181 (Australia) 48 Li JL, McKenna MJ, Wang XN, Fraser SF, 53 Lands LC, Smountas AA, Mesiano G, Brosseau Tel. +61 3 9276 2489, Fax +61 3 9276 3434 Carey MF, Side EA, et al: Low skeletal muscle L, Shennib H, Charbonneau M, et al: Maximal E-Mail [email protected] sarcoplasmic reticulum calcium release post exercise capacity and peripheral skeletal mus- lung transplantation. Am J Respir Crit Care cle function following lung transplantation. J Med 1996;153:A828. Heart Lung Transplant 1999;18:113–120. Role of Cardiopulmonary Exercise Testing 263 in Lung and Heart-Lung Transplantation

Weisman IM, Zeballos RJ (eds): Clinical Exercise Testing. Prog Respir Res. Basel, Karger, 2002, vol 32, pp 264–272 OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO Exercise Responses in Systemic Conditions Obesity, Diabetes, Thyroid Disorders, and Chronic Fatigue Syndrome Kathy E. Sietsema UCLA School of Medicine, Harbor-UCLA Medical Center, Torrance, Calif., USA Summary patients. Each, however, is associated with exercise intol- erance, and may be represented among patients evaluated A wide variety of systemic conditions affect exercise capaci- for exertional dyspnea or fatigue. In addition, some are ty and may present with exercise intolerance. Obesity reduces common co-existing conditions that may complicate the exercise tolerance principally through the increased metabolic clinical evaluation of a patient with other diagnoses, or demands of ambulation, but may also be associated with altera- the interpretation of his or her exercise test. It is therefore tions of ventilatory function, and with circulatory impairment useful to consider what is known regarding how these con- which is often linked to associated hypertension. Diabetes ditions may themselves modify exercise responses. A alters the function of multiple organ systems by virtue of vascu- comprehensive review of the effects of these conditions lar, neuropathic and metabolic mechanisms. Thyroid hormone on exercise function is beyond the scope of this chapter, affects function of virtually every organ system and either which will focus primarily on those findings that are most excess or deficiency may result in impairment of circulatory likely to be reflected in gas exchange and other measure- responses and alteration of skeletal muscle structure and func- ments made during clinical cardiopulmonary exercise tion. The physiologic processes involved in chronic fatigue syn- testing. Where possible, recently published reviews are drome are less well understood, but exercise intolerance is identified for more detailed reference. universal in this condition and significant reductions in exercise capacity may be present. Understanding the effects of these Obesity diverse conditions on metabolic and organ system function is essential for recognizing their expression in parameters de- Obesity is an increasingly prevalent problem with pro- rived from cardiopulmonary exercise testing. found health effects. Aside from the well-recognized asso- ciations with diabetes, atherosclerotic disease and hyper- Introduction tension, obesity may have direct effects on organ system functions that affect responses to exercise testing. In gen- Several distinct systemic conditions are addressed in eral, these are increasingly evident with increasing severi- the following discussion. Exercise testing is not performed ty of obesity, so not all processes identified in this discus- for primary diagnosis of any of these conditions, nor are sion apply to all obese patients. Obesity refers to an excess there established recommendations or guidelines for use of fat, and therefore is most meaningfully defined and of exercise data in clinical decision making for affected quantified by measures of the mass and distribution of adipose tissue. Measures of body composition are not uni-

versally and readily available in all clinical settings, how- of obesity [11]. In contrast to the concentric hypertrophy ever, and therefore obesity is often inferred from the rela- of hypertension, uncomplicated obesity is characterized tionship of weight to height, such as by the body mass by eccentric hypertrophy [8]. Left-ventricular filling pres- index (BMI, weight in kg/height in m2), or percent of ideal sures are reported to be high–normal at rest in severely weight derived from nomograms. While these approaches obese patients, and increase further with exercise, imply- do not actually distinguish fat from lean mass, they are ing associated pulmonary congestion and/or edema [9]. generally useful for describing different magnitudes of Exercise blood pressure is higher in obese compared to obesity. Obesity also raises important issues about the lean subjects, even among those who are normotensive at expression and interpretation of exercise data for address- rest [12]. Changes in right-ventricular structure and func- ing clinical questions. tion occur when alveolar hypoventilation or sleep apnea syndromes complicate obesity, but are beyond the scope Physiologic Effects of Obesity Relevant to of this discussion. Many of the cardiovascular changes Cardiopulmonary Exercise Testing described in obesity appear to be reversible with weight It is well-recognized that obesity can affect pulmonary loss [9, 12]. function [1]. Respiratory system compliance is reduced by the increased mass of the chest wall and resistance to Metabolic effects of obesity include an increase in rest- descent of the diaphragm represented by increased ab- ing metabolic rate due to the added metabolic activity of dominal mass [1]. Work of breathing is therefore in- the fat mass [13]. The metabolic cost of ambulation is creased [2]. Lung volumes may be reduced as a conse- increased to an even greater degree due to the work quence, and increase towards normal following weight required to support the added body mass. From well- loss [3]. Some investigators have found, at least among established principles of exercise training, this situation young adult subjects, that reduction in vital capacity or might be expected to result in increased exercise capacity total lung capacity is uncommon unless obesity is severe, for obese subjects due to the chronic training effects of the i.e. a weight (kg) to height (cm) ratio of 1.0 or greater, cor- metabolic requirements of doing ambulatory activities. responding roughly to weights at least double the ideal Indeed, modeling of data from healthy populations of value. Functional residual capacity is reduced with much varying body composition suggests that an increase in fat lesser degrees of obesity [4], however, and correlates nega- mass is associated with compensatory changes in muscle tively with the extent of overweight. Residual volume is mass to support the added weight, such that peak VO2 preserved, so expiratory reserve volume is reduced, also values normalized to lean mass are similar to those of lean in proportion to the degree of obesity. Hypoxemia, or subjects [14]. Measured fat-free mass is thus a better pre- increased P(A-a)O2 difference is a recognized conse- dictor of maximal VO2 than is total mass, although the quence of obesity [5], even in the absence of obesity relationship between mass and exercise capacity is likely hypoventilation or sleep disordered breathing, and is more complex than the linear relationships often em- attributed to ventilation perfusion mismatching due to ployed in clinical practice [15, 16]. Other metabolic fac- relative underventilation of the lung bases. Hypoxemia is tors relevant to obesity include insulin resistance, which greater in the supine position, and may normalize with not only affects carbohydrate metabolism, but is also asso- upright posture [6]. ciated with impairment in peripheral vascular reactivity. An association between heart failure and obesity has been noted for many years, and strong epidemiologic rela- Findings on Cardiopulmonary Exercise Testing tionships exist between obesity, hypertension, and hyper- Related to Obesity tensive heart disease. Alterations in cardiac structure and Resting VO2 is increased because of increased meta- function, including increased left-ventricular mass and bolically active tissue mass [13]. In the initial transition reductions in parameters of systolic and diastolic function from rest to exercise there is also an exaggerated increase have also been reported in obese subjects who are normo- in VO2 compared to lean subjects, due to the greater work tensive, however [7]. An intrinsic cardiomyopathy of obe- involved in supporting and propelling the weight (on a sity has thus been proposed, the characteristics of which treadmill) or simply lifting large legs against gravity (on a have recently been reviewed [8]. Obese subjects have an cycle ergometer) [17]. Once these initial high metabolic expanded blood volume, even in the absence of heart fail- demands are taken into account, however, the oxygen cost ure, and resting cardiac output is increased [9, 10]. Left- of further increases in work rate follows the same slope in ventricular mass is increased and correlates with severity obese as in lean subjects [18]. That is to say, the slope of the relationship between VO2 and work rate is normal, as Exercise Responses in Systemic Conditions 265

long as data from the initial rest to exercise transition are The widely used convention of indexing peak VO2 to excluded. Any apparent reductions in mechanical effi- body weight results in values that are low, even for dis- ciency, reflected in a greater than normal slope, are likely ease-free obese subjects. Distinction thus need be made due to ergonomic factors related to work expended in pos- between the use of peak VO2 as an expression of cardio- tural control [19], or excessive work of breathing associat- vascular capacity and its use to describe work capacity. ed with high levels of ventilation near peak exercise [17], The former can most meaningfully expressed as peak VO2 as mechanical efficiency of skeletal muscle in obese sub- per lean mass, or as a percent of a reasonably calculated jects is similar to that of lean subjects [20]. predicted value, while the latter might be expressed as peak VO2 indexed to actual body weight, as this reflects As noted above, peak VO2 in obese subjects has a simi- capacity for ambulatory activity. The choice of expression lar relationship to lean body mass as in normal subjects, therefore depends on whether the clinical question relates although the lean mass may in fact be increased to some to physiologic capacity of the organism as a whole, or the degree in response to the chronic training effect of carry- capacity for doing external work above and beyond the ing excess weight. Among the practical issues of exercise work involved in supporting the organism. testing of obese individuals, then, are how to predict what maximal VO2 should be, and how to express the measured Ventilatory patterns may be altered by obesity. The values. Commonly used predicting equations are based lower end-expiratory lung volume at rest leaves less room on regression analyses of normal subjects who were, for for reducing end expiratory volumes during exercise, and, the most part, nonobese. Weight is a positive factor in indeed, moderately obese subjects are found to not de- these equations and using the obese patient’s actual crease end-expiratory volumes during exercise, in con- weight for this may result in inappropriately high pre- trast to nonobese subjects [24]. This being the case, dicted values. Substituting ‘ideal’ weight, calculated from increases in tidal volume can only be accomplished by height, in these calculations results in lower predicted val- increasing inspiratory volumes, and this response appears ues, but these may be inappropriately low as they do not attenuated, perhaps due to the work of breathing associat- take into account the expected normal training response ed with increasing tidal volume in the face of decreased that the ambulatory obese subject should derive from thoracic compliance. Not surprisingly, such limitation in being overweight. Because of this a hybrid approach has tidal volume are more evident in patients with truncal as been recommended in which ideal weight is incorporated compared to peripheral obesity [25]. Although reduced into established formulae, but the calculated value is then tidal volumes could lead to a higher VD/VT in obese sub- increased by 6 ml/min for each kg of excess weight [21]. jects, ventilatory equivalents at rest and exercise at least This value was derived from analysis of the excess VO2 for moderately obese subjects are reported to be similar to required for unloaded cycling for obese subjects [22] rath- those of lean subjects during upright exercise [17, 20]. In er than direct observations of peak VO2 values. It was contrast to most patients with intrinsic lung disease, obese consistent, however, with the small increase in peak VO2 patients exhibit a decrease in abnormalities of oxygena- above that predicted from ideal weight among subjects tion during exercise compared to rest [6]. This has been included in an analysis by the same investigators of 77 attributed to improved ventilation-perfusion matching apparently healthy men [23]. None of the men in that due to increased tidal volumes and reversal of basal atel- analysis were severely obese, and the validity of this ectasis which is more prominent at rest or in the supine approach has not been assessed in larger populations, nor position. specifically in more severely obese individuals, or in women. Systemic blood pressure is higher during exercise in obese compared to lean subjects, even for subjects who From the above it follows that absolute values of both are normotensive at rest [12]. There do not appear to be peak VO2 and the VO2 at the lactate threshold should be systematic effects of obesity on the heart rate response to at least as high, arguably higher, in an obese subject com- exercise, although heart rate may well be high relative to pared to an otherwise identical lean individual. However, work rate, commensurate with the increased metabolic both the weight indexed VO2 values for these parameters, rate. In the presence of heart failure, whether attributable and the work rates associated with them, will likely be to longstanding hypertension, to hypoxema-induced lower in obese subjects. Obese children, for example, have right-sided heart failure resulting from hypoventilation or a greater absolute peak VO2, but lower peak VO2/kg, than sleep apnea syndromes, or to cardiovascular effects intrin- their lean contemporaries [14]. This raises the issue of sic to obesity itself, all of the findings characteristic of how exercise values should be reported for obese persons. heart failure on exercise testing may be expected. 266 Sietsema

Diabetes Findings on Cardiopulmonary Exercise Testing Related to Diabetes Diabetes is an important risk factor for a number of Some investigators have reported that physically ac- diseases which independently limit exercise function, in- tive diabetics without secondary complications may have cluding coronary artery disease, peripheral arterial dis- normal peak work capacity [36] or peak VO2 [37] com- ease, and chronic renal insufficiency. Each of these im- pared to healthy subjects with similar activity patterns. In pairs exercise function, and are addressed in other sections many series, however, especially when not matched for of this book. Even in the absence of overt secondary dis- high levels of physical activity, diabetics have reduced ease processes, however, diabetics may have altered organ peak exercise VO2 compared to age and gender matched system function that could affect exercise responses. Auto- healthy controls [38–42]. Greater reductions in peak VO2 nomic control is impaired by diabetic neuropathy. Altered are reported for diabetics with evidence of autonomic insulin sensitivity affects peripheral vascular resistance neuropathy than for those without [37, 43]. Patients with and metabolic responses. Microvascular changes associat- neuropathy have lower peak heart rates and lower exer- ed with diabetes can have adverse effects on myocardial cise blood pressure, and do not increase cardiac output function, and pulmonary or peripheral gas exchange. normally at peak exercise [43]. Correlation has been iden- Though some of these processes may differ systematically tified between reduction of peak VO2 and the presence between type I and type II diabetics, there are more simi- and severity of diabetic retinopathy and urinary albumin larities than differences between these groups with respect excretion, implying a relationship to microvascular dis- to acute responses to short term exercise, and the following ease [44]. Thus, both autonomic control of circulation and discussion is based on observations from both. microvascular disease are implicated in exercise impair- ment in diabetes. In addition to a reduction in peak VO2, Physiologic Effects of Diabetes Relevant to diabetics are reported to have slower adaptation to sub- Cardiopulmonary Exercise Testing maximal exercise [45], more shallow increase in VO2 rela- Cardiac dysfunction is recognized in diabetics without tive to work rate during graded testing [38, 40] and pro- evidence of obstructive coronary artery disease [26, 27]. longed time course of recovery of heart rate following Diabetic cardiomyopathy is characterized by impaired maximal exercise [36]. Despite reported decreases in spi- left-ventricular diastolic relaxation, which may be identi- rometric volumes and diffusing capacity [34, 35, 42], fied in a high proportion of diabetics [28, 29], and occurs adverse effects of diabetes on the efficiency of pulmonary early in the course of the disease [30]. Diastolic impair- gas exchange or arterial blood gases during exercise have ment is identified even in the absence of hypertension or not been reported [42, 46]. systolic dysfunction [28, 30], but the extent of dysfunction On cardiopulmonary exercise testing therefore diabet- is increased with co-existing systemic hypertension [31]. ics may demonstrate low values for peak VO2, peak work Among diabetics, the presence and degree of diastolic dys- rate, and peak heart rate compared with predicted. A function has been correlated with reduced exercise capac- reduction in peak heart rate may be low due to cardiovas- ity [29]. Although resting systolic function is generally cular autonomic dysfunction and thus cannot be taken to normal in asymptomatic diabetics without clinical evi- indicate submaximal effort on testing. For the same rea- dence of coronary disease, impaired systolic responses to son, heart rate responses may not be appropriate bench- exercise are also reported [32]. This may be attributable to marks for exercise prescription in this population. Other myocardial adrenergic denervation [32] or to microvascu- indices sensitive to cardiovascular response, including the lar disease. There are some data to indicate that cardiac peak O2 pulse and the anaerobic threshold may be function in uncomplicated diabetics correlates with the reduced to the extent that cardiac output is impaired. A adequacy of glycemic control [26, 33]. lower than average slope of VO2 relative to work rate may Reductions in spirometric volumes are reported with be observed, and may reflect prolonged dynamic re- diabetes, and correlate with the duration of diabetes when sponses. Reductions of the slope to values below the nor- other factors are controlled for [34, 35]. Alterations in mal range likely reflect impairment in cardiovascular insulin sensitivity and carbohydrate metabolism due to responses. The ventilatory response to exercise, the venti- diabetes are clearly important to sustained exercise dur- latory equivalents, and arterial blood gas values should be ing which substrate availability is at issue, but of lesser similar to normal unless there is additional co-existing importance in the context of the short protocols used in pulmonary disease. clinical cardiopulmonary exercise testing. Exercise Responses in Systemic Conditions 267

