Pediatric Pulmonology 13:lOl-107 (1992)
Cardiopulmonary Response to Exercise in Anorexia Nervosa Larry Lands, M D , ~Alan Pavilanis, M D , ~Thomas D. Charge, BSC (KI~),' and Allan L. Coates, MD' Summary. Malnutritionis associated with a number of systemic diseases that are often accompanied by severe exercise limitation. Anorexia nervosa (AN) is a disease characterized by malnutrition due to psychological factors rather than systemic disease. Diminished exercise capacity in AN has been attributed to a loss of muscle mass, dysfunction of remaining muscle, and impaired cardiovascular responses. In order to evaluate the role of malnutrition in the cardiopulmonary response to exercise, nine adolescent girls with AN were evaluated during progressive and steady-state exercise testing using a cycle ergometer. Nutritional status was assessed by body mass percentile (BMP) and percent ideal weight (PIWT). Cardiac output was measured by the indirect (C02 rebreathing) Fick method. Maximum work capacity (Wmax) was expressed as a percent of predicted for sex and height, and cardiac output as a percent of predictedfor oxygen consumption. To ensure that the laboratoryvalues were comparable to the predicted values, a control group consisting of ten adolescents was studied concurrently. Wmax was below the 95% confidence interval in six of nine of the AN group (mean 2 SD; 70 -c 22% predicted), whereas two of ten controls were below and one above this interval (112 2 37%). Wmax correlated with nutritional status (BMP: r = 0.75; P < 0.001 ; P I W : r = 0.8.P < 0.001). Ventilatory responses for COPproductionat steady state and for Wmax were appropriate in both groups. Cardiac output was appropriate in both the controls (103 2 12%) and the AN group (104 14%). This was accomplished in the AN group with a relatively low stroke volume (86f 19Oh) and high heart rate (107 2 13%), both expressed as percent predicted for sex and work. With appropriate cardiac and ventilatory responses, exercise capacity in AN appears to be limited primarily by diminished muscle mass and/or dysfunction of the existing muscle. Pediatr Pulmonol. 1992;13:lOl-107. Q 1992 Wiley-hs,Inc.
*
Key words: Malnutrition; BMP; PIW; exercising cardiac output; maximum work capacity; mspoflse to COz production.
INTRODUCTION
Malnutrition in children of industrial nations is most frequently seen in conjunction with other systemic disease; therefore attempts to elucidate the specific role of malnutrition on the cardiopulmonary responses to exercise are frequently confused by the underlying disease. l S z Anorexia nervosa (AN) is a disease characterized primarily by malnutrition in the absence of other underlying disease. Subjects with AN often demonstrate a diminished exercise perf~rmance.~'Most of the decrease in exercise capacity can be attributed to a loss of muscle mass,' but dysfunction of the remaining mass6 may also play a role. This dysfunction may be related to electrolyte disturbances and endocrinological changes that can occur with AN. Cardiovascular responses to exercise may be adversely affected by malnutrition, thus contributing to the diminished exercise capacity seen in systemic diseases. Starvation may result in a loss of myocardial mass and a fall in stroke volume and cardiac output. As body size is concurrently diminished with starvation, resting size-related 0 1992 Wiley-Liss, Inc.
cardiac indices are often maintained.' Cardiac contractility appears n ~ r m a l , ~but . ~ cardiac .~ volume as assessed by echocardiography is reduced, although the reduction is not disproportional to that of the body mass. Abnormal cardiovascular responses to exercise have been noted in AN, including diminished heart rate (HR) and blood pressure responses. Although one study investigated From the Montreal Children's Hospital Research Institute-McGill University, and Hbpital Maisonneuve-Rosemont, Universitk de Montreal,* Montreal, Canada.
'
Received June 12, 1991; (revision) accepted for publication February 13, 1992.
This is publication No. 92013 of the McGill University-Montreal Children's Hospital Research Institute. Funded in part by The Canadian Cystic Fibrosis Foundation. Address correspondence and reprint requests to Dr. A. L. Coates, Respiratory Medicine, Montreal Children's Hospital, 2300 Tupper St., Montreal, Quebec, Canada H3H 1P3.
102
Lands et al.
’”
ejection fraction response during supine work, this may differ from that occurring during upright exercise, which is clinically more relevant. I To date, the cardiopulmonary responses to upright exercise have not been measured in AN. We have undertaken to evaluate the effect of malnutrition without any other systemic disease on cardiac and pulmonary function during exercise. MATERIALS AND METHODS
Nine adolescent females with AN were recruited from the Adolescent Medical Service of the Montreal Children’s Hospital. For inclusion in the study, AN was defined’* as having a 25% loss of body weight, amenorrhea, a consistent psychological disturbance, and laboratory evidence ruling out other disorders that cause weight loss. The patients were studied early in their outpatient treatment phase when all electrolytic and other biochemical imbalances had been corrected but before significant improvement in weight. A normal adolescent group consisting of six boys and four girls was concurrently studied in order to ensure that normal values from the laboratory were comparable to those in the literature used as predicted values. All subjects were questioned and all stated that they were not involved in an active training program. Written informed consent was obtained from the subjects and/or their parents, depending on age. The study had the approval of the institutional ethics committee. Upon entry to the study, subjects were measured for their standing height without shoes, and weighed wearing light clothing. Many measures of nutritional status are available, although none are entirely satisfactory. Percent ideal weight (PIWT)’ is calculated as the ratio of the actual weight of the child to the “ideal” weight of a child the same age, height, and sex. The ideal weight assumes that the weight of the child should score the same percentile ranking as that for height from the National Centre for Health Statistics (NCHS) Growth Charts. For malnourished children who are also stunted, with heights below the published curves, the PIWT cannot be calculated. Body mass percentile (BMP)I3 is calculated as the body mass index (weightlheight,) divided by the expected body mass index for a normal child of the same age and sex growing along the 50th percentile. It is a measure of “thinness” and can be calculated in all children. In the present study, both BMP and PIWT were used. Values less than 90% were considered abnormal. Forced vital capacity (FVC), forced expired volume in 1 second (FEV,), and maximum voluntary ventilation using the 15 second sprint (MVV) were measured using a dry rolling seal spirometer (Gould 5000 IV) and expressed as percent predicted for height, sex, and age.I4 Static lung volumes, slow vital capacity (VC), functional residual capacity (FRC), total lung capacity (TLC), and
