Energy expenditure and substrate utilization in adults with cystic

fibrosis and diabetes mellitus1,2

Sheila A Ward, Jean L Tomezsko, Douglas S Holsclaw, and Albert M Paolone

ABSTRACT survival (2). Management has become even more complicated

Background: The onset of cystic fibrosis–related diabetes mel- with the development of impaired glucose tolerance (26–75% of
litus (CFDM) is often associated with a decline in clinical and patients) and diabetes mellitus (DM) (2.5–12%) probably
nutritional status. because of the destruction of pancreatic tissue. The mean age of
Objective: The purpose of this study was to characterize energy diagnosis for overt diabetes, 19–20 y, is remarkably consistent
expenditure (EE) and substrate utilization during rest, exercise, and the age of 20 y seems to mark a point of rapid decline in the
and recovery from exercise in patients with CF diagnosed with survival of patients with CF-related DM (CFDM), with < 25%
diabetes mellitus. reaching the age of 30 y compared with 60% of the nondiabetic
Design: EE, substrate utilization, minute ventilation, tidal vol- patients with CF (3). Other studies have shown conflicting
ume, and respiratory rate were calculated by indirect calorimetry results with regard to survival, pulmonary function, and nutri-
during rest; a 30-min, low-to-medium-intensity exercise bout on tional status (4–6). The Cystic Fibrosis Foundation has outlined
a treadmill; and a 45-min postexercise recovery period (in reclin- guidelines for CFDM and stressed the complexity of treatment
ing position) in 10 CF, 7 CFDM, and 10 control subjects between of a second chronic illness (7).
18 and 45 y of age. The struggle to maintain good nutritional status continues
Results: In all 3 periods, minute ventilation was higher in the CF throughout life. Causes of malnutrition and growth failure in CF
and CFDM groups than in the control subjects (P < 0.01). Dur- are an inadequate energy intake, some malabsorption despite
ing rest and exercise, the CF and CFDM groups maintained EE pancreatic enzyme supplementation, and increased energy
values at the high end of the normal range of the control sub- expenditure (EE). EE is the energy used for work and heat pro-
jects. However, during recovery, EE was higher in the CF and duced by the body to sustain life processes and activity.
CFDM groups than in the control group (P < 0.01). Increased EE in patients with CF results from respiratory infec-
Conclusions: EE may be higher than usual for the patients with tions, inflammation, fever, theorized increased work of breath-
CF and CFDM during periods of recovery from mild exercise or ing, medications, and possibly the basic genetic defect itself.
activity because of increased work of breathing consistent with Loss of glucose energy in the urine, altered metabolism, along
higher ventilatory requirements. This information may be useful with a possible increased EE can easily put patients with CF and
for patients receiving nutritional counseling who may choose to DM in negative energy balance (8).
exercise regularly, but are concerned about possible weight loss. This study was undertaken to provide a better understanding
Am J Clin Nutr 1999;69:913–9. of CFDM and its role in nutritional status. The purpose of this
study was to characterize the EE and pattern of substrate utiliza-
KEY WORDS Cystic fibrosis, diabetes mellitus, dietary tion of stable ambulatory adult patients with CF and CFDM dur-
intake, energy expenditure, oxygen consumption, exercise, ing rest, exercise, and recovery from exercise and to examine
adults, humans body composition, ventilatory requirements, selected substrates,
and nutrient intakes.

INTRODUCTION
Cystic fibrosis (CF) is the most common, severe genetic dis-
1
order in the United States. A dramatically improved median age From the Pediatric Pulmonary and Cystic Fibrosis Centers, Hahnemann
University Hospital, Philadelphia; the Department of Physical Education,
of survival (29 y) has been reported in recent years (1) because
Temple University, Philadelphia; and the Department of Physical Education
of advances in a variety of treatment regimens. In addition to
and Exercise Science, Norfolk State University, Norfolk, VA.
improved antibiotics and respiratory medications, nutritional 2
Address reprint requests to SA Ward, 2401 Corprew Avenue, Echols
care has improved. In conjunction with pancreatic enzyme sup- Hall, Room 167, Norfolk State University, Norfolk, VA 23504. E-mail:
plements, a high-energy diet with unrestricted fat has greatly [email protected].
aided the prevention of growth failure and malnutrition, both of Received November 13, 1997.
which have been associated with decreased quality of life and Accepted for publication October 20, 1998.