Thyroid Disorders of global ventricular function may remain normal until hypothyroidism is severe [55, 56]. Additional impairment Both hyperthyroidism and hypothyroidism are identi- in cardiac function in hypothyroidism may be attribut- fied as causes of exertional dyspnea. In neither case are able to altered cardiac loading conditions [55] due to the underlying mechanisms entirely clear. Thyroid hor- increased systemic vascular resistance and decreased mone has diverse actions, affecting virtually every organ blood volume [47]. Skeletal muscle oxidative enzyme system, however, including effects on cardiac and skeletal capacity is reduced markedly in hypothyroidism [57] and muscle, vascular tone, and nervous system responses, contractile properties of the muscle are impaired. Exer- each of which are relevant to exercise function. cise of small muscle masses is associated with higher blood lactate levels [58] and altered phosphocreatine Physiologic Effects of Thyroid Disorders Relevant to responses [59] for patients with hypothyroidism com- Cardiopulmonary Exercise Testing pared to euthyroid controls. Overt skeletal muscle myopa- Hyperthyroidism is associated with elevations in meta- thy, manifest as proximal muscle weakness, pain and bolic rate, with attendant increases in resting values of cramping [60] is a well-recognized manifestation of hypo- pulmonary gas exchange, ventilation, and cardiac output. thyroidism. Respiratory as well as locomotor muscles [61] Conversely, these variables are decreased in hypothyroid- may be affected in hypothyroidism. In addition to re- ism, although the changes may be more difficult to detect, duced respiratory muscle strength, abnormalities of respi- at least at rest. Cardiovascular effects of thyroid disorders ratory control are also recognized in thyroid deficiency have recently been reviewed [47]. [62], and either may result in reduced ventilatory volumes In hyperthyroidism, cardiac output is increased dis- and hypercapnia. proportionately to the increased metabolic rate due to the added effects of reduced vascular resistance, increased Consistent with the preceding observations, hypothy- blood volume, increased heart rate and enhancement of roidism appears to impair exercise performance more both contractility and diastolic relaxation of the heart substantially than hyperthyroidism. Baldwin et al. [57] [47]. There is some controversy regarding whether hyper- report a 32% reduction in maximal VO2 in thyroid-defi- thyroidism per se leads to cardiac dysfunction. Although cient rats compared with control animals. In contrast, left-ventricular systolic function at rest is increased in short-term experimental hyperthyroidism in humans re- hyperthyroidism, a decrease in left-ventricular ejection sulted a reduction of peak VO2 of approximately 5% fraction during exercise compared to rest has been re- among otherwise healthy young volunteers [50]. In the lat- ported in some hyperthyroid patients [48, 49]. Clinical ter experiments, heart rate and cardiac output were ele- heart failure is nevertheless uncommon in this condition, vated at rest and throughout exercise, and VO2 at any giv- and, when it occurs, may be related to sustained tachycar- en work rate was higher in the hyperthyroid state. For dia rather than contractile dysfunction [47]. Changes in both hypothyroid and acutely hyperthyroid rats, lactate skeletal muscle are described in experimental hyperthy- threshold occurred at a lower work rate of exercise than roidism, including a shift of fiber type distribution to- for euthyroid animals [63]. wards fast twitch type [50], reduction of muscle mass due to negative protein balance [50], and more rapid deple- Findings on Cardiopulmonary Exercise Testing tion of glycogen during exercise [51]. Given the preserva- Related to Thyroid Disorders tion of cardiac output and muscle blood flow during exer- Studies of exercise function in patients with untreated cise [52], these changes in skeletal muscle are proposed as hyperthyroidism reveal similar findings to the experimen- the principle bases of reduced exercise endurance in tal models described above. Specifically, heart rate and hyperthyroidism [50, 51]. VO2 are high at rest, and remains high relative to any giv- Hypothyroidism is associated with more readily dem- en work rate [64–66]. Kahaly et al. [67] studied 42 hyper- onstrated impairments in cardiac function than hyperthy- thyroid patients with echocardiography and found that roidism, though the clinical findings of altered cardiovas- cardiac output, ejection fraction and stroke volume are all cular activity at rest are generally less apparent. Both sys- high at rest, and systemic vascular resistance low. With tolic [53] and diastolic [54] function of the heart are exercise, the increase in cardiac output was less than in depressed in hypothyroidism due to loss of inotropic and the euthyroid condition, though peak values were margin- lusitropic actions of thyroid hormone on myocardium ally higher. Stroke volume declined from rest to exercise [47]. Despite this, common echocardiographic measures in untreated hyperthyroidism; treatment with popranolol decreased resting but not maximal values, restoring the 268 Sietsema

pattern of increase from rest to exercise. Consistent with a disproportionate increase in heart rate and cardiac out- this, the same authors reported [65] that O2 pulse was nor- put relative to VO2. In hypothyroidism, in contrast, heart mal at rest, but low during exercise and only partially cor- rate is low at rest and the response to exercise is attenuat- rected by treatment with ß-adrenergic blockade. The rate ed. Peak VO2 and lactate threshold are both likely to be of increase of VO2 relative to the increase in work rate has significantly reduced in symptomatic hypothyroidism, been variously reported to be elevated in hyperthyroid though there are few data related to this in human sub- compared to euthyroid subjects [66] or within the normal jects. Additional data regarding exercise responses in sub- range [64, 65]. Some investigators report a decrease in the clinical hypothyroidism would be of particular interest as slope when the same patients are restudied under euthy- the need for therapy for these patients is not fully roid conditions [64, 66], but others do not [65]. In con- defined. trast to the study of short-term experimental hyperthy- roidism [50], patients with clinical disease have been Chronic Fatigue Syndrome reported to have similar VO2 values at the anaerobic threshold and peak exercise when euthyroid compared to Although the etiology of chronic fatigue syndrome when they had been hyperthyroid [65, 66]. Because VO2 is (CFS) remains unclear, there is an extensive literature elevated at rest and relative to work rate in hyperthyroid- characterizing its clinical features, and outlining standard ism, the increase in VO2 from rest to peak exercise, and criteria for case definitions for clinical research purposes the work rates associated with anaerobic threshold and [70, 71]. Common to the several working case definitions peak exercise are higher under euthyroid conditions com- are fatigue of at least 6 months’ duration which persists pared to hyperthyroid. Ventilation is high at rest in the despite rest and reduction in activities, and a negative hyperthyroid states, but not significantly different at peak evaluation for other conditions that would account for the exercise under the two conditions [65]. symptoms. Patients commonly report orthostatic and ex- ercise intolerance and complain that even short periods of There are relatively few reports of exercise responses in exertion result in exaggerated symptoms of fatigue for hypothyroid humans. Short-term hypothyroidism ap- days thereafter. Exercise intolerance is thus a central fea- pears to have minor effects on cardiac contractile proper- ture of CFS, but despite considerable interest, the basis ties [55]. Exercise cardiac output is reduced compared to for it has not been identified. control conditions [55] but it is not clear whether this reduction is disproportionate to the reduction in metabol- Physiologic Effects of Chronic Fatigue Syndrome ic rate. In subclinical hypothyroidism, subtle abnormali- Relevant to Cardiopulmonary Exercise Testing ties of left-ventricular function are reported during exer- Although a great many contradictory reports have been cise [68, 69], but it is doubted that these are of clinical generated regarding one or another abnormality of organ significance. Even in symptomatic patients first present- system function among selected patients, the overall con- ing with thyroid deficiency, abnormalities of left-ventric- sensus is that there is no specific abnormality of neuro- ular diastolic function may be difficult to detect except in muscular, cardiovascular or respiratory function underly- the more severely affected patients [56]. Despite this, ing the condition. reports of impaired skeletal muscle function [58] in sub- Clinical studies have been complicated by the apparent jects with subclinical hypothyroidism suggests that maxi- heterogeneity of the patient population, with many re- mal exercise capacity is likely to be reduced. ported findings affecting only a portion of a study group. Some of the more consistent findings that are reported From the above it can be expected that marked abnor- include certain laboratory studies such as mild reductions malities may be recognized on cardiopulmonary exercise in 24-hour cortisol excretion [72], decreased activity of tests in patients with hyperthyroidism. Resting levels of natural killer cells [73, 74], and other alterations in VO2, VCO2, heart rate and ventilation will all be elevated, immune effector cells or their products [74, 75]. These and will remain high relative to normal values for a given findings have been taken as support for the concept that level of submaximal work. The increase in VO2 relative to patients with CFS have chronic activation of the immune work rate may be higher than average, but does not neces- system, but the basis for this and the relationship of these sarily exceed the upper limits of normal. Peak VO2 is abnormalities to specific symptoms remain to be de- reduced little, if at all, by uncomplicated hyperthyroid- fined. ism; however, peak work rate will likely be reduced due to the shift of VO2 relative to work rate. Despite the hyper- dynamic circulation, O2 pulse may be reduced because of Exercise Responses in Systemic Conditions 269

Findings on Cardiopulmonary Exercise Testing series in which peak VO2 was only mildly reduced [77], Related to CFS but both the work rate and heart rate corresponding to A number of series of patients with CFS undergoing lactate threshold were low in the series finding lower peak cardiopulmonary exercise testing have been reported. VO2 values [81]. The rate of increase of VO2 relative to Several relatively small series of patients found average work rate is only reported in a few studies, and has been peak VO2 values [76, 77] or peak work capacity [78] to be normal [77, 80]. No abnormality in ventilatory responses only mildly reduced compared to age- and gender- to exercise are reported from these studies. matched control subjects, or compared to published nor- mal values [79, 80]. A larger series [81] of 427 patients Some investigators conclude that patients with CFS and 204 age- and gender-matched sedentary control sub- resemble severely deconditioned normal subjects. Decon- jects reports a more significant reduction in peak VO2, ditioning would not be surprising, as meeting the case def- with the mean value of 20.5 ml/min/kg corresponding to initions almost requires that patients be inactive, and only 64% of the value for control subjects. These investi- patients represented in the studies had often been symp- gators and others [77] have identified that a higher pro- tomatic for many years. Cause and effect with respect to portion of patients than control fail to demonstrate find- exercise capacity and exercise behavior in this stetting ings confirming that maximal effort had been attained remains to be established, however. during testing. However, analysis of the subset of patients with reasonably high peak heart rate and RER values still To summarize, exercise testing in patients with CFS showed peak VO2 to be significantly lower than in the often reveals mild-to-moderate reductions in peak VO2, control subjects [81]. The discrepancy in peak VO2 values high submaximal heart rates, but low peak heart rate. As reported by various investigators may reflect recruitment there are no pathognomonic findings of CFS on cardio- bias towards higher functioning patients for inclusion in pulmonary exercise testing, the role of testing is largely for some series. There is, however, clearly a spectrum of exer- evaluating alternative diagnoses, or for prescribing exer- cise performance even among patients meeting the same cise training. With respect to the latter, it may be noted case definition criteria. The relatively well-preserved ex- that despite the history of symptom exacerbation by exer- ercise function of some symptomatic patients argues cise, incremental testing has been done without clinically against the acute cardiorespiratory responses demon- important aggravation of the patients’ conditions. There strated on exercise testing to be the fundamental basis of are findings to support a decline in cognitive function fol- clinical symptoms in CFS. lowing a bout of maximal exercise [87, 88], as well as a Some series, but not all [77], report higher heart rates decrease in spontaneous activity over subsequent days as for patients compared to controls at a given submaximal reflected in accelerometer readings [89]. Both of these level of upright exercise. One [82] found lower heart rate observations are consistent with symptomatic reports of responses during supine exercise. Peak heart rates are prolonged post-exercise aesthenia in this condition. Nev- sometimes [80, 81], but not always [77–79], less than pre- ertheless, Mullis et al. [80] concluded that of 130 patients dicted. When subnormal peak heart rates have been undergoing maximal testing, none had lasting deteriora- noted, this has been variously interpreted to reflect pre- tion in their condition following testing, and only 3% mature limitation of exercise at a submaximal level due to reported exacerbation of symptoms. This is of practical central (i.e. nonmuscular) fatigue [79], or to reflect an importance as exercise training has been reported to intrinsic defect in chronotropic response [81], perhaps improve sense of well-being and level of fitness of patients due to altered autonomic control. The question of auto- with CFS [90]. nomic dysfunction is also raised by another report of a high prevalence of abnormal results on tilt testing [83]. These findings have subsequently been contradicted by others [84], however, and additional studies of cardiovas- cular autonomic responses in CFS patients at rest have not shown consistent abnormalities [85, 86]. Whether the low peak heart rates and peak VO2 values represent intrinsic effects of CFS or are the secondary results of poor maximal performance remains unclear. The lactate threshold has been reported to be normal in a 270 Sietsema