residual volume (RV) were measured in a volume displacement body plethysmograph (J.H. Emerson co.) and expressed as a percent In the presence of a physician, the subjects performed a progressive exercise test until exhaustion on a cycle ergometer as previously described” by having the work load increased by a fixed increment each minute (Jones Stage I). This increment was chosen according to the sex and height of the subject” so that exhaustion occurred within 5 to 10 minutes from the start of the test. At exhaustion, the subject was no longer able to continue pedalling despite encouragement, and was essentially unable to stand unsupported until after a few minutes of rest. Only a test in which it was apparent to the investigator that a true maximum effort had been made was accepted, and this was taken as the subject’s maximum work capacity (Wmax). During the test, HR was continually monitored via EKG (Nellcor N200). Inspired ventilation was measured by a dry gas meter (Parkinson-Cowan). Exhalation was via a uni-directional valve to a variable volume mixing chamber adjusted to the subject’s VC. Mixed expired gas was analyzed for fractional concentrations of oxygen (Applied Electrochemistry S-3A) and carbon dioxide (Beckman LB-2). End-tidal CO, was measured as an estimate of arterial Pcol. HR (via EKG), inspired ventilation, and gas concentrations were recorded on a continuous %channel recorder (Hewlett-Packard 7758A). Oxygen consumption (Vo,) and CO, production (Vco,) were calculated using the nitrogen balance technique -during the last 10 seconds of each workload. The respiratory exchange ratio (RER) was calculated as Vc.,/Vo,. Oxygen saturation by ear oximetry (Nellcor N200) was measured continuously to ensure it was always above 90%. Details of the methodology have been described previously.2JR. ‘ 9 After a rest period of at least 45 minutes, a 5-minute steady-state exercise test at a workload of 50%‘ of the subject’s previously determined Wmax was performed. Steady state was considered achieved when there were no changes in mixed expired 0, and CO, for at least 1 minute. Cardiac output (Q)was assessed at the end of this time using the indirect Fick This CO, rebreathing technique consists of estimating mixed venous partial pressure of CO, (PVm2)by equilibration. The subject, when at steady state, is switched to a circuit at end-expiration so that he or she breathes from a bag containing a mixture of 1 6 1 5 % CO, in 0,. The choice of the CO, concentration is predicated on achieving, after dilution of the physiological deadspace, a Pco, in the alveolar-rebreathing bag system that is approimately ‘y*2” Deep breathing is encouraged to equal to the Pvcc,2. ensure quick, adequate mixing between the lungs and the bag. Once this is achieved, small further transfers of CO, from venous blood to the lung-bag system or vice versa
Cardiopulmonary Response in Anorexia Newosa
103
TABLE 1-Anthropometrlc and Pulmonary Function Parameters of Control Subjects and Anorexia Nenoaa IAN) Patients. Control Sex (M:F) Age (yrs) Height (cm) Weight (kg) BMP PIWT VC (% pred.) FEVi (L) FEVi (76 pred.) FEVIIFVC MVV (% pred.) MVV (in FEVl units)
Outside CIb
6:4 14.8 f 2.3 163 f 14 59 i 13 1 1 1 f 13 107 f 18 99 i 9 3.58 f 0.94 112 f 17 87 5 123 f 30 36.1 k 4.4
+
0110 1110 0110 0110 1110 0110
AN
Outside CIb
0:9 14.7 f 1.9 164 f 7 44 f 5 81 f 9 73 f 12 90 f 8 2.74 f 0.58 93 f 21 8 2 f 18 96 f 16 35.5 f 13.6
819 819 119 119 219 219
aAN, anorexia nervosa; BMP, body mass percentile; PIWT, percent ideal weight; VC, vital capacity; FEVI, forced expiratory volume in 1 second; FVC, forced vital capacity; MVV, maximal voluntary ventilation; % pred., percent of predicted. Values in mean f standard deviation. boutside the 95% confidence interval (>2 SD from mean), nltotal.
will allow for equilibration of the two. This occurs between 6 and 10 seconds, before recirculation can occur and raise the Pvcc,.As Vco2 and arterial CO, have been calculated in the 30 seconds preceding equilibration, Q can be estimated by dividing Vco2by the venous-arterial CO, content difference, content being derived from 21 Pco2. This method has been validated in patients both with and without lung disease.” Downstream correction for mixed venous CO, tension has not been used so as to allow comparison with previous pediatric data. Stroke volume (SV) was calculated as Q/HR averaged over the 30 seconds preceding the rebreathing. Values for Q and SV were expressed as percent of predicted, based on VO2 and on sex and height, respectively. Cardiac output was expressed as percent of predicted for Vo2rather than as an index related to body surface area, because the subjects were evaluated at different absolute work levels and it was more appropriate to compare them on the basis of output for work done. This, as well as differences in body size, required that all results be expressed as percent predicted, based on sex and height to allow for comparison. By using percent of predicted values, it was possible to validate the methodology with the control subjects, thus comparing our study group with previously published standards that take into account sexual differences. Results were expressed as mean standard deviation and as the proportion of each group outside the 95% confidence limits of published normal values. Correlations were calculated using Pearson’s correlation coefficient.
*
RESULTS
Table 1 outlines the characteristics of the subjects as well as the children used as laboratory controls. Lung volumes were minimally lower in the AN group. This
resulted in a lower FVC and FEV, and a normal FEV,/ FVC ratio. RV/TLC ratio was slightly elevated in the AN group but TLC was normal. The MVV was also lower in the AN group when expressed as a percent predicted but not when expressed as the number of FEV I ’s breathed in 15 seconds. The AN group was malnourished, with a low BMP and PIWT (81 9%, 73 12%, mean SD). Resting hemoglobin values were available on seven of the AN subjects and ranged from 11.4 to 14.3 g/lOO mL. Table 2 outlines the responses to exercise. All subjects made maximum efforts during the progressive exercise test as witnessed by progressive elevations of heart rate and ventilation and relative hyperventilation towards completion. The degree of hyperventilation was evident by the RER of 1.46 0.20 for the AN group and 1.26 2 0.18 for the control group at Wmax. The subjects were exhausted at the time of cessation of exercise and unable to stand without support immediately following the test. Using the prediction equationi8for work in children, the subjects with AN performed less work than expected (mean, 70 22% predicted). Using 80% of Wmax predicted as the lower limit of normal, 6/9 patients were abnormal (Fig. 1). Regarding nutritional status, maximum work capacity achieved correlated with percent ideal weight for the combined groups (Fig. 2) (PIWT: r = 0.8, P < 0.001). It correlated with body mass percentile in a similar way (BMP: r = 0.75, P < 0.001). It should be noted that 8/10 of the control subjects achieved greater than 80% Wmax with a mean of 112 & 37% predicted. The two who were low both exceeded 70% predicted. One subject was above the upper limit of the confidence interval. Ventilation at Wmax was assessed by comparing the result with the normal plot of ventilation versus Wmax. Normally, the maximum workload achieved, which determines the metabolic demand, is the determining factor
*
*
*
*
*
Landsetal.