Am J Clin Nutr 1999;69:913–9. Printed in USA. © 1999 American Society for Clinical Nutrition 913
914 WARD ET AL

SUBJECTS AND METHODS Physical characteristics and body composition

Subjects Measurements of weight, height, skinfold thicknesses (biceps,

triceps, subscapular, and suprailiac with Lange calipers; Cam-
A total of 27 men and women aged 18–45 y were recruited bridge Scientific Instruments, Cambridge, MD), and midarm cir-
and represented 3 study groups. The subjects with CF were cumference were made in triplicate by one observer (SAW).
recruited from the CF patient populations at Hahnemann Uni- Body density (9); percentage body fat, fat mass, and fat-free
versity Hospital, the Medical College of Pennsylvania, and the mass (FFM; 10); and midarm muscle circumference (11) were
University of Pennsylvania Medical Center (all in Philadelphia). also measured. Percentage ideal body weight (12), body surface
CF patients were grouped according to those without DM (CF area (by nomogram), and body mass index (13) were calculated.
group, n = 3 women and 7 men) and those with DM (CFDM
group, n = 2 women and 5 men). All patients were clinically sta- Ventilatory indexes
ble. No patient with CF was allowed to participate if they were Pulmonary function tests were performed with a Spiro Ana-
severely malnourished (≤ 75% of ideal body weight), receiving lyzer ST-90 (Futuremed, Deer Park, NY). The forced vital capac-
supplemental oxygen during the daytime, or taking thyroid or ity (FVC) and forced expiratory volume in 1 s (FEV1), which
cardiac medication. evaluate the resistance properties of the airways and the strength
Eight of the 10 subjects in the CF group and all the CFDM of the expiratory muscles, were determined (14). Minute venti-
patients were pancreatic insufficient. Eight of the 10 subjects in · · ·
lation (VE), tidal volume (VT), oxygen consumption (VO2), and
the CF group were using bronchodilators, 1 subject was not, and respiratory rate (RR) were also measured during the same time
1 subject’s bronchodilator status was unknown. All CFDM sub- periods and with the same procedure as described for EE.
jects were using bronchodilators. However, on the morning of
the study, use of bronchodilators and all other medications was Blood analyses
not allowed. Of the 10 patients in the CF group, 3 were DF508 Approximately 16 mL blood was drawn from a catheter
homozygous, 4 were DF508 heterozygous (other mutations were inserted into the left brachial vein at the end of the rest, exercise,
S549N and W1282X with 2 genes unidentified), 2 were and postexercise recovery periods. After centrifugation at 4 8C
W1282X/3849+10, and 1 was G551D with an unidentified gene. for 10–15 min at 2739–5590 3 g (3500–5000 rpm), all blood
Of the 7 patients in the CFDM group, 1 was DF508 homozygous samples were separated, stored at 240 8C, and analyzed accord-
and 5 were DF508 heterozygous (other genes were 621+1 and ing to hospital laboratory procedures at Hahnemann University
R553X with 2 genes unidentified; one was 441 with an unidenti- Hospital (glucose, triacylglycerol, and urea) or the General Clin-
fied gene. ical Research Center at Temple University (catecholamines,
Control subjects (n = 5 women and 5 men) were healthy with glucagon, and insulin).
no known medical conditions and were recruited from hospital Blood drawn for triacylglycerol and urea measurements was
and community sources. The control group was age- and weight- placed in a heparin-containing tube, shaken, and placed on ice
matched to the CF and CFDM groups, but not individually so. until centrifuged and analyzed with the glycerol phosphate oxi-
The duration of diabetes in the CFDM group was 6 ± 3.9 y and dase enzymatic (Boehringer Mannheim, Mannheim, Germany)
the CFDM group was significantly older than the control group and urease method (Boehringer Mannheim), respectively. Blood
(P < 0.03). No persons known to be alcohol or drug addicted or samples for glucagon measurements were placed in refrigerated
febrile the morning of the study were allowed to participate. The tubes containing 0.2 mL Trasylol (Bayer, West Haven, CT),
study was conducted at Hahnemann University Hospital. All sub- shaken, and placed on ice until centrifuged and analyzed by
jects gave written consent and the protocol was approved by the radioimmunoassay. Blood samples for catecholamine analyses
Committee on Human Subjects, Hahnemann University Hospital. were placed in refrigerated Amersham tubes (Piscataway, NJ),
shaken, placed on ice, centrifuged within 15 min, and analyzed
Study design
(15). Blood samples for insulin measurements were clotted at
The CF, CFDM, and control groups were studied in the early room temperature for 30 min, centrifuged, and analyzed by the
morning after a 12-h overnight fast. Subjects in the CFDM group standard double-antibody immunoassay (Pharmacia Diagnostics,
were studied 8–10 h after their last regularly scheduled meal and Uppsala, Sweden). Blood samples for glucose measurement were
evening medication administration. Water was allowed and clotted at room temperature for 15–20 min, centrifuged, and ana-
encouraged during the fasting period and all medications and lyzed by the hexokinase method (Boehringer Mannheim).
regular therapies, including chest physiotherapy and bron-
chodilators, were withheld. Screening blood glucose tests were Dietary intake
performed and temperatures were taken on the morning of the Nutrient intakes were derived from a self-reported dietary
study. Patients in the CFDM group were also screened after the record maintained by the subjects for 3 d before the study day
exercise and recovery periods. Measurements were taken for after they had been given verbal and written instructions. Sub-
weight, height, body composition, and lung status. Subjects jects weighed, measured, and recorded food intakes. All enzyme
reclined quietly and rested for 30 min before baseline measure- supplements for each meal were also recorded. Energy intake
ments of EE, substrate utilization, ventilatory indexes, and and dietary analyses were conducted by using a computerized
selected blood indexes were made; the same measurements were program, NUTRITIONIST III (Salem, OR).
made during the exercise and recovery periods. No medications
or therapies, such as bronchodilators, were administered after the Energy expenditure and substrate utilization
exercise period. For patients in the CFDM group, poststudy EE and substrate (carbohydrate, fat, and protein) utilization
medication was adjusted and given with the poststudy meal were calculated from the amount of oxygen utilized and carbon
immediately after the study was completed. dioxide produced during 3 time periods on the same day by
 CF, DIABETES MELLITUS, AND ENERGY EXPENDITURE 915