References 1 Luce JM: Respiratory complications of obesity. 18 Freyschuss U, Melcher A: Exercise energy ex- filling in insulin-dependent and non-insulin- Chest 1980;78:626–631. penditure in extreme obesity: Influence of er- dependent diabetic patients. Clin Cardiol gometry type and weight loss. Scand J Clin Lab 1997;20:536–540. 2 Sharp JT, Henry JP, Sweany SK: The total Invest 1978;38:753–759. 34 Davis TM, Knuiman M, Kendall P, Vu H, work of breathing in normal and obese men. J Davis WA: Reduced pulmonary function and Clin Invest 1964;43:728–734. 19 Dempsey JA, Reddan W, Balke B, Rankin J: its associations in type 2 diabetes: The Freman- Work capacity determinants and physiologic tle Diabetes Study. Diabetes Res Clin Pract 3 Thomas PS, Cowen ER, Hulands G, Milledge cost of weight-supported work in obesity. J 2000;50:153–159. JS: Respiratory function in the morbidly obese Appl Physiol 1966;21:1815–1820. 35 Barrett-Connor E, Frette C: NIDDM, impaired before and after weight loss. Thorax 1989;44: glucose tolerance, and pulmonary function in 382–386. 20 Anton-Kuchly B, Roger P, Varene P: Determi- older adults: The Rancho Bernardo Study. Dia- nants of increased energy cost of submaximal betes Care 1996;19:1441–1444. 4 Ray CS, Sue DY, Bray G, Hansen JE, Wasser- exercise in obese subjects. J Appl Physiol 1984; 36 Radice M, Rocca A, Bedon E, Musacchio N, man K: Effects of obesity on respiratory func- 56:18–23. Morabito A, Segalini G: Abnormal response to tion. Am Rev Respir Dis 1983;128:501–506. exercise in middle-aged NIDDM patients with 21 Wasserman K, Hansen JE, Sue DY, Casaburi and without autonomic neuropathy. Diabet 5 Holley HS, Milic-Emili J, Becklake MR, Bates R, Whipp BJ: Normal values; in Principles of Med 1996;13:259–265. DV: Regional distribution of pulmonary venti- Exercise Testing and Interpretation, ed 3. Phil- 37 Veves A, Saouaf R, Donaghue VM, Mullooly lation and perfusion in obesity. J Clin Invest adelphia, Lippincott Williams & Wilkins, CA, Kistler JA, Giurini JM, Horton ES, Fielding 1967;46:475–481. 1999, pp 143–164. RA: Aerobic exercise capacity remains normal despite impaired endothelial function in the mi- 6 Hakala K, Mustajoki P, Aittomaki J, Sovijarvi 22 Wasserman K, Whipp BJ: Excercise physiology cro- and macrocirculation of physically active A: Improved gas exchange during exercise after in health and disease. Am Rev Respir Dis IDDM patients. Diabetes 1997;46:1846–1852. weight loss in morbid obesity. Clin Physiol 1975;112:219–249. 38 Regensteiner JG, Sippel J, McFarling ET, Wol- 1996;16:229–238. fel EE, Hiatt WR: Effects of non-insulin-depen- 23 Hansen JE, Sue DY, Wasserman K: Predicted dent diabetes on oxygen consumption during 7 Alpert MA, Lambert CR, Terry BE, Cohen values for clinical exercise testing. Am Rev treadmill exercise. Med Sci Sports Exerc 1995; MV, Mukerji V, Massey CV, Hashimi MW, Respir Dis 1984;129:S49–55. 27:875–881. Panayiotou H: Interrelationship of left ventric- 39 Baraldi E, Monciotti C, Filippone M, Santuz P, ular mass, systolic function and diastolic filling 24 Babb TG, Buskirk ER, Hodgson JL: Exercise Magagnin G, Zanconato S, Zacchello F: Gas in normotensive morbidly obese patients. Int J end-expiratory lung volumes in lean and moder- exchange during exercise in diabetic children. Obes Relat Metab Disord 1995;19:550–557. ately obese women. Int J Obes 1989;13:11–19. Pediatr Pulmonol 1992;13:155–160. 40 Bottini P, Tantucci C, Scionti L, Dottorini ML, 8 Alpert MA: Obesity cardiomyopathy: Patho- 25 Li J, Li S, Feuers RJ, Buffington CK, Cowan Puxeddu E, Reboldi G, Bolli GB, Casucci G, physiology and evolution of the clinical syn- GS: Influence of body fat distribution on oxy- Santeusanio F, Sorbini CA: Cardiovascular re- drome. Am J Med Sci 2001;321:225–236. gen uptake and pulmonary performance in sponse to exercise in diabetic patients: In- morbidly obese females during exercise. Respi- fluence of autonomic neuropathy of different 9 Kaltman AJ, Goldring RM: Role of circulatory rology 2001;6:9–13. severity. Diabetologia 1995;38:244–250. congestion in the cardiorespiratory failure of 41 Katoh J, Hara Y, Kurusu M, Miyaji J, Naruta- obesity. Am J Med 1976;60:645–653. 26 Uusitupa MI, Mustonen JN, Airaksinen KE: ki K: Cardiorespiratory function as assessed by Diabetic heart muscle disease. Ann Med 1990; exercise testing in patients with non-insulin- 10 Alexander JK, Dennis DW, Smith WG, Amad 22:377–386. dependent diabetes mellitus. J Int Med Res KH, Ducan WC, Austin RD: Blood volume, 1996;24:209–213. cardiac output and distribution of systemic 27 Fein FS, Sonnenblick EH: Diabetic cardiomyo- 42 Wanke T, Formanek D, Auinger M, Zwick H, blood flow in extreme obesity. Cardiovasc Res pathy. Cardiovasc Drugs Ther 1994;8:65–73. Irsigler K: Pulmonary gas exchange and oxygen Center Bull 1962;1:39–43. uptake during exercise in patients with type 1 di- 28 Zabalgoitia M, Ismaeil MF, Anderson L, Mak- abetes mellitus. Diabet Med 1992;9:252–257. 11 Lauer MS, Anderson KM, Kannel WB, Levy lady FA: Prevalence of diastolic dysfunction in 43 Roy TM, Peterson HR, Snider HL, Cyrus J, D: The impact of obesity on left ventricular normotensive, asymptomatic patients with Broadstone VL, Fell RD, Rothchild AH, Sa- mass and geometry. The Framingham Heart well-controlled type 2 diabetes mellitus. Am J mols E, Pfeifer MA: Autonomic influence on Study. JAMA 1991;266:231–236. Cardiol 2001;87:320–323. cardiovascular performance in diabetic sub- jects. Am J Med 1989;87:382–388. 12 Ben-Dov I, Grossman E, Stein A, Shachor D, 29 Poirier P, Garneau C, Bogaty P, Nadeau A, 44 Estacio RO, Regensteiner JG, Wolfel EE, Jef- Gaides M: Marked weight reduction lowers Marois L, Brochu C, Gingras C, Fortin C, fers B, Dickenson M, Schrier RW: The associa- resting and exercise blood pressure in morbidly Jobin J, Dumesnil JG: Impact of left ventricu- tion between diabetic complications and exer- obese subjects. Am J Hypertens 2000;13:251– lar diastolic dysfunction on maximal treadmill cise capacity in NIDDM patients. Diabetes 255. performance in normotensive subjects with Care 1998;21:291–295. well-controlled type 2 diabetes mellitus. Am J 45 Regensteiner JG, Bauer TA, Reusch JE, Bran- 13 Nielsen S, Hensrud DD, Romanski S, Levine Cardiol 2000;85:473–477. denburg SL, Sippel JM, Vogelsong AM, Smith JA, Burguera B, Jensen MD: Body composition S, Wolfel EE, Eckel RH, Hiatt WR: Abnormal and resting energy expenditure in humans: 30 Di Bonito P, Cuomo S, Moio N, Sibilio G, oxygen uptake kinetic responses in women Role of fat, fat-free mass and extracellular Sabatini D, Quattrin S, Capaldo B: Diastolic with type II diabetes mellitus. J Appl Physiol fluid. Int J Obes Relat Metab Disord 2000;24: dysfunction in patients with non-insulin-de- 1998;85:310–317. 1153–1157. pendent diabetes mellitus of short duration. 46 Minette P, Buysschaert M, Rahier J, Veriter C, Diabet Med 1996;13:321–324. Frans A: Pulmonary gas exchange in life-long 14 Goran M, Fields DA, Hunter GR, Herd SL, nonsmoking patients with diabetes mellitus. Weinsier RL: Total body fat does not influence 31 Nicolino A, Longobardi G, Furgi G, Rossi M, Respiration 1999;66:20–24. maximal aerobic capacity. Int J Obes Relat Zoccolillo N, Ferrara N, Rengo F: Left ventric- Metab Disord 2000;24:841–848. ular diastolic filling in diabetes mellitus with and without hypertension. Am J Hypertens 15 Nevill AM, Holder RL: Scaling, normalizing, 1995;8:382–389. and per ratio standards: An allometric modeling approach. J Appl Physiol 1995;79:1027–1031. 32 Scognamiglio R, Avogaro A, Casara D, Crepal- di C, Marin M, Palisi M, Mingardi R, Erle G, 16 Batterham AM, Vanderburgh PM, Mahar MT, Fasoli G, Dalla Volta S: Myocardial dysfunc- Jackson AS: Modeling the influence of body tion and adrenergic cardiac innervation in pa- size on V(O2) peak: Effects of model choice and tients with insulin-dependent diabetes melli- body composition. J Appl Physiol 1999;87: tus. J Am Coll Cardiol 1998;31:404–412. 1317–1325. 33 Astorri E, Fiorina P, Contini GA, Albertini D, 17 Whipp BJ, Davis JA: The ventilatory stress of Magnati G, Astorri A, Lanfredini M: Isolated exercise in obesity. Am Rev Respir Dis 1984; and preclinical impairment of left ventricular 129:S90–92. Exercise Responses in Systemic Conditions 271

47 Klein I, Ojamaa K: Thyroid hormone and the 62 Zwillich CW, Pierson DJ, Hofeldt FD, Lufkin chronic fatigue syndrome. Am J Med 1996; cardiovascular system. N Engl J Med 2001; EG, Weil JV: Ventilatory control in myxedema 100:634–640. 344:501–509. and hypothyroidism. N Engl J Med 1975;292: 78 Gibson H, Carroll N, Clague JE, Edwards RH: 662–665. Exercise performance and fatiguability in pa- 48 Forfar JC, Muir AL, Sawers SA, Toft AD: tients with chronic fatigue syndrome. J Neurol Abnormal left ventricular function in hyper- 63 Zarzeczny R, Pilis W, Langfort J, Kaciuba- Neurosurg Psychiatry 1993;56:993–998. thyroidism: Evidence for a possible reversible Uscilko H, Nazar K: Influence of thyroid hor- 79 Kent-Braun JA, Sharma KR, Weiner MW, cardiomyopathy. N Engl J Med 1982;307: mones on exercise tolerance and lactate thresh- Massie B, Miller RG: Central basis of muscle 1165–1170. old in rats. J Physiol Pharmacol 1996;47:503– fatigue in chronic fatigue syndrome. Neurology 513. 1993;43:125–131. 49 Iskandrian AS, Rose L, Hakki AH, Segal BL, 80 Mullis R, Campbell IT, Wearden AJ, Morriss Kane SA: Cardiac performance in thyrotoxico- 64 Ben-Dov I, Sietsema KE, Wasserman K: O2 RK, Pearson DJ: Prediction of peak oxygen sis: Analysis of 10 untreated patients. Am J uptake in hyperthyroidism during constant uptake in chronic fatigue syndrome. Br J Sports Cardiol 1983;51:349–352. work rate and incremental exercise. Eur J Appl Med 1999;33:352–356. Physiol Occup Physiol 1991;62:261–267. 81 De Becker P, Roeykens J, Reynders M, McGre- 50 Martin WH, Spina RJ, Korte E, Yarasheski gor N, De Meirleir K: Exercise capacity in KE, Angelopoulos TJ, Nemeth PM, Saffitz JE: 65 Kahaly GJ, Nieswandt J, Wagner S, Schlegel J, chronic fatigue syndrome. Arch Intern Med Mechanisms of impaired exercise capacity in Mohr-Kahaly S, Hommel G: Ineffective car- 2000;160:3270–3277. short duration experimental hyperthyroidism. diorespiratory function in hyperthyroidism. J 82 Montague TJ, Marrie TJ, Klassen GA, Bewick J Clin Invest 1991;88:2047–2053. Clin Endocrinol Metab 1998;83:4075–4078. DJ, Horacek BM: Cardiac function at rest and with exercise in the chronic fatigue syndrome. 51 McAllister RM, Delp MD, Laughlin MH: Thy- 66 Kimura H, Kawagoe Y, Kaneko N, Fessler HE, Chest 1989;95:779–784. roid status and exercise tolerance: Cardiovas- Hosoda S: Low efficiency of oxygen utilization 83 Bou-Holaigah I, Rowe PC, Kan J, Calkins H: cular and metabolic considerations. Sports during exercise in hyperthyroidism. Chest The relationship between neurally mediated Med 1995;20:189–198. 1996;110:1264–1270. hypotension and the chronic fatigue syndrome. JAMA 1995;274:961–967. 52 McAllister RM, Sansone JC, Laughlin MH: 67 Kahaly GJ, Wagner S, Nieswandt J, Mohr- 84 Poole J, Herrell R, Ashton S, Goldberg J, Buch- Effects of hyperthyroidism on muscle blood Kahaly S, Ryan TJ: Stress echocardiography in wald D: Results of isoproterenol tilt table test- flow during exercise in rats. Am J Physiol 1995; hyperthyroidism. J Clin Endocrinol Metab ing in monozygotic twins discordant for 268:H330–335. 1999;84:2308–2313. chronic fatigue syndrome. Arch Intern Med 2000;160:3461–3468. 53 Bough EW, Crowley WF, Ridgway C, Walker 68 Arem R, Rokey R, Kiefe C, Escalante DA, 85 Soetekouw PM, Lenders JW, Bleijenberg G, H, Maloof F, Myers GS, Daniels GH: Myocar- Rodriguez A: Cardiac systolic and diastolic Thien T, van der Meer JW: Autonomic func- dial function in hypothyroidism: Relation to function at rest and exercise in subclinical hy- tion in patients with chronic fatigue syndrome. disease severity and response to treatment. pothyroidism: Effect of thyroid hormone thera- Clin Auton Res 1999;9:334–340. Arch Intern Med 1978;138:1476–1480. py. Thyroid 1996;6:397–402. 86 De Becker P, Dendale P, De Meirleir K, Cam- pine I, Vandenborne K, Hagers Y: Autonomic 54 Wieshammer S, Keck FS, Waitzinger J, Henze 69 Forfar JC, Wathen CG, Todd WT, Bell GM, testing in patients with chronic fatigue syn- E, Loos U, Hombach V, Pfeiffer EF: Acute Hannan WJ, Muir AL, Toft AD: Left ventricu- drome. Am J Med 1998;105:22S–26S. hypothyroidism slows the rate of left ventricu- lar performance in subclinical hypothyroidism. 87 LaManca JJ, Sisto SA, DeLuca J, Johnson SK, lar diastolic relaxation. Can J Physiol Pharma- Q J Med 1985;57:857–865. Lange G, Pareja J, Cook S, Natelson BH: col 1989;67:1007–1010. Influence of exhaustive treadmill exercise on 70 Bates DW, Buchwald D, Lee J, Kith P, Doolit- cognitive functioning in chronic fatigue syn- 55 Wieshammer S, Keck FS, Waitzinger J, Kohler tle TH, Umali P, Komaroff AL: A comparison drome. Am J Med 1998;105:59S–65S. J, Adam W, Stauch M, Pfeiffer EF: Left ven- of case definitions of chronic fatigue syndrome. 88 Blackwood SK, MacHale SM, Power MJ, tricular function at rest and during exercise in Clin Infect Dis 1994;18(suppl 1):S11–15. Goodwin GM, Lawrie SM: Effects of exercise acute hypothyroidism. Br Heart J 1988;60: on cognitive and motor function in chronic 204–211. 71 Fukuda K, Straus SE, Hickie I, Sharpe MC, fatigue syndrome and depression. J Neurol Dobbins JG, Komaroff A: The chronic fatigue Neurosurg Psychiatry 1998;65:541–546. 56 Tielens ET, Pillay M, Storm C, Berghout A: syndrome: A comprehensive approach to its 89 Sisto SA, Tapp WN, LaManca JJ, Ling W, Changes in cardiac function at rest before and definition and study. International Chronic Fa- Korn LR, Nelson AJ, Natelson BH: Physical after treatment in primary hypothyroidism. tigue Syndrome Study Group. Ann Intern Med activity before and after exercise in women Am J Cardiol 2000;85:376–380. 1994;121:953–959. with chronic fatigue syndrome. QJM 1998;91: 465–473. 57 Baldwin KM, Ernst SB, Herrick RE, Hooker 72 Demitrack MA, Dale JK, Straus SE, Laue L, 90 Fulcher KY, White PD: Randomised con- AM, Mullin WJ: Exercise capacity and cardiac Listwak SJ, Kruesi MJ, Chrousos GP, Gold trolled trial of graded exercise in patients with function in trained and untrained thyroid-defi- PW: Evidence for impaired activation of the the chronic fatigue syndrome. BMJ 1997;314: cient rats. J Appl Physiol 1980;49:1022–1026. hypothalamic-pituitary-adrenal axis in patients 1647–1652. with chronic fatigue syndrome. J Clin Endocri- 58 Monzani F, Caraccio N, Siciliano G, Manca L, nol Metab 1991;73:1224–1234. Kathy E. Sietsema, MD Murri L, Ferrannini E: Clinical and biochemi- Division of Respiratory and Critical Care cal features of muscle dysfunction in subclini- 73 Whiteside TL, Friberg D: Natural killer cells Physiology and Medicine cal hypothyroidism. J Clin Endocrinol Metab and natural killer cell activity in chronic fatigue Harbor-UCLA Medical Center 1997;82:3315–3318. syndrome. Am J Med 1998;105:27S–34S. 1000 West Carson Street Torrance, CA 90509 (USA) 59 Argov Z, Renshaw PF, Boden B, Winokur A, 74 Barker E, Fujimura SF, Fadem MB, Landay Tel. +1 310 222 3803, Fax +1 310 328 9849 Bank WJ: Effects of thyroid hormones on skele- AL, Levy JA: Immunologic abnormalities asso- E-Mail [email protected] tal muscle bioenergetics: In vivo phosphorus- ciated with chronic fatigue syndrome. Clin In- 31 magnetic resonance spectroscopy study of fect Dis 1994;18(suppl 1):S136–141. humans and rats. J Clin Invest 1988;81:1695– 1701. 75 Buchwald D, Komaroff AL: Review of labora- tory findings for patients with chronic fatigue 60 Khaleeli AA, Griffith DG, Edwards RH: The syndrome. Rev Infect Dis 1991;13(suppl 1): clinical presentation of hypothyroid myopathy S12–18. and its relationship to abnormalities in struc- ture and function of skeletal muscle. Clin En- 76 Riley MS, O’Brien CJ, McCluskey DR, Bell docrinol (Oxf) 1983;19:365–376. NP, Nicholls DP: Aerobic work capacity in patients with chronic fatigue syndrome. BMJ 61 Martinez FJ, Bermudez-Gomez M, Celli BR: 1990;301:953–956. Hypothyroidism. A reversible cause of dia- phragmatic dysfunction. Chest 1989;96:1059– 77 Sisto SA, LaManca J, Cordero DL, Bergen MT, 1063. Ellis SP, Drastal S, Boda WL, Tapp WN, Natel- son BH: Metabolic and cardiovascular effects of a progressive exercise test in patients with 272 Sietsema