104
TABLE 2-Exercise Testing Parameters of Control Subjects and Anorexia Nervosa Patients’ Control
Outside CIb
186 f 52 3.2 f 0.6 112 f 37 67 f 9 182 f 12 157 10 98 f 8 93 7 103 f 12 1.49 f 0.4 105 f 20
Wmax (watts) Wmax (wattslkg) Wmax (96 pred.) VEmax (% M W ) HR at Wmax (bpm) HR at SS (bpm) HR at Wmax (% pred.) HR at SS (% pred.) Q at SS (% pred.) Vo2 at SS (Llmin) SV at ss (8 Dred.)
98 f 16 2.2 f 0.4 70 f 22 61 -120 169 f 16 144 f 13 105 f 8 107 f 13 104-1 14 0.88 -1 0.2 8 6 f 19
3/10
* *
Outside Clb
AN
0110 0110 0110 2/10 2/10
719
019 219 019 1/9 619
aWmax, maximum work capacity; Vmax,.minute ventilation at Wmax; MVV, maximal voluntary ventilation; HR, heart rate; SS,steady state; Q,cardiac output;Vo,, oxygen uptake;SV, stroke volume; % pred., percent of predicted. Values in mean 5~ standard deviation. boutside the 95% confidence interval (>2 SD from mean), n/total. 225 200
120
-
+ 100
-
7
1
n
0
TI Q
175
.TI 0
150
2 80 -
K
E
3
W
100
X 0
50 J, 140
I
I
I
150
160
170
Ht
1
1
180
190
40
(cm)
50
Fig. 1. Height (Ht) versus maxlmum work capacity (Wmax). 0 , anorexia nervosa patlents; 0 , control females. Regression line from data of Godfreyla for females with the 95% confidence intervals (k 34W).
1
I
I
I
75
100
125
150
PIWT Fig. 2. Percent Ideal weight (PIWT) versus maxlmum work capacity (Wmax). 0, anorexia nervosa patlents; 0, control females; 0 ,control males. Correlation (Pearson’s) coefficient: r = 0.8, P < 0.001.
of maximum ventilation. Ventilation was appropriate for work done in all subjects. Maximum ventilation as a percentage of the 15 second MVV was similar in both groups and, as is expected in normal individuals, was well below the MVV (67 f 9% for controls, 61 20% for AN). At steady state, oo2was appropriate for work done. Q was maintained in the AN group (104 14% predicted) (Fig. 3), but with a relatively high HR (107 & 13% predicted) and low SV (86 ? 19% predicted). Stroke volume correlated weakly with both indices of nutritional status for the combined groups: PIWT, r = 0.51; P C 0.03 (Fig. 4) and BMP, r = 0.52, P < 0.02. The relationship between ventilation and Vco2at steady state was in keeping with published normal values” (Fig. 5).
*
*
DISCUSSION
The present study has shown that impaired exercise ability in AN is related to poor nutritional status. Ventilatory responses are appropriate for Vco, and Wmax. There is preservation of Q but this is achieved with a lower SV and higher heart rate. The normal Q response coupled with hemoglobin concentrations greater than 10 g/100 mL would suggest that oxygen delivery is not a limiting factor for exercise. In conjunction with the normal ventilatory responses, it appears that exercise capacity in AN is limited by loss of muscle mass or dysfunction of existing muscle, as has been suggested by This inability to perform work in AN is consistent with previous studies.4*5.8-10
Cardiopulmonary Response 25
6o
.C
E 20
1 '.
0
-
in Anorexia Newosa
1
105
0
U
I-
3
k 3
15-
0 0
1
g
10-
4 V
0
5 !
I
0.5
I
0.5
I
1 .o 1.5 2.0 0X Y G,EN CON S U M PT I0N (I / rnin)
0
n
u
125
-
P)
c 0
.U
!! 100 -
L
K
> v1
0
__,6___
0
v
75
50
-
0
I
I
I
1.5
2.0
2.5
Vco, (L/rnin)
Output (dl at Fig. 3. OVgen consumption * O J 5096 of Wmax. Symbols as In Figure 2. Regression line from data of Godfrey." 150
1 .o
t
Fig. 5. Ventilation (VE) versus carbon dioxide production(V-J at Steady state. Symbols as In Figure 2. Regression from data Jones.17
Previous work in malnutrition has generally studied l o demonstrated cardiac function at r e ~ t . ~ , ~One , ~ study '' that the ejection fraction during supine work using radionuclide cineangiography did increase appropriately, but the workload at which measurements were made was not delineated and because of differences between upright and supine work, cardiac responses may not be comparable. I ' The present study using a non-invasive technique demonstrated that at Yi-Wmax, Q was maintained but with a smaller SV and increased HR. The question arises as to whether this compensation persists at Wmax. Previous work in normal subjects has shown that an individual attains more than 90% of maximum SV with
derived was maximal in the control subjects. The low HR previously observed in subjects with AN at rest is in contrast with the higher than expected one for work performed in the present study. The finding of a lower HR at maximal exercise compared to controls reported by NuMalnutrition can result in functional impairment of del et aL4 pertains only to absolute values. If the HR for muscle. Caloric deprivation tends to lead to selective controls and patients is expressed as a percent expected atrophy of type I1 fibers, decreasing the force-developing for work, the patients with AN have a relative tachycarpotential of a muscle and leading to premature fatigue.6 dia. Interpreting the relative resting bradycardia in AN Cellular integrity and intracellular ion contents can also patients with the echocardiographic data suggesting a be adversely affected.6 Subjects with AN often have de- normal for body mass at rest would clearly imply that pletion of potassium stores,6 and thus again, may be the resting SV of these patients is a larger percentage of predisposed to fatigue. Animal studies have shown that maximal than it is in normal subjects. Cardiac output fatigue is related to potassium efflux from exercising during exercise increases first by a joint increase in both muscle fibers.22There is also the possibility of a detrain- HR and SV until SV is maximal, whereupon further ing effect in the AN group. This may augment the cate- increases are due to increases in the HR alone. This cholamine response to exercise and hence increase the would suggest that patients with AN would be expected to achieve a maximal SV earlier in the course of exercise heart rate response. Fig. 4. Percent ideal weight (PIWT) versus stroke volume (SV). Symbols as in Figure 2. Correlation: r = 0.51, P < 0.03.
Lands et al.