TABLE 1
Physical characteristics, body-composition, resting energy expenditure (REE), and lung function values for the 3 study groups1
CF group CFDM group Control group
Variable (n = 7 M and 3 F) (n = 5 M and 2 F) (n = 5 M and 5 F)
Age (y) 31.1 ± 7.30 33.1 ± 1.762 25.7 ± 4.00
Weight (kg) 63.8 ± 8.46 63.0 ± 12.10 61.8 ± 11.37
Height (cm) 172.4 ± 10.00 169.0 ± 6.23 167.9 ± 9.51
Percentage of IBW (%) 98.5 ± 15.04 104.9 ± 14.09 102.7 ± 11.20
Percentage body fat (%) 19.6 ± 8.02 20.4 ± 6.00 23.1 ± 6.53
Fat-free mass (kg) 51.5 ± 9.87 50.0 ± 9.25 47.9 ± 11.36
Fat mass (kg) 12.2 ± 4.40 13.0 ± 4.77 13.9 ± 3.19
Body mass index 21.4 ± 1.96 21.9 ± 3.29 22.0 ± 2.21
Body surface area (m2) 1.75 ± 0.16 1.71 ± 0.18 1.70 ± 0.20
Skinfold-thickness sum (mm) 42.4 ± 16.94 45.9 ± 16.49 48.2 ± 12.08
REE
(kJ/d) 6774.3 ± 1077.80 6957.6 ± 1459.80 6083.5 ± 1155.62
(kJ ? kg body wt 21 ? d21) 107.1 ± 17.57 112.5 ± 24.69 98.7 ± 9.62
(kJ ? kg FFM 21 ? d21) 134.7 ± 28.03 141.0 ± 26.36 129.3 ± 15.48
Predicted REE (kJ) 6572.6 ± 830.11 6391.9 ± 817.97 6428.3 ± 963.16
Percentage of predicted REE (kJ) 104.0 ± 17.32 109.2 ± 20.15 94.4 ± 7.95
FEV1 (% of predicted) 45.7 ± 15.60 50.3 ± 22.20 95.5 ± 13.703
FVC (% of predicted) 62.6 ± 12.20 64.9 ± 17.10 92.3 ± 14.00
1–
x ± SD. CF, cystic fibrosis; CFDM, CF plus diabetes mellitus; IBW, ideal body weight; FFM, fat-free mass; FEV1, forced expiratory volume in 1 s;
FVC, forced vital capacity.
2
Significantly different from the control group, P < 0.05 (post hoc Tukey-Kramer’s test).
3
Significantly different from the CF and CFDM groups, P < 0.001 (post hoc Tukey-Kramer’s test).