Weisman IM, Zeballos RJ (eds): Clinical Exercise Testing. Prog Respir Res. Basel, Karger, 2002, vol 32, pp 273–281 OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO Clinical Exercise Testing during Pregnancy and the Postpartum Period Mary L. O’Toole Raul Artal Women’s Exercise Research Laboratory, Department of Obstetrics, Gynecology and Women’s Health, Saint Louis University, St. Louis, Mo., USA Summary women are sedentary, but see pregnancy as a time to develop healthy lifestyle habits including participation in Cardiopulmonary exercise testing of pregnant women may regular physical activity. Other young women have or be done for either diagnostic or nondiagnostic reasons. The develop medical conditions during their pregnancies for most common reason to exercise test healthy young women, which regular physical activity would be beneficial. Indi- pregnant or not, is to acquire information with which to develop viduals from each of these groups may benefit from car- individualized exercise prescriptions. Pregnant women can be diopulmonary exercise testing to assist in the planning of tested either during weight-bearing activity (e.g., treadmill walk- physical activity during pregnancy. ing) or during an activity that is independent of body weight (e.g., cycle ergometry). Semi-recumbent cycle ergometry appears to As with other populations, cardiopulmonary exercise be the most comfortable modality. Short incremental protocols testing of pregnant women may be done for either diag- (8–12 min) to maximal subjective effort provide good informa- nostic or nondiagnostic reasons. Diagnostic tests are done tion and appear to be safe during all stages of pregnancy for both to establish the safety of participation while nondiagnos- mother and fetus. Single-stage submaximal protocols are also tic tests may be used to assess fitness, provide a basis for safe and may provide useful information, particularly when sub- exercise prescription or monitor progress in an exercise strate or hormonal responses are of interest. For proper inter- program. Pregnant women, however, present unique con- pretation of results, individuals involved in conducting exercise cerns regarding exercise testing and participation. Unlike testing of pregnant women should be aware of the anatomical usual diagnostic tests for disease conditions, the well- and physiological changes that may alter cardiopulmonary re- being of the fetus and the exercising mother must be sponses to exercise during pregnancy. established. Thus, all tests, while not strictly diagnostic, are used to establish safety of exercise participation dur- Introduction ing pregnancy. Pregnancy is associated with profound anatomical and physiological changes that alter cardio- Many young women consider physical activity to be an pulmonary responses both at rest and during exercise. integral part of their lives and wish to continue regular Individuals involved in conducting cardiopulmonary ex- activity throughout their pregnancies. Many other young ercise testing of pregnant women must be aware of these changes in order to choose appropriate testing modalities and protocols as well as to be able to correctly interpret exercise test results.

Anatomic/Physiologic Changes Associated with testing. Such changes include an increase in blood vol- Pregnancy ume, heart rate, and stroke volume as well as resultant cardiac output, and a decrease in systemic vascular resis- The most obvious changes with the potential to affect tance [11, 14, 15]. Blood volume expands to 40–50% exercise test responses during pregnancy are the increase above resting levels, reaching a peak towards the end of in body weight and fat. Average weight gain during preg- the second trimester [16]. Increases in both plasma vol- nancy is approximately 12.5 kg, approximately a 20% ume (approximately 50%) and red cell volume (20%) con- increase in body weight for most women [1]. Weight gain tribute to the increase in blood volume [17–19]. Associat- increases with increasing gestation such that expected ed increases in end-diastolic volumes in conjunction with weight gain during the first trimester is only 1–3 kg, while slightly increased contractility result in larger stroke vol- 6–8 kg increases occur during the second trimester and umes than during the nonpregnant state [20]. Resting 3.5–4 kg increases are usual during the third trimester. Of heart rates start to rise as early as 8 weeks of gestation this, 2.5 kg is estimated to be increased maternal fat [2, 3]. (approximately 8 bpm) and reach a peak by 32 weeks of Although the time course for changes in maternal body fat gestation that is about 20 bpm higher than in the nonpreg- are difficult to precisely quantify, most evidence suggests nant state [21, 22]. Most studies show that maternal that percent body fat increases during the first and second stroke volume increases by 10% by the end of the first trimesters, then plateaus or decreases slightly during the trimester and is followed by a 20% increase in heart rate third trimester [1, 4]. during the second and third trimesters [23]. The increased cardiac output results in a chronic volume overload of the The gains in weight and body fat result in changes in heart. As in other situations with chronic volume over- posture, balance, and locomotor patterns. Protrusion of load, the left-ventricular chamber increases in size with the abdomen with accompanying increased lumbar lordo- little or no change in wall thickness [24–26]. Despite this sis and anterior pelvic rotation results in altered spinal increase in cardiac output, mean arterial pressure de- mechanics and displacement forward of the woman’s line creases by 5–10 mm Hg by the middle of the second tri- of gravity. These changes result in low back pain for mester and then gradually increases back to prepregnancy approximately 50% of pregnant women and may alter levels [13]. The decreased blood pressures are mediated their ability to perform an adequate exercise test during by venous relaxation, greater venous capacitance and walking on a treadmill [5]. An upward displacement of the reduced peripheral resistance [27, 28]. In particular, there diaphragm alters pulmonary mechanics so that residual is an increase in the vasculature of the uterine circulation volume, expiratory reserve volume and functional residu- and decreases in vascular resistance in the skin, kidney al capacity are all reduced at rest. However, inspiratory and uteroplacental circulations [23]. These hemodynamic capacity is increased so that vital capacity is not changed changes appear to establish a circulatory reserve necessary and resting pulmonary function is well preserved [6–8]. to provide nutrients and oxygen to both mother and fetus Increase diaphragmatic work and greater use of accessory at rest and during moderate exercise. muscle of breathing do result in an increased work of breathing that is partially compensated for by decreased Clinical Screening Prior to Exercise Testing during bronchial smooth muscle tone and reduced pulmonary Pregnancy resistance [9]. Pretest screening should be designed to determine In addition to the anatomic changes, pregnancy affects whether sufficient physiologic reserve exists to accommo- the pulmonary system by increasing ventilatory sensitivi- date maternal and fetal needs plus additional demands of ty. The respiratory center has a decreased threshold to exercise testing. A thorough pretest medical history respond to PCO2 and an increased sensitivity to any rise should include both past and present information. Sug- in PCO2 [6, 7, 11–13]. These changes are thought to be gested components for this exam include any medical secondary to increased circulating progesterone levels and diagnoses, findings from previous physical exams, a thor- result in larger tidal volumes and minute ventilations. ough history of any symptoms, recent illnesses, orthoped- Resting arterial PO2 levels rise to approximately 100 mm ic problems, medication use, other habits such as use of Hg and arterial PCO2 levels decrease to 30–32 mm Hg. caffeine, alcohol or tobacco, exercise history, work history Respiratory alkalosis results and is only partially compen- with emphasis on physical demands, and family history of sated for by urinary excretion of bicarbonate. Pregnancy also causes changes in the cardiovascular system that have an impact on responses during exercise 274 O’Toole/Artal

Table1. Contraindications to exercise testing* Table 2. Contraindications to exercise testing or participation during pregnancy Absolute Recent significant change in resting ECG suggestive of ischemia, Absolute Hemodynamically significant heart disease myocardial infarction or other acute cardiac event Constrictive lung disease Unstable angina History of three or more spontaneous abortions Uncontrolled arrhythmias causing symptoms or hemodynamic Incompetent cervix/cerlage Multiple gestation compromise Persistent second-or-third-trimester bleeding Severe symptomatic aortic stenosis Placenta previa Uncontrolled congestive heart failure Premature labor during the prior or current pregnancy Acute pulmonary embolus or pulmonary infarction Ruptured membranes Acute myocarditis or pericarditis Pregnancy-induced hypertension Suspected or known dissecting aneurysm Acute infection Relative Anemia Breech presentation in last trimester Relative Cardiac arrhythmia or palpitations Left main coronary stenosis Chronic bronchitis Stenotic valvular heart disease of moderate severity Uncontrolled diabetes Significant electrolyte abnormalities Morbid obesity Severe arterial hypertension at rest (systolic BP 1 200 mm Hg; Extreme underweight History of bleeding during current pregnancy diastolic BP 1 110 mm Hg) History of extremely sedentary lifestyle Tachy- or bradyarrhythmias History of intrauterine growth restriction Hypertrophic cardiomyopathy or other forms of outflow tract History of precipitous labor Hypertension obstruction Orthopedic limitations Disorders (neuromuscular, musculoskeletal or rheumatoid) that are Seizure disorder Thyroid disease exacerbated by exercise Heavy smoker High-degree atrioventricular block Ventricular aneurysm Uncontrolled metabolic disease (e.g. diabetes, thrytoxicosis or myxedema) Chronic infectious disease (e.g. mononucleosis, hepatitis, AIDS) * Adapted from ref. [29]. cardiopulmonary or metabolic disease [29]. Contraindi- pregnant as well as nonpregnant women. The ‘gold stan- cations to exercise testing because of medical conditions dard’ tests for this purpose are tests to assess cardiopul- in the nonpregnant state are equally applicable during monary fitness by measuring maximal aerobic capacity pregnancy (table 1). In addition, certain obstetrical condi- (VO2max). These tests are typically incremental tests with tions may develop in a pregnant woman regardless of her work rates starting at low intensity and progressing up to previous activity or fitness level that disqualify her from maximal effort. In nonpregnant individuals, incremental safely participating in an exercise test. Use of the submaximal tests to a preset endpoint (such as 85% age- PARmed-X for Pregnancy may be an aid in identifying predicted maximal heart rate) are sometimes used to pre- these obstetrical conditions [30]. Table 2 is suggested as a dict maximal exercise capacity and develop individual- guide to determining the appropriateness of exercise test- ized exercise plans. Because of progressive changes in ing or participation during pregnancy for individual pa- hemodynamics, pulmonary function, and energy cost of tients [see also ref. 31]. exercise during the course of gestation, submaximal tests are less useful during pregnancy and may be quite mis- Exercise Test Considerations leading when used for this purpose (see below). Exercise tests may also be used and to monitor progress in response Reasons to Exercise Test Pregnant Women to an exercise training program and to adjust the exercise The most common reason to exercise test healthy prescription appropriately over time. In nonpregnant women is to acquire information with which to develop women, either maximal or submaximal tests may be relia- individualized exercise prescriptions. This is true for bly used. In pregnant women, submaximal tests may be used in most cases. However, if precise quantification of Clinical Exercise Testing during Pregnancy 275 and the Postpartum Period