106
than normals. The finding of an HR in the patients with AN well above that considered sufficient for a maximal SV in normals clearly would imply that a maximal SV was achieved in these patients. Evidence to the contrary would be a Q less than expected for oxygen consumption in patients with AN. This was not the case in the present study. In an effort to see if “cardiac insufficiency” (defined for the purpose of this study as an inappropriately low Q for work accomplished or was a factor, the SV was multiplied by the HR at Wmax to give an estimate of maximum Q. Using this estimate, Q in the AN group compared to controls was preserved even at Wmax. This would imply either that cardiac insufficiency at maximum work is not the main limiting factor to exercise or that the onset of cardiac insufficiency results in immediate cessation of work. The methods used in this study were not sufficiently sensitive to answer this question. In the present study SV correlated with nutritional status, albeit weakly. Part of the reason for the lack of a strong correlation is that SV, unlike Wmax, is calculated and not measured directly. Hence, the scatter in the relationship may be due to small inherent inaccuracies in any or all component parts of the determination of SV. It is unclear, however, to what extent this lower SV might be due to the elevated response in HR, a loss of cardiac muscle mass, decreased filling pressures, or decreased myocardial contractility. At the same workload, HR responses are higher when the work is performed by a small The subjects muscle mass as compared to a large with AN had low weight for height so that at any given they were performing with a reduced muscle mass, compared to the normal population. This may be the reason for the increased HR response in these patients. The finding of slight reductions in VC and FEV, ,and a mildly increased RV/TLC in the AN group is due to respiratory muscle weakness associated with malnutrition. This has been reported in a recent p~blication.’~ In conclusion, our subjects with AN demonstrated a reduced exercise capacity. The cardiac output response was adequate although this was achieved with a relatively elevated heart rate. Ventilatory responses appeared appropriate. Hemoglobin concentrations were not in a range to cause diminished oxygen delivery. This would suggest a peripheral limitation to exercise capacity, in large part due to the loss of muscle mass. The impact of peripheral muscle function on cardiopulmonary responses to exercise in patients with malnutrition associated with systemic or other organ system disease’.’ needs further investigation.
vo2)
vo2,
REFERENCES 1. Marcotte JE, Canny GJ, Grisdale R, Desmond K, Corey M,
Zinman R, Levison H, Coates AL. Effect of nutritional status on
exercise performance in advanced cystic fibrosis. Chest. 1986; 90:375-379. 2. Hortop J, Desmond KJ, Coates AL. The mechanical effects of
expiratory airflow limitation on cardiac performance in cystic fibrosis. Am Rev Respir Dis. 1988; 137:132-137. 3. Moodie DS, Salcedo E. Cardiac function in adolescents and young adults with anorexia nervosa. J Adolesc Health Care. 1983; 49-14. 4. Nude1 DB, Gootman N, Nussbaum MP, Shenker IR. Altered
exercise performance and abnormal sympathetic response to exercise in patients with anorexia nervosa. J Pediatr. 1984; 105:34-37. 5 . Davies CTM, Barnes C, Godfrey S. Body composition and maximal exercise performance in children. Hum Biol. 1972; 44:195214. 6. Russel DM, Walker PM, Lester LA, et al. Metabolic and struc-
7. 8.
9.
10.
11.
12.
13.
tural changes in skeletal muscle during hypocaloric dieting. Am J Clin Nutr. 1984; 39503-513. Webb JG, Kiess MC, Chan-Yan CC. Malnutrition and the heart. CMAJ 1986; 135:753-758. Heymsfield SB, Bethel RA, Ansley JD, Gibbs DM, Felner JM, Nutter DO. Cardiac abnormalities in cachectic patients before and during nutritional repletion. Am Heart J. 1978; 9 5 5 8 4 5 9 4 . St. John Sutton MG, Plappert T, Crosby L, Douglas P, Mullen J , Recchek N. Effects of reduced left ventricular mass on chamber architecture, load and function: a study of anorexia nervosa. Circulation. 1985; 72:991-1000. Goltdiener JS, Gross HA, Henry WL, Borer JS, Ebert MH. Effects of self-induced starvation on cardiac size and function in anorexia nervosa. Circulation. 1978; 58:425-433. Marcotte JE, Perrault H, Drblik SP. Caughlan M, Lamarre A. Hernodynamic response to upright and supine exercise in mild to severe cystic fibrosis patients. Am Rev Respir Dis. 1989; 139:A7 I. Work Group to Revise DSM-111. Diagnostic and Statistical Manual of Mental Disorders, 3rd ed, revised. Washington, DC: American Psychiatric Association, 1987. Coates AL, Boyce P, Muller D, Mearns M, Godfrey S. The role of nutritional status, airway obstruction, hypoxia, and abnormalities in serum lipid composition in limiting exercise tolerance in children with cystic fibrosis. Acta Paediatr Scand. 1980;
69:353-358. 14. Knudson RJ, Slatin RC, Lebowitz MD, Burrows B. The maximal expiratory flow-volume curve. Am Rev Respir Dis. 1976; I13:587400. 15. Desmond KJ, Demizio DL, Allen PD, Beaudry PH, Coates AL.
An alternate method for the determination of functional residual capacity in a plethysmograph. Am Rev Respir Dis. 1988; 1371273-276. 16. Polgar G, Promadhat V.Pulmonary Function Testing in Children. Philadelphia: WB Saunders, 1971. 17. Jones NL. Clinical Exercise Testing. Philadelphia: WB Saunders, 1988. 18. Godfrey S . Exercise Testing in Children. Philadelphia: WB Saunders, 1974. 19. Coates AL, Desmond K, Asher MI, Hortop J , Beaudry PH. The
effect of digoxin on exercise capacity and exercising cardiac function in cystic fibrosis. Chest. 1982; 82543-547. 20. Mahler DA, Matthay RA, Snyder PE, Neff RK, Loke J . Determination of cardiac output at rest and during exercise by carbon dioxide rebreathing method in obstructive airway disease. Am Rev Respir Dis. 1985; 131:73-78. 21. McHardy GJR. Relationship between the difference in pressure and content of carbon dioxide in arterial and venous blood. Clin Sci. 1967; 32:299-309. 22. Lindinger MI, Heigenhauser GJF. Ion fluxes during tetanic stimu-
Cardiopulmonary Response in Anorexia Nervosa lation in isolated perfused rat hindlimbs. Am J Physiol. 1988; 254:Rl17-R126. 23. Astrand P-0, Cuddy TE, Saltin B, Steinberg J. Cardiac output during sub-maximal and maximal work. J Appl Physiol. 1964; I 9:268-274. 24. Lewis SF, Taylor WF, Graham RM, Pettinger WE, Schutte JG, Blomquist CG. Cardiovascular responses to exercise as functions
107
of absolute and relative workload. J Appl Physiol. 1983; 54: 13141323. 25. Lands L, Desmond KJ,Demizio D, Pavilanis A, Coates AL. The effects of nutritional status and hyperinflation on respiratory muscle strength in children and young adults. Am Rev Respir Dis. 1990; 141:1506-1 509.