indirect calorimetry with a MetaScope Metabolic Analyzer body weight, and body composition in comparisons by group
(Cybermedic, Boulder, CO). This machine used a paramagnetic (17). Specifically, a model was constructed that included sex,
oxygen analyzer and an infrared carbon dioxide analyzer. The body composition (as FFM), weight, and group (CF, CFDM, and
hood flow range was 0–40 L. The machine flow was autoset to control groups) in a multiple general linear model procedure. EE
1 of 16 speeds based on 0.5% CO2 and calibrated at the begin- was analyzed by using a repeated-measures (within 3 between,
ning of each subject session. A comfortable one-way valve, or mixed) ANOVA model constructed for each transformation of
CPAP cushion flex, ventilated mask (Bird Life Design, Dallas) the dependent variable (Table 2). Contrast testing was per-
was worn on the face to cover the mouth and nose while expired formed post hoc for group comparisons in the recovery time
gases were collected and analyzed. In a quiet, thermoneutral period. A one-way ANOVA with Tukey-Kramer’s test was used
room, EE was measured during rest for 45 min while the subject to test for significant differences between groups in age, weight,
reclined comfortably; during a 30-min, low-to-medium-inten- height, percentage ideal body weight, percentage body fat, FFM,
sity exercise bout (95 kg ? m21 ? min21) on a treadmill; and dur- fat mass, body mass index, body surface area, and the sum of
ing a 45-min postexercise recovery period (in a reclining posi- skinfold thicknesses. Significance was set at the 0.05 level in all
tion). Data for the first 10 min of the resting EE (REE) tests.
measurement were deleted. During the exercise period, the rate
of perceived exertion, heart rate, blood pressure, and oxygen
saturation were monitored. The predicted EE values were cal- RESULTS
culated by using WHO equations based on age, weight, and sex The physical characteristics, body composition, baseline and
(16). Before the study, the reliability and validity of the method REE values, and resting lung function of the 3 groups are pre-
were verified with multiple studies of individual subjects and a sented in Table 1. The diabetic status of the CFDM group was as
methanol burn test, respectively. A 60-mL urine sample was col- follows: stable to well controlled in 5 patients, fair in 1 patient,
lected during the study period, from which 24-h urinary urea and unknown in 1 patient. In the CFDM group, diabetes was
excretion was determined to calculate the nonprotein respira- treated with insulin in 6 subjects and with the hypoglycemic
tory quotient (Hahnemann University Hospital). The respiratory agent Glucotrol (Pratt Pharmaceuticals, New York) in 1 subject.
quotient was calculated as carbon dioxide production divided by There were no significant differences in any of the body-compo-
·
VO2, both of which were measured by using the metabolic cart. sition or anthropometric data between groups. All 3 groups had
good nutritional status on the basis of mean values for percentage
Statistical analyses of ideal body weight. Two subjects in the CF group had mild-to-
Substrate utilization, ventilatory indexes, and blood data were moderate malnutrition, one subject in the CFDM group was
analyzed by using a 3 3 3 factorial analysis of variance underweight, and one control subject had mild malnutrition (12).
(ANOVA) (version 4.1; SPSS, Inc, Chicago) with repeated
measures across each time period and group. For the initial EE and substrate utilization
analysis of EE between groups (Table 1), multivariate ANOVA There was no significant difference in EE between the 3 groups
(MANOVA) was used to account for the known effects of sex, during rest or exercise (Table 2). However, during recovery, EE
916 WARD ET AL