fitness changes during pregnancy is important (such as in be adjusted in small amounts. Older mechanically braked research studies), maximal exercise tests are necessary. ergometers (e.g. Monark) require proper pedal speed to be maintained to obtain the desired resistance. This may be In addition to using exercise testing to guide healthy difficult for some women, particularly late in pregnancy pregnant women in an exercise program, exercise testing when the protruding abdomen may interfere will leg may be used to guide exercise training for women with cycling movements. Most newer models, however, are gestational diabetes. Testing procedures are similar to electromagnetically braked so that appropriate resistance those used with healthy women with the addition of blood will result across a wide range of pedal rates. Cycle ergom- glucose monitoring before and after exercise. Exercise eters are relatively inexpensive, require little space and tests that monitor substrate utilization and/or hormonal, are easily transportable. Upright cycling has frequently cardiovascular or pulmonary responses over a prolonged been used to test pregnant women. We have also used the period of time to a single submaximal intensity (see semi-recumbent cycle for sedentary, overweight pregnant below) may also be useful in guiding exercise of women women because of its more comfortable seat and body with gestational diabetes. Additional uses of exercise tests position. Physiologic monitoring is considerably easier during pregnancy are to evaluate suspected deficits in car- during cycling compared with treadmill testing. Arm diopulmonary or metabolic function and to monitor dis- ergometers have been used with pregnant women and ease-related decreases in physical function. have been shown to be safe (no uterine contractions dur- ing or immediately postexercise) [32]. However, we do Selection of Testing Mode not recommend routine use of arm ergometry to test preg- Pregnant women can be tested either during weight- nant women because of the small muscle mass used and bearing activity, such as treadmill walking or during an resultant local arm and shoulder fatigue. activity that is independent of body weight, such as cycle ergometry. Maximal Exercise Tests Weight-Bearing Exercise. Treadmill walking is a com- Protocol Selection. The basic pattern of maximal test monly used weight-bearing exercise testing mode. The protocols for pregnant women is not different from that main advantage of treadmill testing is that it provides a for nonpregnant women. Selection of a specific protocol familiar type of exercise stress (i.e. walking). Treadmill should be in accord with the purpose of the test, the out- tests can be used for women with marked differences in come measure(s) of interest and the individual being fitness level because of the wide range of speeds and tested. For example, if the purpose is to individualize a grades available. The treadmill test should be chosen walking program, a treadmill protocol would be appro- when the individualized exercise prescription is to be a priate. If, on the other hand, sedentary obese patients are walking program. Several disadvantages are associated to be tested, a semirecumbent cycle ergometer protocol with treadmill testing. Good balance and locomotor skills may be more appropriate. Treadmill protocols that call are necessary for safe and accurate testing. Although most for constant speed with increases in grade have been used treadmills have a front and at least one side rail for indi- with pregnant women [33] as have those using a constant viduals to steady themselves, accurate cardiopulmonary treadmill incline with increases in speed [34]. The tread- and metabolic measurements will only result when the mill protocol that we prefer to use with pregnant women is test is done without holding the handrails. Other disad- the ACT (Activity Counseling Trial) protocol [35]. Brief- vantages of treadmill testing include the expense of a ly, in the ACT protocol, participants are taught to walk on treadmill, inability to move it easily and difficulty in mak- the treadmill at a low speed until they can walk without ing some measurements, such as taking blood pressure. holding on to the handrails. This is particularly important For these reasons, treadmill testing may not be appro- for pregnant women whose balance may be altered by her priate for pregnant women late in gestation when balance pregnancy. Treadmill speed is then increased slowly until is difficult, when pregnancy-induced hypertension may be a speed that elicits a heart rate equivalent to 60–70% of present or for sequential measurements through the age-predicted maximal heart rate is identified. After a course of pregnancy. brief rest, the test begins at the identified speed and zero Weight-Supported Exercise. Cycle ergometers provide percent grade. The test proceeds with 2% grade incre- an option for non-weight-bearing testing and may be more ments every 2 min (approximately 1 MET/stage incre- useful during pregnancy. Cycle ergometers may be tradi- ments) until maximal effort is reached. This protocol is a tional upright leg cycles, semi-recumbent cycles or arm particularly good option for women who have been active cycles. Regardless of specific type, resistance can usually 276 O’Toole/Artal

Table 3. General indications for stopping an exercise test in low-risk cycling test on a semi-recumbent cycle ergometer to be adults* more comfortable for obese, pregnant women, particu- larly during the third trimester. The semi-recumbent cycle Onset of angina or angina-like symptoms protocol utilized in our laboratory employs 10-Watt incre- Significant drop (20 mm Hg) in systolic blood pressure or failure of ments (approximately 0.5 MET increments) each minute. We have found this protocol to be useful in evaluating systolic blood pressure to rise with an increase in exercise women with gestational diabetes who tend to be over- intensity weight or obese and sedentary. Spinnewijn et al. [40] have Excessive rise in blood pressure (SBP 1 260 mm Hg; used a tethered swimming protocol to evaluate exercise DBP 1 115 mm Hg) performance in a weightless environment. In this proto- Signs of poor perfusion: light-headedness, confusion, ataxia, pallor, col, the swimmer swims breast stroke attached to a pulley cyanosis, nausea, or cold and clammy skin system with adjustable weights. Initial resistance is 0.5 kg Failure of heart rate to increase with increase in intensity with 0.5 kg being added each minute until the swimmer Noticeable change in heart rhythm can no longer sustain the pull. Regardless of the protocol Subject requests to stop chosen, it is advisable to have an experienced obstetrical Physical or verbal manifestations of severe fatigue nurse use a cardiotocometer to monitor the fetus for 15– Failure of the testing equipment 20 min before and after the exercise test. Measurements of fetal baseline heart rate, frequency of accelerations and * Adapted from ref. [29]. decelerations, and fetal heart rate variability can then be used, according to the guidelines developed by the Na- Table 4. Warning signs to stop exercising while pregnant* tional Institute of Child Health and Human Develop- ment, to document lack of harm to the fetus from the Shortness of breath exercise test [41]. Dizziness Headache (may be early sign of pre-eclampsia or pregnancy-induced Safety of Maximal Exercise Testing Protocols. Maxi- mal testing using short protocols (8–12 min) appears to be hypertension) safe during all stages of pregnancy for both mother and Chest pain fetus [36, 42]. Usual monitoring during maximal testing Muscle weakness of pregnant women should include a 12-lead ECG, heart Calf pain or swelling (need to rule out thrombophelbitis) rate (HR), blood pressure (BP), and rating of perceived Regular good-quality contractions that change the cervix exertion (RPE). When energy expenditure, substrate utili- Decreased fetal movement zation or pulmonary function are of interest, gas exchange Amniotic fluid leakage measurements should be added. The same precautions Generalized swelling (early sign of pre-eclampsia) and indications for stopping an exercise test that are used Pain of hips, back, or symphysis pubis with nonpregnant individuals should be observed (ta- ble 3). * Adapted from Artal and Sherman [10] and Kulpa [55]. Additionally, reasons to stop an exercise test of a preg- walkers or joggers prior to pregnancy. It is also appro- nant woman include the warning signs listed in table 4. priate for more sedentary women who wish to participate in a walking program during pregnancy. When adminis- However, in the absence of signs or symptoms of exer- tered correctly, this protocol results in maximal effort cise intolerance, a pregnant woman should be able to safe- being reached in 8–12 min regardless of initial fitness ly continue a cardiopulmonary exercise test until she level. reaches subjective maximal effort (volitional fatigue). Fe- tal heart rate responses to a short, maximal exercise test Most commonly, maximal exercise tests during preg- have been reported to be minimal and transient. The most nancy use upright leg cycle ergometer protocols. Protocols common response is a mild tachycardia immediately after that have been reported are remarkably similar. After a 2- exercise that returns to baseline levels within 20–30 min to 4-min warm-up during which the patient pedals against [43–48]. Caution is advised, however, that these data are little or no resistance, resistance is increased by 20–30 W/ derived from normal pregnancies in healthy women. min to volitional fatigue [34, 36–40]. Upright leg cycle Screening, including estimates of fetal weight, should be ergometry protocols have been used successfully from 16 made before clearing an apparently healthy, pregnant to 35 weeks of gestation to 12 weeks postpartum. In addi- woman for a maximal exercise test. tion to upright cycling protocols, we have found the leg Clinical Exercise Testing during Pregnancy 277 and the Postpartum Period

Expected Responses to Maximal Exercise Tests during peak lactate in combination with lower respiratory ex- Pregnancy. Studies have compared physiologic responses change ratios suggest that carbohydrate utilization is to maximal cardiopulmonary exercise testing of pregnant reduced, perhaps as a protective mechanism to spare glu- women with responses of nonpregnant women. Both cose for the fetus. A lesser O2 debt (excess postexercise cross-sectional and longitudinal study designs have been oxygen consumption) has been reported and suggests that used to study maximal responses at various times smaller changes in temperature, catecholamines, calcium throughout gestation and postpartum. Some data [34] ions, fatty acids, or restoration of glycogen, ATP and crea- support the use of postpartum values as control values tine phosphate stores have occurred [36]. These responses (i.e. not different from pre-pregnancy values), while oth- may aid in maintenance of fetal well being during strenu- ers [49] suggest that true control values (not recently preg- ous exercise. nant) and postpartum values may result in different out- comes in ventilatory and metabolic comparisons. These Treadmill testing has been used less frequently to assess authors also reported on a sub-sample of women (n = 4) physiologic function at maximal exertion in pregnant who were studied 3 months before gestation to 6 months women. Artal et al. [33] studied responses to maximal postpartum. The average 7-week postpartum values at weight-bearing exercise (treadmill walking) in pregnant maximal exercise were not different from prepregnancy women late in the 2nd trimester or early 3rd trimester in or 6-month postpartum values, suggesting that compari- comparison with non-pregnant women. At maximal exer- son of pregnancy values with either pre- or postpregnancy cise, they found lower tidal volume, VO2max, VCO2max values provides an appropriate control. Upright cycling and RER max in the pregnant women than the controls. In has been the most common testing mode, but data are also contrast, Lotgering et al. [34] reported that maximal power available for treadmill testing and tethered swimming. and maximal oxygen uptake were unchanged by pregnan- cy. It seems likely that these differences may be the result of There is consensus that cycling VO2peak is unaffected inherent differences between cross-sectional [33] and lon- by pregnancy [34, 36, 40, 50–52]. Peak power (Watts) gitudinal [34] study designs. Lotgering et al. [34] reported may be slightly less (4%) at 35 (but not at 15 or 25) weeks that findings for weight-bearing exercise (treadmill) were of gestation in comparison with 7 weeks postpartum [34]. similar to those for non-weight-bearing exercise (cycle Maximal heart rate may be unchanged [36] or slightly ergometry). One observed difference was that VCO2max lower [34, 40] during pregnancy. Efficiency of leg cycling appeared to be more blunted during treadmill exercise at maximal effort appears to be unchanged during preg- causing RERmax to be lower during treadmill exercise than nancy [34, 36]. Cross-sectional analyses suggest that maxi- cycle ergometry. Artal et al. [33] noted that the ventilatory mal CO2 production (VCO2max) is slightly lower and may equivalent for oxygen was statistically unchanged, but that be sufficiently low to decrease maximal respiratory ex- there was a consistent trend for it to be slightly increased change ratio (RER max) [36]. Longitudinal analysis dem- (fig. 1). Thus, data from both studies suggest that maximal onstrated that VCO2max progressively decreased (6, 9 weight-bearing exercise results in greater augmented respi- and 11% at 15, 25 and 35 weeks of gestation) during preg- ratory sensitivity in comparison to non-weight-bearing nancy [34]. RER max appears to be approximately 5–7% exercise. As would be expected from nonpregnant re- lower late in pregnancy in comparison with nonpregnant sponses, values for all other physiologic variables studied conditions [34, 36]. Maximal minute ventilation can be were higher during treadmill exercise than during cycle expected to be approximately 5–7% greater during preg- ergometry [34]. nancy, mainly as a result of a higher maximal tidal vol- ume and no change in breathing frequency. Thus, the ven- Two studies have investigated the physiologic re- tilatory equivalent of oxygen (VE/VO2) should increase sponses to swimming during pregnancy in comparison to and the ventilatory equivalent of carbon dioxide (VE/ postpartum responses [40, 50]. Spinnewijn et al. [40] VCO2) should decrease. This was documented in a longi- reported no difference in VO2peak or HR peak between tudinal study [34], but was not evident in data from a pregnancy and postpartum. As with other modes of exer- recent, cross-sectional study [36]. In the cross-sectional cise, VCO2peak was lower (F10%) during pregnancy and study, none of the respiratory gas exchange variables that resulted in a lower RER peak. Both VE peak (F4%) and were measured (VE, PETO2, PETCO2) were different be- Vt peak (F8%) were slightly increased, but statistically tween pregnant and nonpregnant women at maximal unchanged. Also similar to results with other exercise exercise [36]. Peak lactates have been reported to be modes, peak lactates were lower during pregnancy. decreased in pregnant women [36, 39, 50, 53]. The lower McMurray et al. [50] had previously reported marked dif- ferences in responses (17% lower VO2 peak to swimming 278 O’Toole/Artal