Cardiopulmonary Response to Exercise in Anorexia Nervosa Larry Lands, M D , ~Alan Pavilanis, M D , ~Thomas D. Charge, BSC (KI~),' and Allan L. Coates, MD' Summary. Malnutritionis associated with a number of systemic diseases that are often accompanied by severe exercise limitation. Anorexia nervosa (AN) is a disease characterized by malnutrition due to psychological factors rather than systemic disease. Diminished exercise capacity in AN has been attributed to a loss of muscle mass, dysfunction of remaining muscle, and impaired cardiovascular responses. In order to evaluate the role of malnutrition in the cardiopulmonary response to exercise, nine adolescent girls with AN were evaluated during progressive and steady-state exercise testing using a cycle ergometer. Nutritional status was assessed by body mass percentile (BMP) and percent ideal weight (PIWT). Cardiac output was measured by the indirect (C02 rebreathing) Fick method. Maximum work capacity (Wmax) was expressed as a percent of predicted for sex and height, and cardiac output as a percent of predictedfor oxygen consumption. To ensure that the laboratoryvalues were comparable to the predicted values, a control group consisting of ten adolescents was studied concurrently. Wmax was below the 95% confidence interval in six of nine of the AN group (mean 2 SD; 70 -c 22% predicted), whereas two of ten controls were below and one above this interval (112 2 37%). Wmax correlated with nutritional status (BMP: r = 0.75; P < 0.001 ; P I W : r = 0.8.P < 0.001). Ventilatory responses for COPproductionat steady state and for Wmax were appropriate in both groups. Cardiac output was appropriate in both the controls (103 2 12%) and the AN group (104 14%). This was accomplished in the AN group with a relatively low stroke volume (86f 19Oh) and high heart rate (107 2 13%), both expressed as percent predicted for sex and work. With appropriate cardiac and ventilatory responses, exercise capacity in AN appears to be limited primarily by diminished muscle mass and/or dysfunction of the existing muscle. Pediatr Pulmonol. 1992;13:lOl-107. Q 1992 Wiley-hs,Inc.
*
Key words: Malnutrition; BMP; PIW; exercising cardiac output; maximum work capacity; mspoflse to COz production.
INTRODUCTION
Malnutrition in children of industrial nations is most frequently seen in conjunction with other systemic disease; therefore attempts to elucidate the specific role of malnutrition on the cardiopulmonary responses to exercise are frequently confused by the underlying disease. l S z Anorexia nervosa (AN) is a disease characterized primarily by malnutrition in the absence of other underlying disease. Subjects with AN often demonstrate a diminished exercise perf~rmance.~'Most of the decrease in exercise capacity can be attributed to a loss of muscle mass,' but dysfunction of the remaining mass6 may also play a role. This dysfunction may be related to electrolyte disturbances and endocrinological changes that can occur with AN. Cardiovascular responses to exercise may be adversely affected by malnutrition, thus contributing to the diminished exercise capacity seen in systemic diseases. Starvation may result in a loss of myocardial mass and a fall in stroke volume and cardiac output. As body size is concurrently diminished with starvation, resting size-related 0 1992 Wiley-Liss, Inc.
cardiac indices are often maintained.' Cardiac contractility appears n ~ r m a l , ~but . ~ cardiac .~ volume as assessed by echocardiography is reduced, although the reduction is not disproportional to that of the body mass. Abnormal cardiovascular responses to exercise have been noted in AN, including diminished heart rate (HR) and blood pressure responses. Although one study investigated From the Montreal Children's Hospital Research Institute-McGill University, and Hbpital Maisonneuve-Rosemont, Universitk de Montreal,* Montreal, Canada.
'
Received June 12, 1991; (revision) accepted for publication February 13, 1992.
This is publication No. 92013 of the McGill University-Montreal Children's Hospital Research Institute. Funded in part by The Canadian Cystic Fibrosis Foundation. Address correspondence and reprint requests to Dr. A. L. Coates, Respiratory Medicine, Montreal Children's Hospital, 2300 Tupper St., Montreal, Quebec, Canada H3H 1P3.
102
Lands et al.
’”
ejection fraction response during supine work, this may differ from that occurring during upright exercise, which is clinically more relevant. I To date, the cardiopulmonary responses to upright exercise have not been measured in AN. We have undertaken to evaluate the effect of malnutrition without any other systemic disease on cardiac and pulmonary function during exercise. MATERIALS AND METHODS
Nine adolescent females with AN were recruited from the Adolescent Medical Service of the Montreal Children’s Hospital. For inclusion in the study, AN was defined’* as having a 25% loss of body weight, amenorrhea, a consistent psychological disturbance, and laboratory evidence ruling out other disorders that cause weight loss. The patients were studied early in their outpatient treatment phase when all electrolytic and other biochemical imbalances had been corrected but before significant improvement in weight. A normal adolescent group consisting of six boys and four girls was concurrently studied in order to ensure that normal values from the laboratory were comparable to those in the literature used as predicted values. All subjects were questioned and all stated that they were not involved in an active training program. Written informed consent was obtained from the subjects and/or their parents, depending on age. The study had the approval of the institutional ethics committee. Upon entry to the study, subjects were measured for their standing height without shoes, and weighed wearing light clothing. Many measures of nutritional status are available, although none are entirely satisfactory. Percent ideal weight (PIWT)’ is calculated as the ratio of the actual weight of the child to the “ideal” weight of a child the same age, height, and sex. The ideal weight assumes that the weight of the child should score the same percentile ranking as that for height from the National Centre for Health Statistics (NCHS) Growth Charts. For malnourished children who are also stunted, with heights below the published curves, the PIWT cannot be calculated. Body mass percentile (BMP)I3 is calculated as the body mass index (weightlheight,) divided by the expected body mass index for a normal child of the same age and sex growing along the 50th percentile. It is a measure of “thinness” and can be calculated in all children. In the present study, both BMP and PIWT were used. Values less than 90% were considered abnormal. Forced vital capacity (FVC), forced expired volume in 1 second (FEV,), and maximum voluntary ventilation using the 15 second sprint (MVV) were measured using a dry rolling seal spirometer (Gould 5000 IV) and expressed as percent predicted for height, sex, and age.I4 Static lung volumes, slow vital capacity (VC), functional residual capacity (FRC), total lung capacity (TLC), and
residual volume (RV) were measured in a volume displacement body plethysmograph (J.H. Emerson co.) and expressed as a percent In the presence of a physician, the subjects performed a progressive exercise test until exhaustion on a cycle ergometer as previously described” by having the work load increased by a fixed increment each minute (Jones Stage I). This increment was chosen according to the sex and height of the subject” so that exhaustion occurred within 5 to 10 minutes from the start of the test. At exhaustion, the subject was no longer able to continue pedalling despite encouragement, and was essentially unable to stand unsupported until after a few minutes of rest. Only a test in which it was apparent to the investigator that a true maximum effort had been made was accepted, and this was taken as the subject’s maximum work capacity (Wmax). During the test, HR was continually monitored via EKG (Nellcor N200). Inspired ventilation was measured by a dry gas meter (Parkinson-Cowan). Exhalation was via a uni-directional valve to a variable volume mixing chamber adjusted to the subject’s VC. Mixed expired gas was analyzed for fractional concentrations of oxygen (Applied Electrochemistry S-3A) and carbon dioxide (Beckman LB-2). End-tidal CO, was measured as an estimate of arterial Pcol. HR (via EKG), inspired ventilation, and gas concentrations were recorded on a continuous %channel recorder (Hewlett-Packard 7758A). Oxygen consumption (Vo,) and CO, production (Vco,) were calculated using the nitrogen balance technique -during the last 10 seconds of each workload. The respiratory exchange ratio (RER) was calculated as Vc.,/Vo,. Oxygen saturation by ear oximetry (Nellcor N200) was measured continuously to ensure it was always above 90%. Details of the methodology have been described previously.2JR. ‘ 9 After a rest period of at least 45 minutes, a 5-minute steady-state exercise test at a workload of 50%‘ of the subject’s previously determined Wmax was performed. Steady state was considered achieved when there were no changes in mixed expired 0, and CO, for at least 1 minute. Cardiac output (Q)was assessed at the end of this time using the indirect Fick This CO, rebreathing technique consists of estimating mixed venous partial pressure of CO, (PVm2)by equilibration. The subject, when at steady state, is switched to a circuit at end-expiration so that he or she breathes from a bag containing a mixture of 1 6 1 5 % CO, in 0,. The choice of the CO, concentration is predicated on achieving, after dilution of the physiological deadspace, a Pco, in the alveolar-rebreathing bag system that is approimately ‘y*2” Deep breathing is encouraged to equal to the Pvcc,2. ensure quick, adequate mixing between the lungs and the bag. Once this is achieved, small further transfers of CO, from venous blood to the lung-bag system or vice versa
Cardiopulmonary Response in Anorexia Newosa
103
TABLE 1-Anthropometrlc and Pulmonary Function Parameters of Control Subjects and Anorexia Nenoaa IAN) Patients. Control Sex (M:F) Age (yrs) Height (cm) Weight (kg) BMP PIWT VC (% pred.) FEVi (L) FEVi (76 pred.) FEVIIFVC MVV (% pred.) MVV (in FEVl units)
Outside CIb
6:4 14.8 f 2.3 163 f 14 59 i 13 1 1 1 f 13 107 f 18 99 i 9 3.58 f 0.94 112 f 17 87 5 123 f 30 36.1 k 4.4
+
0110 1110 0110 0110 1110 0110
AN
Outside CIb
0:9 14.7 f 1.9 164 f 7 44 f 5 81 f 9 73 f 12 90 f 8 2.74 f 0.58 93 f 21 8 2 f 18 96 f 16 35.5 f 13.6
819 819 119 119 219 219
aAN, anorexia nervosa; BMP, body mass percentile; PIWT, percent ideal weight; VC, vital capacity; FEVI, forced expiratory volume in 1 second; FVC, forced vital capacity; MVV, maximal voluntary ventilation; % pred., percent of predicted. Values in mean f standard deviation. boutside the 95% confidence interval (>2 SD from mean), nltotal.
will allow for equilibration of the two. This occurs between 6 and 10 seconds, before recirculation can occur and raise the Pvcc,.As Vco2 and arterial CO, have been calculated in the 30 seconds preceding equilibration, Q can be estimated by dividing Vco2by the venous-arterial CO, content difference, content being derived from 21 Pco2. This method has been validated in patients both with and without lung disease.” Downstream correction for mixed venous CO, tension has not been used so as to allow comparison with previous pediatric data. Stroke volume (SV) was calculated as Q/HR averaged over the 30 seconds preceding the rebreathing. Values for Q and SV were expressed as percent of predicted, based on VO2 and on sex and height, respectively. Cardiac output was expressed as percent of predicted for Vo2rather than as an index related to body surface area, because the subjects were evaluated at different absolute work levels and it was more appropriate to compare them on the basis of output for work done. This, as well as differences in body size, required that all results be expressed as percent predicted, based on sex and height to allow for comparison. By using percent of predicted values, it was possible to validate the methodology with the control subjects, thus comparing our study group with previously published standards that take into account sexual differences. Results were expressed as mean standard deviation and as the proportion of each group outside the 95% confidence limits of published normal values. Correlations were calculated using Pearson’s correlation coefficient.
*
RESULTS
Table 1 outlines the characteristics of the subjects as well as the children used as laboratory controls. Lung volumes were minimally lower in the AN group. This
resulted in a lower FVC and FEV, and a normal FEV,/ FVC ratio. RV/TLC ratio was slightly elevated in the AN group but TLC was normal. The MVV was also lower in the AN group when expressed as a percent predicted but not when expressed as the number of FEV I ’s breathed in 15 seconds. The AN group was malnourished, with a low BMP and PIWT (81 9%, 73 12%, mean SD). Resting hemoglobin values were available on seven of the AN subjects and ranged from 11.4 to 14.3 g/lOO mL. Table 2 outlines the responses to exercise. All subjects made maximum efforts during the progressive exercise test as witnessed by progressive elevations of heart rate and ventilation and relative hyperventilation towards completion. The degree of hyperventilation was evident by the RER of 1.46 0.20 for the AN group and 1.26 2 0.18 for the control group at Wmax. The subjects were exhausted at the time of cessation of exercise and unable to stand without support immediately following the test. Using the prediction equationi8for work in children, the subjects with AN performed less work than expected (mean, 70 22% predicted). Using 80% of Wmax predicted as the lower limit of normal, 6/9 patients were abnormal (Fig. 1). Regarding nutritional status, maximum work capacity achieved correlated with percent ideal weight for the combined groups (Fig. 2) (PIWT: r = 0.8, P < 0.001). It correlated with body mass percentile in a similar way (BMP: r = 0.75, P < 0.001). It should be noted that 8/10 of the control subjects achieved greater than 80% Wmax with a mean of 112 & 37% predicted. The two who were low both exceeded 70% predicted. One subject was above the upper limit of the confidence interval. Ventilation at Wmax was assessed by comparing the result with the normal plot of ventilation versus Wmax. Normally, the maximum workload achieved, which determines the metabolic demand, is the determining factor
*
*
*
*
*
Landsetal.