Fig. 1. Ventilatory equivalent of oxygen (VE/VO2) before and during symptom-limited treadmill walking to maximal effort in pregnant vs. nonpregnant controls. From Artal et al. [33]. during pregnancy, but methodical differences in testing gle stage can be used. During a single-stage test, the partic- protocols likely caused these apparent differences. Spin- ipant exercises on an ergometer (treadmill, leg cycle newijn et al. [40] reported that during pregnancy, swim- ergometer, etc.) for a prolonged period of time at a single, ming VO2peak was 11% lower and VCO2peak was 25% submaximal exercise intensity. Measures of cardiovascu- lower in comparison with cycling values. In accord with lar variables such as heart rate and blood pressure are the these lower values, RER peak, VE peak, Vt peak and peak mainstays of single-stage tests and are assessed over time. lactate were also lower in swimming than cycling. These Other uses include tracking substrates (e.g. glucose, lac- responses are similar to those expected in nonpregnant tate, fatty acids), electrolytes (e.g. sodium, potassium) individuals. and/or hormones (e.g. insulin, epinephrine) for a pro- longed period of time. Single-stage tests may also be used Submaximal Exercise Tests to measure energy expenditure (caloric cost) of a specific Protocol Selection. Not all cardiopulmonary exercise exercise intensity. They may also be used to assess the rel- testing of pregnant women needs to be maximal exercise ative contribution of carbohydrate and fat as substrate testing. Submaximal tests are appropriate when informa- during that exercise intensity and may be particularly use- tion about safety or the impact of a therapeutic interven- ful in quantifying caloric expenditure of an activity. tion at a particular exercise intensity is required. Steady- state responses reflect homeostatic adjustment to a sub- Expected Responses to Submaximal Exercise Tests maximal exercise level and can be used to assess appro- during Pregnancy. Comparisons of physiologic responses priateness of cardiovascular, pulmonary or metabolic re- to submaximal exercise during pregnancy with responses sponses. Several submaximal stages can be used or a sin- by nonpregnant controls suggest that few differences can be expected during non-weight-bearing exercise when in- Clinical Exercise Testing during Pregnancy 279 and the Postpartum Period

dividuals are carefully matched. In active women (all reg- of engorged breasts [55]. Additionally, nursing before ularly exercised 3–6 times per week), ventilatory thresh- exercise will avoid the potential problems associated with old during cycle ergometry was not different between increased acidity of milk secondary to any build-up of lac- pregnant and nonpregnant groups, nor was the respiratory tic acid. A recent study of the short-term effects of dieting compensation threshold or work efficiency [36, 39]. Other plus aerobic exercise on lactation performance concluded physiologic responses at single submaximal exercise in- that this was a safe practice that resulted in optimal tensity, defined as an oxygen uptake of 100 ml less than weight loss without affecting lactation performance [56]. the ventilatory threshold, were compared in these sub- Finally, a return to physical activity following pregnancy jects. There was no significant effect of pregnancy on car- has been associated with decreased postpartum depres- diovascular (HR), metabolic (VO2) or respiratory re- sion, but only if the exercise is stress relieving and not sponses except for significantly increased ventilatory stress provoking [57]. equivalents for oxygen and carbon dioxide and signifi- cantly decreased end tidal and arterial carbon dioxide lev- Conclusions els in the pregnant women [36]. These data support the well-documented increase in respiratory drive during Pregnancy is associated with profound anatomical and pregnancy and demonstrate that the increased drive con- physiological changes that alter cardiopulmonary re- tinues to be operative during submaximal exercise. sponses to clinical exercise testing. Nevertheless, exercise tests can be used during pregnancy for the same reasons Exercise Testing and Prescription during the they are used in a nonpregnant population (e.g. diagnostic Postpartum Period reasons, development of individualized exercise prescrip- tions, and monitoring progress in a training program) as Many of the physiologic and morphologic changes of long as appropriate testing modalities and protocols are pregnancy persist 4–6 weeks postpartum. Thus, pre-preg- used. Semirecumbent cycle ergometry appears to be the nancy exercise routines should be resumed gradually most comfortable modality, but treadmill walking or based on a woman’s physical capability. The competitive upright cycle ergometry may also be used. Short protocols athlete with an uncomplicated pregnancy may resume (8–12 min) to maximal subjective effort appear to be safe training as early a 2 weeks postpartum. No known mater- during all stages of pregnancy for both mother and fetus. nal complications are associated with resumption of train- Single-stage submaximal protocols also provide useful ing [54]. There are, however, anecdotal reports of failure information particularly when therapeutic interventions, of infants to gain weight as rapidly as expected in strenu- substrate utilization (especially for diabetics), or hormon- ously training mothers. Failure to gain weight is associat- al responses are of interest. Individuals responsible for the ed with decreased milk production that may be secondary conduct and interpretation of exercise testing in pregnant to inadequate fluid or nutritional intake to balance train- women should be aware of expected responses to submax- ing-induced outputs. Nursing women should feed the imal and maximal exercise during pregnancy. infant before exercising in order to avoid the discomfort References 4 Wolfe L: Pregnancy; in Skinner JS (ed): Exer- 8 Ratigan TR: Anatomic and physiologic cise Testing and Exercise Prescription for Spe- changes of pregnancy: Anesthetic consider- 1 Girandola RN, Khodiguian N, Artal R, Wis- cial Cases, ed 2. Philadelphia, Lea & Febiger, ations. J Am Assoc Nurse 1983;51:38–42. well RA: Body composition in pregnancy; in 1993, pp 363–385. Artal R, Wiswell RA, Drinkwater BL (eds): 9 Gee JB, Packer BS, Millen JE, Robin ED: Pul- Exercise and Pregnancy. Baltimore, Williams 5 Artal R, Friedman MJ, McNitt-Gregg JL: Or- monary mechanics during pregnancy. J Clin & Wilkins, 1991, pp 99–108. thopedic problems in pregnancy. Physician Invest 1967;46:945–952. Sportsmed 1990;18:93–105. 2 Brown JE: Nutrition for Your Pregnancy. Min- 10 Artal R, with Sherman C: Exercise during preg- neapolis, University of Minnesota Press, 1983. 6 Alaily AB, Carroll KB: Pulmonary ventilation nancy: Safe and beneficial for most. Physician in pregnancy. Br J Obstet Gynaecol 1978;85: Sportsmed 1999;27(8):51–75. 3 Palin D, Rankine D: Nutrition in pregnancy – 518–524. national guidelines: A summary. J Can Diet Assoc 1987;47:209. 7 Knuttgen HG, Emerson K Jr: Physiological response to pregnancy at rest and during exer- cise. J Appl Physiol 1974;36:549–553. 280 O’Toole/Artal

11 Artal R, Wiswell RA, Drinkwater BL (eds): 28 Goodrich SM, Wood JE: The effect of estra- 44 Artal R, Rutherford S, Romem Y, Kammula Exercise in Pregnancy, ed 2. Baltimore, Wil- diol-17beta on peripheral venous distensibility RK, Dorey FJ, Wiswell RA: Fetal heart rate liams & Wilkins, 1991. and velocity of venous blood flow. Am J Obstet responses to maternal exercise. Am J Obstet Gynecol 1966;96:407–412. Gynecol 1986;155:729–733. 12 Pernoll ML, Metcalfe J, Kovach PA, Wachtel R, Dunham MJ: Ventilation during rest and 29 American College of Sports Medicine: ACSM’s 45 Watson WJ, Katz VL, Hackney AC, Gall MM, exercise in pregnancy and postpartum. Resp Guidelines for Exercise Testing and Prescrip- McMurray RG: Fetal responses to maximal Physiol 1975;25:295–310. tion, ed 6. Philadelphia, Lippincott Williams & swimming and cycling exercise during preg- Wilkins, 2000, pp 36, 50, 80. nancy. Obstet Gynecol 1991;77:382–386. 13 Romem Y, Masaki DI, Artal R: Physiological and endocrine adjustments to pregnancy; in 30 Canadian Society for Exercise Physiology: 46 Webb KA, Wolfe LA, McGrath MJ: Effects of Artal R, Wiswell RA, Drinkwater BL (eds): Physical Activity Readiness Medical Examina- acute and chronic maternal exercise on fetal Exercise and Pregnancy. Baltimore, Williams tion for Pregnancy. Ottawa, Canadian Society heart rate. J Appl Physiol 1994;77:2207–2213. & Wilkins, 1991, pp 9–29. for Exercise Physiology, 1996. 47 O’Neill ME: Maternal rectal temperature and 14 Clark SL, Cotton DB, Lee W, Bishop C, Hill T, 31 American College of Obstetricians and Gyne- fetal heart rate responses to upright cycling in Southwick J, Pivarnik J, Spillman T, DeVore cologists: Exercise during Pregnancy and the late pregnancy. Br J Sports Med 1996;30:32– GR, Phelan J: Central hemodynamic assess- Postpartum Period, Tech Bull #189. Washing- 35. ment of normal term pregnancy. Am J Obstet ton, American College of Obstetricians and Gynecol 1989;161:1439–1442. Gynecologists, 1994. 48 Manders MAM, Sonder GJB, Mulder EJH, Visser GHA: The effects of maternal exercise 15 Wolfe LA, Ohtake PJ, Mottola MF, McGrath 32 Jovanovic-Peterson L, Durak EP, Peterson on fetal heart rate and movement patterns. Ear- MJ: Physiological interactions between preg- CM: Randomized trial of diet versus diet plus ly Hum Dev 1997;48:237–247. nancy and aerobic exercise. Exerc Sport Sci cardiovascular conditioning on glucose levels Rev 1989;17:295–351. in gestational diabetes. Am J Obstet Gynecol 49 Jaque-Fortunato SV, Wiswell RA, Khodiguian 1989;161:415–419. N, Artal R: A comparison of the ventilatory 16 Morton MJ: Maternal hemodynamics in preg- responses to exercise in pregnant, postpartum nancy; in Artal R, Wiswell RA, Drinkwater BL 33 Artal R, Wiswell R, Romem Y, Dorey F: Pul- and nonpregnant women. Seminar Perinatol (eds): Exercise and Pregnancy. Baltimore, Wil- monary responses to exercise in pregnancy. Am 1996;20:263–276. liams & Wilkins, 1991. J Obstet Gynecol 1986:154:378–383. 50 McMurray RG, Hackney AC, Katz VL, Gall 17 Pirani BB, Campbell DM, MacGillivray I: 34 Lotgering FK, Van Doorn MB, Struijk PC, M, Watson WJ: Pregnancy-induced changes in Plasma volume in normal first pregnancy. J Pool J, Wallenburg HCS: Maximal aerobic ex- the maximal physiological responses during Obstet Gynaecol Br Commonw 1973;80:884– ercise in pregnant women: Heart rate, O2 con- swimming. J Appl Physiol 1991;71:1454– 887. sumption, CO2 production, and ventilation. J 1459. Appl Physiol 1991;70:1016–1023. 18 Lund CJ, Donovan JC: Blood volume during 51 Sady SP, Carpenter MW, Thompson PD, Sady pregnancy. Am J Obstet Gynecol 1967;98:394– 35 Blair SN, Applegate WB, Dunn AL, Ettinger MA, Haydon B, Coustan DR: Cardiovascular 403. WH, Haskell WL, King AC, Morgan TM, Shih response to cycle exercise during and after JA, Simons-Morton DG: Activity Counseling pregnancy. J Appl Physiol 1989;66:336–341. 19 Scott DE: Anemia in pregnancy. Obstet Gyne- Trial (ACT): Rationale, design, and methods. col Annu 1972;1:219–244. Med Sci Sports Exerc 1998;30:1097–1106. 52 Sady MA, Haydon BB, Sady SP, Carpenter MW, Thompson PD, Coustan DR: Cardiovas- 20 Longo LD: Maternal blood volume and cardiac 36 Heenan AP, Wolfe LA, Davies GAL: Maximal cular response to maximal cycle exercise during output during pregnancy: A hypothesis of en- exercise testing in late gestation: Maternal re- pregnancy and at two and seven months post- docrinologic control. Am J Physiol 1983;245: sponses. Obstet Gynecol 2001;97:127–134. partum. Am J Obstet Gynecol 1990;162:1181– R720–R729. 1185. 37 South-Paul JE, Rajjagopal KR, Tenholder MF: 21 Bader RA, Bader ME, Rose PJ, Braunwald E: The effect of participation in a regular exercise 53 Clapp JF III, Wesley M, Sleamaker RH: Ther- Haemodynamics at rest and during exercise in program upon aerobic capacity during preg- moregulatory and metabolic responses to jog- normal pregnancy as studied by cardiac cathe- nancy. Obstet Gynecol 1988;71:175–179. ging prior to and during pregnancy. Med Sci terization. J Clin Invest 1955;34:1524. Sports Exerc 1987;19:124–130. 38 Lotgering FK, Struijk PC, Van Doorn MB, 22 Wilson M, Morganti AG, Zervoudakis I, Letch- Spinnewijn WEM, Wallenburg HCS: Anaero- 54 Hale RW, Milne L: The elite athlete and exer- er RL, Romney BM, Von Oeyon P, Papera S, bic threshold and respiratory compensation in cise in pregnancy. Semin Perinat 1996;20:277– Sealey JE, Laragh JH: Blood pressure, the un- pregnant women. J Appl Physiol 1995;78: 284. ine-aldosterone system and sex steroids 1772–1777. throughout normal pregnancy. Am J Med 55 Kulpa P: Exercise during pregnancy and post 1980;68:97. 39 Wolfe LA, Walker RMC, Bonen A, McGrath partum; in Agostini R (ed): Medical and Or- MJ: Effects of pregnancy and chronic exercise thopedic Issues of Active and Athletic Women. 23 Pivaranik JM: Cardiovascular responses to on respiratory responses to graded exercise. J Philadelphia, Hanley & Belfus, 1994, pp 191– aerobic exercise during pregnancy and postpar- Appl Physiol 1994;76:1928–1936. 199. tum. Semin Perinatol 1996;20:242–249. 40 Spinnewijn WEM, Wallenburg HCS, Struijk 56 McCrory MA, Nommsen-Rivers LA, Mole PA, 24 Katz R, Karliner JS, Resnik R: Effects of a nat- PC, Lotgering FK: Peak ventilatory responses Lonnerdal B, Dewey KG: Randomized trial of ural volume overload state (pregnancy) in left during cycling and swimming in pregnant and short-term effects of dieting compared with ventricular performance in normal human sub- nonpregnant women. J Appl Physiol 1996;81: dieting plua aerobic exercise on lactation per- jects. Circulation 1978;48:434. 738–742. formance. Am J Clin Nutr 1999;69:959–967. 25 Mashini IS, Albazzaz SJ, Fadel HE, Abdulla 41 Electronic fetal heart rate monitoring: Re- 57 Koltyn KF, Schultes SS: Psychological effects AM, Hadi HA, Harp R, Devoe LD: Serial non- search guidelines for interpretation. National of an aerobic exercise session and a rest session invasive evaluation of cardiovascular hemody- Institute of Child Health and Human Develop- following pregnancy. J Sports Med Phys Fit- namics during pregnancy. Am J Obstet Gyne- ment Research Planning Workshop. Am J Ob- ness 1997;37:287–291. col 1987;156:1208–1213. stet Gynecol 1997;177:1385–1390. Mary L. O’Toole, PhD 26 Rubler S, Damani PM, Pinto ER: Cardiac size 42 Macphail A, Davies GAL, Victory R, Wolfe Department of Obstetrics, Gynecology and and performance during pregnancy estimated LA: Maximal exercise testing in late gestation: Women’s Health with echocardiography. Am J Cardiol 1977;40: Fetal responses. Obstet Gynecol 2000;96:565– 6420 Clayton Road, Suite 290 534–540. 570. St. Louis, MO 63117 (USA) Tel. +1 314 781 2543, Fax +1 314 645 6173 27 Goodrich SM, Wood JE: Peripheral venous 43 Wolfe LA, Brenner IKM, Mottola MF: Mater- E-Mail [email protected] distensibility and velocity of venous blood flow nal exercise, fetal well-being, and pregnancy during pregnancy or during oral contraceptive outcome. Exerc Sport Sci Rev 1994;22:145– therapy. Am J Obstet Gynecol 1964;90:740. 194. Clinical Exercise Testing during Pregnancy 281 and the Postpartum Period