104
TABLE 2-Exercise Testing Parameters of Control Subjects and Anorexia Nervosa Patients’ Control
Outside CIb
186 f 52 3.2 f 0.6 112 f 37 67 f 9 182 f 12 157 10 98 f 8 93 7 103 f 12 1.49 f 0.4 105 f 20
Wmax (watts) Wmax (wattslkg) Wmax (96 pred.) VEmax (% M W ) HR at Wmax (bpm) HR at SS (bpm) HR at Wmax (% pred.) HR at SS (% pred.) Q at SS (% pred.) Vo2 at SS (Llmin) SV at ss (8 Dred.)
98 f 16 2.2 f 0.4 70 f 22 61 -120 169 f 16 144 f 13 105 f 8 107 f 13 104-1 14 0.88 -1 0.2 8 6 f 19
3/10
* *
Outside Clb
AN
0110 0110 0110 2/10 2/10
719
019 219 019 1/9 619
aWmax, maximum work capacity; Vmax,.minute ventilation at Wmax; MVV, maximal voluntary ventilation; HR, heart rate; SS,steady state; Q,cardiac output;Vo,, oxygen uptake;SV, stroke volume; % pred., percent of predicted. Values in mean 5~ standard deviation. boutside the 95% confidence interval (>2 SD from mean), n/total. 225 200
120
-
+ 100
-
7
1
n
0
TI Q
175
.TI 0
150
2 80 -
K
E
3
W
100
X 0
50 J, 140
I
I
I
150
160
170
Ht
1
1
180
190
40
(cm)
50
Fig. 1. Height (Ht) versus maxlmum work capacity (Wmax). 0 , anorexia nervosa patlents; 0 , control females. Regression line from data of Godfreyla for females with the 95% confidence intervals (k 34W).
1
I
I
I
75
100
125
150
PIWT Fig. 2. Percent Ideal weight (PIWT) versus maxlmum work capacity (Wmax). 0, anorexia nervosa patlents; 0, control females; 0 ,control males. Correlation (Pearson’s) coefficient: r = 0.8, P < 0.001.
of maximum ventilation. Ventilation was appropriate for work done in all subjects. Maximum ventilation as a percentage of the 15 second MVV was similar in both groups and, as is expected in normal individuals, was well below the MVV (67 f 9% for controls, 61 20% for AN). At steady state, oo2was appropriate for work done. Q was maintained in the AN group (104 14% predicted) (Fig. 3), but with a relatively high HR (107 & 13% predicted) and low SV (86 ? 19% predicted). Stroke volume correlated weakly with both indices of nutritional status for the combined groups: PIWT, r = 0.51; P C 0.03 (Fig. 4) and BMP, r = 0.52, P < 0.02. The relationship between ventilation and Vco2at steady state was in keeping with published normal values” (Fig. 5).
*
*
DISCUSSION
The present study has shown that impaired exercise ability in AN is related to poor nutritional status. Ventilatory responses are appropriate for Vco, and Wmax. There is preservation of Q but this is achieved with a lower SV and higher heart rate. The normal Q response coupled with hemoglobin concentrations greater than 10 g/100 mL would suggest that oxygen delivery is not a limiting factor for exercise. In conjunction with the normal ventilatory responses, it appears that exercise capacity in AN is limited by loss of muscle mass or dysfunction of existing muscle, as has been suggested by This inability to perform work in AN is consistent with previous studies.4*5.8-10
Cardiopulmonary Response 25
6o
.C
E 20
1 '.
0
-
in Anorexia Newosa
1
105
0
U
I-
3
k 3
15-
0 0
1
g
10-
4 V
0
5 !
I
0.5
I
0.5
I
1 .o 1.5 2.0 0X Y G,EN CON S U M PT I0N (I / rnin)
0
n
u
125
-
P)
c 0
.U
!! 100 -
L
K
> v1
0
__,6___
0
v
75
50
-
0
I
I
I
1.5
2.0
2.5
Vco, (L/rnin)
Output (dl at Fig. 3. OVgen consumption * O J 5096 of Wmax. Symbols as In Figure 2. Regression line from data of Godfrey." 150
1 .o
t
Fig. 5. Ventilation (VE) versus carbon dioxide production(V-J at Steady state. Symbols as In Figure 2. Regression from data Jones.17
Previous work in malnutrition has generally studied l o demonstrated cardiac function at r e ~ t . ~ , ~One , ~ study '' that the ejection fraction during supine work using radionuclide cineangiography did increase appropriately, but the workload at which measurements were made was not delineated and because of differences between upright and supine work, cardiac responses may not be comparable. I ' The present study using a non-invasive technique demonstrated that at Yi-Wmax, Q was maintained but with a smaller SV and increased HR. The question arises as to whether this compensation persists at Wmax. Previous work in normal subjects has shown that an individual attains more than 90% of maximum SV with
derived was maximal in the control subjects. The low HR previously observed in subjects with AN at rest is in contrast with the higher than expected one for work performed in the present study. The finding of a lower HR at maximal exercise compared to controls reported by NuMalnutrition can result in functional impairment of del et aL4 pertains only to absolute values. If the HR for muscle. Caloric deprivation tends to lead to selective controls and patients is expressed as a percent expected atrophy of type I1 fibers, decreasing the force-developing for work, the patients with AN have a relative tachycarpotential of a muscle and leading to premature fatigue.6 dia. Interpreting the relative resting bradycardia in AN Cellular integrity and intracellular ion contents can also patients with the echocardiographic data suggesting a be adversely affected.6 Subjects with AN often have de- normal for body mass at rest would clearly imply that pletion of potassium stores,6 and thus again, may be the resting SV of these patients is a larger percentage of predisposed to fatigue. Animal studies have shown that maximal than it is in normal subjects. Cardiac output fatigue is related to potassium efflux from exercising during exercise increases first by a joint increase in both muscle fibers.22There is also the possibility of a detrain- HR and SV until SV is maximal, whereupon further ing effect in the AN group. This may augment the cate- increases are due to increases in the HR alone. This cholamine response to exercise and hence increase the would suggest that patients with AN would be expected to achieve a maximal SV earlier in the course of exercise heart rate response. Fig. 4. Percent ideal weight (PIWT) versus stroke volume (SV). Symbols as in Figure 2. Correlation: r = 0.51, P < 0.03.
Lands et al.
106
than normals. The finding of an HR in the patients with AN well above that considered sufficient for a maximal SV in normals clearly would imply that a maximal SV was achieved in these patients. Evidence to the contrary would be a Q less than expected for oxygen consumption in patients with AN. This was not the case in the present study. In an effort to see if “cardiac insufficiency” (defined for the purpose of this study as an inappropriately low Q for work accomplished or was a factor, the SV was multiplied by the HR at Wmax to give an estimate of maximum Q. Using this estimate, Q in the AN group compared to controls was preserved even at Wmax. This would imply either that cardiac insufficiency at maximum work is not the main limiting factor to exercise or that the onset of cardiac insufficiency results in immediate cessation of work. The methods used in this study were not sufficiently sensitive to answer this question. In the present study SV correlated with nutritional status, albeit weakly. Part of the reason for the lack of a strong correlation is that SV, unlike Wmax, is calculated and not measured directly. Hence, the scatter in the relationship may be due to small inherent inaccuracies in any or all component parts of the determination of SV. It is unclear, however, to what extent this lower SV might be due to the elevated response in HR, a loss of cardiac muscle mass, decreased filling pressures, or decreased myocardial contractility. At the same workload, HR responses are higher when the work is performed by a small The subjects muscle mass as compared to a large with AN had low weight for height so that at any given they were performing with a reduced muscle mass, compared to the normal population. This may be the reason for the increased HR response in these patients. The finding of slight reductions in VC and FEV, ,and a mildly increased RV/TLC in the AN group is due to respiratory muscle weakness associated with malnutrition. This has been reported in a recent p~blication.’~ In conclusion, our subjects with AN demonstrated a reduced exercise capacity. The cardiac output response was adequate although this was achieved with a relatively elevated heart rate. Ventilatory responses appeared appropriate. Hemoglobin concentrations were not in a range to cause diminished oxygen delivery. This would suggest a peripheral limitation to exercise capacity, in large part due to the loss of muscle mass. The impact of peripheral muscle function on cardiopulmonary responses to exercise in patients with malnutrition associated with systemic or other organ system disease’.’ needs further investigation.
vo2)
vo2,
REFERENCES 1. Marcotte JE, Canny GJ, Grisdale R, Desmond K, Corey M,
Zinman R, Levison H, Coates AL. Effect of nutritional status on
exercise performance in advanced cystic fibrosis. Chest. 1986; 90:375-379. 2. Hortop J, Desmond KJ, Coates AL. The mechanical effects of
expiratory airflow limitation on cardiac performance in cystic fibrosis. Am Rev Respir Dis. 1988; 137:132-137. 3. Moodie DS, Salcedo E. Cardiac function in adolescents and young adults with anorexia nervosa. J Adolesc Health Care. 1983; 49-14. 4. Nude1 DB, Gootman N, Nussbaum MP, Shenker IR. Altered
exercise performance and abnormal sympathetic response to exercise in patients with anorexia nervosa. J Pediatr. 1984; 105:34-37. 5 . Davies CTM, Barnes C, Godfrey S. Body composition and maximal exercise performance in children. Hum Biol. 1972; 44:195214. 6. Russel DM, Walker PM, Lester LA, et al. Metabolic and struc-
7. 8.
9.
10.
11.
12.
13.
tural changes in skeletal muscle during hypocaloric dieting. Am J Clin Nutr. 1984; 39503-513. Webb JG, Kiess MC, Chan-Yan CC. Malnutrition and the heart. CMAJ 1986; 135:753-758. Heymsfield SB, Bethel RA, Ansley JD, Gibbs DM, Felner JM, Nutter DO. Cardiac abnormalities in cachectic patients before and during nutritional repletion. Am Heart J. 1978; 9 5 5 8 4 5 9 4 . St. John Sutton MG, Plappert T, Crosby L, Douglas P, Mullen J , Recchek N. Effects of reduced left ventricular mass on chamber architecture, load and function: a study of anorexia nervosa. Circulation. 1985; 72:991-1000. Goltdiener JS, Gross HA, Henry WL, Borer JS, Ebert MH. Effects of self-induced starvation on cardiac size and function in anorexia nervosa. Circulation. 1978; 58:425-433. Marcotte JE, Perrault H, Drblik SP. Caughlan M, Lamarre A. Hernodynamic response to upright and supine exercise in mild to severe cystic fibrosis patients. Am Rev Respir Dis. 1989; 139:A7 I. Work Group to Revise DSM-111. Diagnostic and Statistical Manual of Mental Disorders, 3rd ed, revised. Washington, DC: American Psychiatric Association, 1987. Coates AL, Boyce P, Muller D, Mearns M, Godfrey S. The role of nutritional status, airway obstruction, hypoxia, and abnormalities in serum lipid composition in limiting exercise tolerance in children with cystic fibrosis. Acta Paediatr Scand. 1980;
69:353-358. 14. Knudson RJ, Slatin RC, Lebowitz MD, Burrows B. The maximal expiratory flow-volume curve. Am Rev Respir Dis. 1976; I13:587400. 15. Desmond KJ, Demizio DL, Allen PD, Beaudry PH, Coates AL.
An alternate method for the determination of functional residual capacity in a plethysmograph. Am Rev Respir Dis. 1988; 1371273-276. 16. Polgar G, Promadhat V.Pulmonary Function Testing in Children. Philadelphia: WB Saunders, 1971. 17. Jones NL. Clinical Exercise Testing. Philadelphia: WB Saunders, 1988. 18. Godfrey S . Exercise Testing in Children. Philadelphia: WB Saunders, 1974. 19. Coates AL, Desmond K, Asher MI, Hortop J , Beaudry PH. The
effect of digoxin on exercise capacity and exercising cardiac function in cystic fibrosis. Chest. 1982; 82543-547. 20. Mahler DA, Matthay RA, Snyder PE, Neff RK, Loke J . Determination of cardiac output at rest and during exercise by carbon dioxide rebreathing method in obstructive airway disease. Am Rev Respir Dis. 1985; 131:73-78. 21. McHardy GJR. Relationship between the difference in pressure and content of carbon dioxide in arterial and venous blood. Clin Sci. 1967; 32:299-309. 22. Lindinger MI, Heigenhauser GJF. Ion fluxes during tetanic stimu-
Cardiopulmonary Response in Anorexia Nervosa lation in isolated perfused rat hindlimbs. Am J Physiol. 1988; 254:Rl17-R126. 23. Astrand P-0, Cuddy TE, Saltin B, Steinberg J. Cardiac output during sub-maximal and maximal work. J Appl Physiol. 1964; I 9:268-274. 24. Lewis SF, Taylor WF, Graham RM, Pettinger WE, Schutte JG, Blomquist CG. Cardiovascular responses to exercise as functions
107
of absolute and relative workload. J Appl Physiol. 1983; 54: 13141323. 25. Lands L, Desmond KJ,Demizio D, Pavilanis A, Coates AL. The effects of nutritional status and hyperinflation on respiratory muscle strength in children and young adults. Am Rev Respir Dis. 1990; 141:1506-1 509.