Weisman IM, Zeballos RJ (eds): Clinical Exercise Testing. Prog Respir Res. Basel, Karger, 2002, vol 32, pp 282–299 OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO Clinical Exercise Testing in Children John Faheya Dan Nemetb Dan M. Cooperb aDepartment of Pediatrics, Section of Cardiology, Yale University School of Medicine, New Haven, Conn., and bCenter for the Study of Health Effects of Exercise in Children, University of California, Irvine, College of Medicine, Orange, Calif., USA Summary disability. We now believe that exercise in children is not merely play, but rather, is essential for healthy growth and In children, exercise is not merely play but a fundamental development. There is mounting research identifying biologic regulator of growth and development. Recognizing physiological, molecular, and structural mechanisms that this, the pediatric exercise laboratory will need to focus not link exercise with processes of growth and development in only on traditional cardiorespiratory responses to exercise, health and disease. Moreover, there appear to be signifi- but also on testing strategies that gauge the dynamic relation- cant long-term health consequences for the child who is ship between exercise and growth. The pediatric exercise lab- unable to optimally engage in physical activity because of oratory performs critical diagnostic procedures relevant to environmental, social and psychological, or physiological childhood lung (e.g. cystic fibrosis or asthma), cardiac (e.g. con- barriers. In this chapter, we review essential knowledge genital heart disease), and metabolic diseases (e.g. diabetes). In that links physical activity with growth and that provides addition, exercise testing in children will increasingly be relied the basis for innovative uses of exercise in children from upon to develop exercise therapies for children with condi- both a diagnostic and a therapeutic perspective. tions ranging from obesity to mitochondrial disease. The clini- cian referring a child for pediatric exercise consultation must Exercise and Physical Activity in Healthy Children gain insight not only into the child’s functional capability at the moment, but also, whether or not the level of physical activity It is now known that exercise in healthy younger chil- of his or her patient optimizes the many long-term benefits of dren is characterized by many brief pulses (about 85 per exercise for the overall process of growth and development. hour, mean of 20 s duration) that vary greatly in intensity [1]. The majority of these exercise bouts are relatively low Introduction intensity, but in about 20%, the exercise intensity is likely to be above the anaerobic or lactate threshold [2–4] Children are, arguably, the most naturally physically (fig. 1). During these high-intensity exercise bouts, pertur- active human beings. However, surprisingly little is yet bations in circulating levels of growth hormone (GH) [5], understood about the role of exercise in: (1) the growth insulin-like growth factor-I (IGF-I) [6], and inflammatory and development of healthy children; (2) gaining diagnos- mediators (e.g. interleukin-6) might occur [7]. Current tic insight into pediatric pathophysiology, and (3) provid- research suggests that physical activity may influence ing novel therapies for children with chronic disease and growth and development by activating these anabolic and catabolic mediators.

Fig. 1. Representative example of activity patterns derived from physiologically based exercise testing was derived from direct observation in a 7-year-old boy over a 24-min period. Bouts of efforts to understand maximal exercise response for sev- physical activity are randomly spaced. Most bouts were in the low- eral reasons. First, in the early part of the 20th Century, intensity range (below the LT – the dashed line, estimated for chil- biomedical scientists were occupied with discovering how dren this age); however, a substantial portion of the energy expendi- to make better soldiers and physical laborers – activities ture due to physical activity results from the less frequent, high-inten- that required substantial energy expenditure. Secondly, sity bouts. Reprinted with permission from Berman et al. [1]. observations from the same time period showed that humans achieved a true maximal oxygen uptake (VO2) With very few exceptions, all children, whether or not (i.e. a point at which the individual could increase exter- they have a chronic illness, should be encouraged to be nal work with no accompanying increase in oxygen up- active and fit. But what has been difficult to do is to arrive take). This was very intriguing to investigators who felt at physiologically sound definitions for ‘fitness’ in both that the VO2max must indicate critically important fea- the healthy child and in the child with chronic disease. In tures of the physiological response to exercise. this context, laboratory tests of exercise capability in chil- dren have become increasingly relevant. Exercise testing As a consequence, equipment and testing protocols can reproducibly measure respiratory, cardiac, and meta- designed for adults are still frequently used to test chil- bolic function under stress, and the values derived from a dren in many laboratories. There are, however, several given child can then be compared with normal data. This important considerations that require different testing information can be used to better define the exercise ‘pre- approaches in evaluating children. The laboratory must scription’ because the initial response characteristics be able to accommodate large discrepancies in subject serve as baseline values to gauge the later effects of train- age, size, strength, coordination, fitness level and atten- ing or other therapies. Other fundamental questions that tion span. The personnel must be comfortable in working can be answered by exercise testing are: (1) at what level of with younger children and adolescents. The 6-year-old intensity does exercise become dangerous; (2) what are boy may be eager to perform, but intimidated and scared the mechanisms that limit physical activity in a particular by the equipment and monitors; the 16-year-old girl may child, and (3) to what extent can these limitations be ame- want to do her best, but worried about how much sweat liorated by exercise itself? the exercise testing bout may produce. Pediatric exercise laboratories must have a variety of innovative protocols Approaches to Exercise Testing in Children and equipment that can be readily modified. Cycle er- gometers must have adjustable seats, handle bars, and Although children are naturally active, exercise testing pedals. The treadmills must have additional or adjustable in children and adolescents was initially an outgrowth of side rails. Finally, it must be recognized that the VO2max the larger experience in adults. The history of modern test which focused on peak function may not be suitable to determine limitations in function. Ironically, determin- ing impaired metabolic and physiologic responses to exer- cise is a major goal of exercise testing in the child with chronic disease or disability. In addition, the physiologic responses of children dur- ing exercise are different than in adults, and, indeed, mature. ‘Normal’ values must be found that adequately account for growth related changes in body mass, intrinsic muscle function, and neuromotor maturation. An over- view of some valid normal values for exercise testing in children can be found in a variety of sources [8, 9]. Ulti- mately, each pediatric exercise laboratory must ensure that any reference value for normal values is validated and appropriate for the ethnic and socioeconomic mix of children tested in a particular geographical area. Ideally, the pediatric exercise laboratory should be a separate facility, with dedicated personnel and equipment, pre- senting a ‘child friendly’ environment. Clinical Exercise Testing in Children 283

Fig. 2. Relationship between VE-VCO2 slope during exercise and age in children and adolescents. These data confirm earlier observations from Cooper et al. [3]. Ventilation per VCO2 production is greater in younger children. Reprinted with permission from Nagano et al. [20]. Maturation of the Physiologic Responses of spite of this, children typically have higher heart rates Children to Exercise during exercise when compared to adults. Young children have higher resting heart rates than adults, with resultant Research over the past 50–70 years has demonstrated elevation of the exercise heart rates. Further, children key areas in which the cardiorespiratory responses to have significantly higher maximal heart rates, in the range exercise differ between prepubertal children and adults. of 200–215 beats/min, persisting until late adolescence These include: (thereafter, maximal heart rates tend to decrease with age). For all of these reasons, heart rates tend to be signifi- (1) Reduced mechanical efficiency and increased oxy- cantly higher during pediatric exercise tests and should gen cost of work [10, 11]. not be interpreted as a sign of distress or dysfunction. As in adults, the systolic blood pressure rises progressively (2) Generally faster oxygen uptake, heart rate, and ven- during progressive exercise, while the diastolic blood pres- tilatory kinetics at the onset of and in recovery from a sure either stays the same or only slightly increases. In bout of exercise [11]. general, the systolic blood pressure rarely exceeds 200 mm Hg in a healthy pediatric population [14]. (3) Inability to achieve as high a level of lactate and hydrogen ion concentration (or to increase intramuscular The ventilatory response to exercise is also different in Pi/PCr ratios by magnetic resonance spectroscopy) [12]. children. Ventilatory responses are appropriately normal- ized to CO2 production since it is the latter that is physio- (4) Increased ventilatory cost of carbon dioxide pro- logically driving ventilation during exercise. Numerous duction [13]. studies of exercise in children (including our own) have clearly demonstrated that the slope of the relationship As for adults, in order for children to perform progres- between ventilation and CO2 production (fig. 2) is greater sively increasing workload protocols, there must be pro- in children compared with adults. The mechanism for this gressive, proportional increases in both cardiac output is likely related to a lower CO2 set point, relatively less and ventilation in order to deliver oxygen to the exercis- CO2 storage in the body fat and muscle, and altered sensi- ing muscles and eliminate carbon dioxide from the ve- tivity of respiratory control centers [10, 13, 19, 20]. nous blood. Similar to adult responses, children can achieve 5-fold increases in cardiac output [14–16], and Like adults, children will achieve a lactate or anaerobic 12-fold increases in minute ventilation [13, 17] over rest- threshold [3], but because of their relatively inability to ing values. However, the changes that produce these produce large concentrations of lactate [21] and to lower increases are quite different. Stroke volume, when in- dexed to body surface area, seems to be the same in chil- dren and adults both at rest and during exercise [18]. In 284 Fahey/Nemet/Cooper

pH, relatively greater respiratory ‘noise’, and their in- Noninvasive cardiac output and stroke volume mea- creased VE relative to VCO2, it is more technically diffi- surements during exercise are possible in pediatric pa- cult in children to precisely determine the threshold using tients, although used more for research than clinical stud- noninvasive gas exchange measurements (i.e. the slope ies. The techniques are the same as used for adults. Both changes in R, PetO2, or VE/VO2 are simply not as great as CO2 rebreathing and acetylene-helium rebreathing tech- in adults). Similarly, while in adults, oxygen uptake con- niques have been employed successfully in children [23, tinues to increase even for constant work rate, above-AT 24]. These require the absence of significant ventilation- exercise, the slope of the oxygen uptake response for perfusion mismatching, as well as no intracardiac or intra- above-AT constant work rate exercise (phase 3) in chil- pulmonary shunting; both of which significantly limit the dren is quite low and often difficult to observe [10]. usefulness of this type of measurement in the pediatric Although the precise mechanisms for these maturational laboratory. In addition, these techniques also require a differences in lactate kinetics are not known, some work- high degree of patient cooperation. Other techniques may ers have suggested that immature anaerobic glycolytic have more use for the pediatric patient. Doppler and two- pathways may well lead to the relative inability of chil- dimensional echocardiographic measurements are fre- dren to produce lactate at levels equivalent to those found quently easier in children since they tend to have better in adults. acoustic windows [25]. Electrical bioimpedance measure- ments have recently been significantly improved, require Exercise Measurements no patient cooperation, and may be used in children as small as 15 kg, but are awaiting correlation with more At least three leads of the standard surface electrocar- accepted techniques. diogram should be routinely and continuously monitored, and the examiner should have the ability to view various Oxygen saturation measurements have become almost combinations of the twelve available leads. Lead place- a routine part of exercise testing, especially in pediatric ment is the same in children as in adults, as is the skin pulmonary and cardiac patients, since the introduction of preparation. Smaller leads are available for children to pulse oximetry. Both finger and ear oximeters have been avoid overlap or misplacement. The heart rate is mea- used, but both have the disadvantage of being inaccurate sured from the electrocardiogram. Blood pressure is mea- at higher work rates due to motion artifact or tight grip- sured at each level of exercise using an appropriately sized ping of the handlebars during cycle ergometry. These blood pressure cuff. The recommendation is that the blad- problems can be largely overcome using a reflectance oxy- der of the cuff completely encircles the arm, and the cuff gen probe, which is placed over the temporal artery on the should cover two-thirds of the distance from the antecubi- forehead and held in place with a headband [26]. tal fossa to the axilla. Too small a cuff will produce falsely elevated measurements, whereas too large a cuff will min- Protocols for Exercise Testing imally depress the blood pressure measurement [22]. Ob- viously, a large assortment in cuff sizes must be available, There is no single, standardized exercise protocol rou- and the examiner should err on the side of too large a cuff tinely used or recommended for exercise testing in the if there is a question. pediatric population, and with good reason. The protocol must be selected, and sometimes modified, based on the The use of respiratory gas analyzers has become rou- information required, as well as the age, health and fitness tine in pediatric exercise laboratories. Variables such as level of the subject. There are two general types of exercise VO2, VCO2, ventilation, tidal volume, and respiratory testing: (1) progressively increasing workloads to maxi- rate can be monitored continuously during the tests, on a mum, with no rest period between changes in work incre- breath-by-breath basis. Mouthpieces are available in a ments, and (2) a brief constant work rate [8]. variety of sizes, allowing metabolic measurements in chil- dren as young as 6–7 years of age. Tightly fitting masks Progressively increasing workloads to maximum are are also available for small children who do not tolerate the most common in use in pediatric laboratories, and the mouthpiece or nose clip, but the tendency for leaks is may use either a treadmill or cycle ergometer. There are greater. Equipment to perform resting and postexercise many benefits to these protocols and they can answer a spirometry, and inspiratory and expiratory flow volume variety of clinical questions. They can measure the level loops can usually be easily obtained. of aerobic fitness (VO2max). Maximal stress yields confi- dence about the safety of strenuous exercise for patients, Clinical Exercise Testing in Children 285

parents, and caretakers. They maximally stress the respi- cise in children with a variety of diseases and disabilities ratory system, and can precipitate both exercise-induced (see below). It has been argued, for example, that onset stridor and/or bronchospasm. They also maximally stress and offset gas exchange and heart rate kinetics (i.e. the the heart by maximizing myocardial oxygen demands (i.e. time required for physiological systems to respond to and maximal rate-pressure product). recover from exercise) are as important as maximal values in assessing impairment. Brief exercise protocols more There are, however, drawbacks, limitations, and criti- closely mimic exercise patterns of children and are more cisms of this form of maximal exercise testing. It is felt to likely to be tolerated than either maximal exercise tests or be an ‘unnatural’ form of exercise, and not representative long bouts of sustained, heavy exercise. This approach has of patterns of physical activity of normal children [27]. been used successfully in children with chronic disease Neither children, adults, nor virtually any mammal nor- (fig. 3, 4). mally exercise to maximum, and VO2max may only occur in the exercise laboratory. The study requires cooperation Recovery kinetics may determine the child’s ability to on the part of the patient, substantial coaxing on the part do repetitive bouts of exercise, and therefore be more pre- of the laboratory technician, and a true maximal study is dictive of functional ability. This form of testing is not as very difficult to obtain in the young child, either on the stressful to the children and they are easier for the chil- treadmill or the cycle. In fact, only about 20–30% of chil- dren to perform [27]. It may be that combinations of sev- dren actually achieve a true plateau of oxygen uptake, eral types of exercise testing will be needed for the long- diagnostic of VO2max. Finally, in children with chronic term management of children with significant cardiac or heart or lung disease, it is the rare laboratory technician pulmonary disease (fig. 5–7). who will cajole and coax the subjects to the same extent as he or she would a healthy child, particularly in the higher Pediatric stress tests can be performed using either the exercise intensities where pH is lower and the child is like- cycle or the treadmill. Cycle ergometers are more com- ly to experience uncomfortable levels of dyspnea and monly used for pediatric stress tests than for adult tests. fatigue. Many laboratories have both available since each has advantages and disadvantages. Cycle ergometers have Although the clinical utility of maximal values can be several significant advantages for clinical exercise testing. questioned, much useful information is available from Due to the stability of the trunk and arms, there tends to progressive exercise protocols. Gas exchange and HR be less artifact in the measurement of the ECG and blood responses continuously change during progressive tests, pressure. It is easier to do metabolic measurements as the and often the relationships between these changing vari- mouthpiece can be supported by a mechanical arm and ables (referred to as dynamic relationships) can be quanti- the children do not bear the weight or have to wear a cum- fied using straightforward analytic techniques. As noted bersome halo device. There is easier access to the patient, above, the LAT in children and young adults can be deter- making cardiac output and echocardiographic measure- mined noninvasively from the dynamic responses of gas ments more convenient. The cycle is also easier and quiet- exchange during progressive exercise [3, 28]. Simple er with essentially no risk of injury to the patient and is linear regression analysis of HR and VO2 can provide preferred if laboratory space is limited. The seat height noninvasive indicators of cardiac function [29–31]. The and handle bar height are adjustable. However, it is fre- slope of the regression changes systematically with matu- quently inconvenient to change the length of the pedal ration and with diseases (e.g. congenital heart disease arm. Many children less than 6 years of age cannot sustain [32]). Similarly, linear regression analysis of VE and VCO2 pedaling at even the lowest work rates, and a treadmill can be used to assess respiratory efficiency and control must be used. A final but critical point is that with the during exercise [33]. The real clinical value of progressive cycle ergometer, the external work performed is known exercise testing for children may prove to be in the rich precisely while on the treadmill, the actual work done can cardiorespiratory data obtained during the submaximal at best only be estimated. phases rather than from the final, single, data point at peak or maximal power. Treadmills have the advantage that children can easily walk and then run at a slow pace. Meaningful tests can be There is to date less experience with brief, constant performed on children as young as three years of age. As work rate tests [32]. Nonetheless, there are practical bene- more muscle groups are used in running as opposed to fits to these protocols, and there is a growing body of cycling, VO2max values tend to be about 10% higher on research demonstrating the efficacy of these protocols in the treadmill when compared to the cycle ergometer. identifying impaired cardiorespiratory responses to exer- There are several disadvantages to the treadmill. Safety 286 Fahey/Nemet/Cooper

Fig. 3. Comparison among total work per- formed, total work per body weight, and heart rate by end-exercise in controls and CF subjects using multiple, constant work rate protocols scaled to each child’s LT. Total work performed was significantly higher in controls (* p ! 0.001). However, both CF (i.e. with and without nonsteroidal anti- inflammatory agents – NSAIDS) and con- trol groups reached the same heart rate by end-exercise. No effect of ibuprofen use was observed in the CF subjects. Data reprinted with permission from Tirakitsoon- torn [2001]. Fig. 4. Effect of exercise on plasma lactate and serum GH in both concerns are greater on the treadmill due to the risks of controls and CF subjects. GH and lactate levels increased significantly falling on the moving belt. An observer must be positioned during exercise and were the same in both control and CF groups. No behind young children to support them if they fall or stum- effect of ibuprofen was observed in the CF subjects. This approach ble. Increased body movement leads to a great deal of arti- demonstrates the potential utility of brief, scaled, constant work rate fact on the ECG. The blood pressure is harder to measure protocols in gauging metabolic responses to exercise in children. Data when the patient is running. Treadmills are frequently reprinted with permission from Tirakitsoontorn [2001]. loud and threatening to young children. Treadmills are more expensive and require more space in the lab. There are a variety of standardized protocols that have been used for cycle ergometry in children. As noted, most pediatric exercise laboratories use progressive protocols (incremental or ramp-type) in which the child exercises to the limit of his or her tolerance. One should gauge the rate of work rate increase so that the total time on the cycle ergometer should be about 10–14 min, allowing sufficient data density for accurate analysis. In healthy children, the rate of increase might range from 5–10 W/min in 6-year- olds to 15–25 W/min in 18-year-olds. As clinicians gain experience, their ability to select the ‘right’ work rate increment improves. The Bruce protocol is the most commonly used tread- mill protocol in the pediatric exercise laboratory [34]. Developed for adults, this protocol consists of 3-min stages with an increase at each stage in both the speed and grade of the treadmill. There is no modification in the protocol for age or size. Most pediatric laboratories have modified the Bruce protocol for small children by includ- ing two stages at the beginning of the protocol with low belt speed and lower grades. Another treadmill protocol used for children is the James protocol [35]. As for the James protocol, the 3-min stages make metabolic mea- surements difficult. Children also find the 3-min stages at Clinical Exercise Testing in Children 287

Fig. 5. Relationship between VO2 and HR during progressive exer- Fig. 6. HR response before, during, and in the recovery from 1 min of cise in individual 12-year-old control and Fontan subjects. In both exercise in a 14-year-old Fontan subject. The recovery kinetics were cases, there was a characteristic largely linear relationship between quantified using a single exponential as shown. Data reprinted with the two variables (solid lines indicate best-fit lines by linear regres- permission from Troutman et al. [32]. sion). As demonstrated in these two subjects, the slope of the rela- tionship (VO2/HR) was lower in the Fontan patients. Data reprinted with permission from Troutman et al. [32]. the lower work levels boring and become distracted. The studies are excessively long in highly fit individuals due to the nonstressful early stages. The Balke protocol uses a constant treadmill speed with increases in treadmill grade every minute [36]. The treadmill speed is chosen based on patient age. This pro- tocol is well suited for unfit, obese, or chronically ill chil- dren where most of the test is spent walking. This is a disadvantage for the highly fit subjects. Modified ‘run- ning’ Balke protocols can be developed starting at a higher initial grade, and choosing a faster treadmill velocity. Metabolic measurements are also easier with the Balke protocol and its modifications [37]. Indications for Pediatric Exercise Tests Indications for exercise testing in pediatric patients are Fig. 7. HR and O2 recovery times for control and Fontan subjects. In varied and often much different than in the adult popula- controls, recovery times were longer following the higher work rate tion. Exercise testing may be needed for diagnostic pur- protocols (* p ! 0.05). In Fontan subjects, recovery times were pro- poses or to institute a therapeutic exercise program. Inter- longed compared with both the same absolute (2 W/kg) and relative esting examples are included in table 1. (3.5 W/kg) protocols in control subjects (** p ! 0.001). Data reprinted with permission from Troutman et al. [32]. A useful approach toward the indications for exercise testing in children is to group them based on the special interests of the physicians or other health care profession- 288 Fahey/Nemet/Cooper

Table1. Examples of current clinical uses of exercise in children and chronic lung disease associated with prematurity [e.g. adolescents bronchopulmonary dysplasia (BPD) and gastroesopha- geal reflux (GER)]. In addition, upper airway obstruction Diagnostic is probably more common in the pediatric age range, and Elucidation of bronchial reactivity may be classified as congenital, acquired, or functional. Growth hormone deficiency Symptoms associated with exercise may occur with all of Preparticipation sports physical in children these diseases. Establishing reduced physical activity as an etiology for obesity Evaluating fitness and the efficacy of therapy in patients with a Asthma and Exercise Currently, the most common use of diagnostic exercise variety of congenital heart diseases testing is in the evaluation of exercise-induced asthma Evaluating long-term outcomes of surgical repair of single-ventricle (EIA). EIA, a common feature of asthma especially in chil- dren, is characterized by a short and sometimes severe congenital heart lesions asthmatic attack following exercise. As EIA occurs in Determination of optimal hematocrit in patients with congenital about 60–80% of asthmatic children it can be used as one of the diagnostic tests to establish the disease; moreover, anemia an exercise challenge is a more specific indicator of asthma Quantifying the functional disability caused by chronic diseases of than either histamine or methacholine challenges [38]. During exercise, lung function changes little or may childhood and developing an exercise prescription even improve. Toward the end of exercise or few minutes Predicting outcomes in cystic fibrosis after, lung function begins to fall. The maximum fall in Establishing the diagnosis of the long QT syndrome lung function occurs 5–10 min after the end of exercise after which recovery in lung function will take place in Therapeutic and rehabilitation about 30–45 min. In a small proportion of asthmatic sub- Training respiratory muscles in patients with chronic lung disease jects a late phase reaction may develop [39–41]. An adjunct to chest physiotherapy in patients with cystic fibrosis The upper limit of post-exercise fall in FEV1 (mean + Stabilization of glycemia in insulin-dependent diabetes 2 SD) in normal children was found to be 6–8% [42]. In An adjunct to diet in the treatment of childhood obesity asthmatic children the severity of EIA may be influenced Nonpharmacologic treatment of juvenile hypertension by the severity of asthma [43] and by pre-exposure to Increasing school and social participation in children with allergens [44]. Moreover, the severity, duration and type of exercise may influence the severity of EIA. Running, as osteogenesis imperfecta compared to swimming under the same inspired air con- Amelioration of exercise induced asthma ditions and work intensity, will result in much more EIA Pediatric cardiac rehabilitation [45]. It is also important to note that the recovery from Improving functional performance in children with cerebral palsy EIA differs in younger compared with older children. For example, Hofstra et al. [46] recently demonstrated that 7- als who are making the referrals. For example, the ques- to 10-year-olds with EIA improved FEV1 by a mean of tions to be answered by the test may be much different if 1.60%/min following the challenge, but improvement in the child is referred by a pediatric pulmonologist as 11- to 12-year-olds was significantly prolonged (0.54%/ opposed to a pediatric cardiologist. However, it is not min). Practical guidelines for the performance of exercise uncommon to find multiplicity of mechanisms – the child challenge testing for the diagnosis of asthma are reviewed referred by the general pediatrician for ‘feeling faint’ dur- elsewhere [8]. ing exercise may prove to have exercise-induced broncho- spasm. Using this approach, the patient referrals could be Cystic Fibrosis and Exercise broken down into four categories: (1) pediatric pulmo- The clinician attempting to prescribe a program of nary; (2) pediatric cardiology; (3) other pediatric subspe- exercise training for children and adolescents with cystic cialties, and (4) general pediatrics. Each will be dealt with fibrosis (CF) faces a dilemma. Exercise may promote in a separate section. There will obviously be overlap health in CF in part by stimulating growth factors and between the categories; for example, shortness of breath tissue anabolism (enhanced bone mineralization, in- with exercise may be a question asked in all categories, creased muscle hypertrophy, mitochondrial density and and this grouping will only serve as a framework. Pediatric Pulmonary Referrals Pediatric pulmonologists deal with a variety of chronic lung and airway diseases. The most important chronic pediatric lung diseases are asthma, cystic fibrosis, and Clinical Exercise Testing in Children 289

Fig. 8. The relationship between VE-VCO2 during a progressive exer- cise test in a 9-year-old girl with CF. Since VE is driven by VCO2, these data tend to have a high signal-to-noise ratio. The slope of the VE-VCO2 relationship is easily calculated using standard linear regression techniques. Reprinted with permission from Moser et al. [53]. capillarization, and increased insulin sensitivity). How- Fig. 9. Dynamic variables of progressive exercise tests in CF subjects ever, even in healthy children, it is now known that the (closed circles) and controls (open circles) as a function of body very same process of exercise, if sufficiently intense, can weight. ¢VO2/¢WR data are shown in the upper panel, and ¢VE/ stimulate inflammatory cytokines and lead to a catabolic ¢VCO2 in the lower panel. Both variables were significantly abnor- state [7, 47–49]. Finding the optimal level of physical mal in CF subjects. Reprinted with permission from Moser et al. activity in children and adolescents with CF is difficult [53]. because the underlying disease is associated with in- creased basal energy expenditure [50–53], hypoxemia, progresses and lung function deteriorates, exercise toler- malnutrition, and inflammation, all of which promote tis- ance diminishes. sue catabolism even at rest. Children and adolescents with cystic fibrosis are Cystic fibrosis is a multisystem disease, with the lung, known to have reduced exercise tolerance, but despite this gut, and pancreas being most severely affected. There is there appear to be therapeutic benefits of exercise in these an increased viscosity of secretions that block the airways. subjects. Indeed, fitter CF patients may have an improved Additionally, inflammatory processes in the lung (second- outcome [54]. The precise mechanism of the exercise ary to chronic bacterial and viral infections) lead to fur- impairment remains largely unknown, but clearly, nutri- ther viscosity increases and worse obstruction. Pulmo- tional and cardiorespiratory factors play a role [55]. nary disease progresses through bronchitis/bronchiolitis, bronchiectasis, emphysema, and restrictive changes, and Although VO2peak or max is usually reduced in CF accounts for 95% of the deaths of patients with cystic patients, more recent studies have begun to identify those fibrosis. It is a progressive disease that eventually results factors which may contribute to the exercise limitation in in hypoxemia and pulmonary hypertension. Individuals these children. Using progressive exercise protocols on with cystic fibrosis experience an average 2% per year the cycle ergometer and breath-to-breath measurements decline in their FEV1. Many patients with mild lung dys- of gas exchange, ¢VO2/¢HR was found to be low in CF function have normal exercise tolerance. As the disease subjects and ¢VE/¢VCO2 high (fig. 8, 9). The slope of the 290 Fahey/Nemet/Cooper