TABLE 2
Energy expenditure in the 3 study groups during rest, exercise, and recovery1
Energy expenditure CF group (n = 10) CFDM group (n = 7) Control group (n = 10)
Rest
(kJ/min) 4.7045 ± 0.7489 4.8317 ± 1.0125 4.2246 ± 0.8033
(kJ ? kg body wt21 ? min21) 0.0745 ± 0.0121 0.0782 ± 0.0171 0.0686 ± 0.0067
(kJ ? kg FFM21 ? min21) 0.0937 ± 0.0197 0.0979 ± 0.0180 0.0895 ± 0.0109
Exercise
(kJ/min) 15.9841 ± 2.8911 15.8992 ± 3.2551 15.0674 ± 2.6275
(kJ ? kg body wt21 ? min21) 0.2531 ± 0.0498 0.2540 ± 0.0314 0.2498 ± 0.0527
(kJ ? kg FFM21 ? min21) 0.3192 ± 0.0741 0.3205 ± 0.0464 0.3268 ± 0.0711
Recovery
(kJ/min) 5.0526 ± 0.9330 5.3497 ± 0.7782 3.7982 ± 0.80332
(kJ ? kg body wt21 ? min21) 0.0795 ± 0.0113 0.0862 ± 0.0130 0.0619 ± 0.00922
(kJ ? kg FFM21 ? min21) 0.0996 ± 0.0163 0.1079 ± 0.0117 0.0807 ± 0.01252
1–
x ± SD. CF, cystic fibrosis; CFDM, CF plus diabetes mellitus; FFM, fat-free mass.
2
Significantly different from the CF and CFDM groups, P < 0.01 (post hoc Tukey-Kramer’s test).

was significantly higher in the CF and CFDM groups than in the pmol/L) and recovery (2.08 ± 1.12 pmol/L), and norepinephrine
control group when measured as kJ/min, kJ ? kg body wt21 ? min21, was significantly higher (P < 0.001) during exercise
and kJ ? kg FFM21 ? min21. There were no significant differences (0.033 ± 0.019 nmol/L) than during rest (0.019 ± 0.009 nmol/L)
between the 3 groups in the incremental change of any EE variable and recovery (0.021 ± 0.009 nmol/L). There were no significant
between the rest and exercise periods and between the exercise differences in triacylglycerol, urea, insulin, or glucagon across
and recovery periods. There were no significant differences in any the 3 time periods (rest, exercise, and recovery) in the CF,
of the substrate utilization variables between the 3 groups during CFDM, and control groups and no significant differences in the
exercise or recovery (Table 3). Neither the CF nor the CFDM respiratory quotient, the nonprotein respiratory quotient, or glu-
groups reached the dietary goal of > 120% of the recommended cose variables for the CF and control groups.
dietary allowance for energy or the suggested energy intake from In the CFDM group, the respiratory quotient was significantly
fat of 35–40% (18, 19). higher during exercise than during rest and recovery and signifi-
cantly higher during rest than during recovery (P < 0.04). The
Ventilatory indexes nonprotein respiratory quotient was significantly higher during
Ventilatory indexes during rest, exercise, and recovery are rest than during exercise and recovery and significantly higher
· ·
presented in Table 4. During rest and exercise, VT, RR, and VO2 during exercise than during recovery (P < 0.03). The blood glu-
21 21
(mL ? kg ? min ) were not significantly different between the 3 cose concentration was significantly higher (P < 0.01) during
·
groups; VE was significantly higher in the CF and CFDM groups rest (8.60 ± 2.90 mmol/L) than during exercise (7.28 ± 1.61
·
than in the control group. During recovery, VT was significantly mmol/L) and recovery (7.29 ± 2.09 mmol/L). The percentage
higher in the CFDM group than in the control group (P < 0.02) carbohydrate utilization was significantly higher in the CF and
· ·
and VE and VO2 were significantly higher in the CF and CFDM control groups during exercise than during recovery (P < 0.01)
groups than in the control group (P < 0.001). There was no signi- and significantly higher in the CFDM group during exercise than
ficant difference in the RR between groups. during rest and recovery (P < 0.01). Within all 3 groups, the per-
centage of protein utilization was significantly lower during
Blood indexes exercise than during rest and recovery (P < 0.001), but there
Differences in all of the blood indexes were examined were no significant differences in the percentage lipid utiliza-
between the 3 study groups. Glucagon was higher in the CFDM tion. There were also no significant differences in the incremen-
group (162.5 ± 3.63 ng/L) than in the CF group (102.4 ± 32.3 tal changes in substrate utilization between time periods
ng/L) during rest (P < 0.04). Differences in all variables were (between rest and exercise and between exercise and recovery)
also examined within each study group. Within the CF group, across the 3 groups.
the blood concentration of epinephrine was significantly higher
(P < 0.01) during exercise (3.96 ± 1.93 pmol/L) than during rest
(2.22 ± 1.21 pmol/L). Norepinephrine was also significantly DISCUSSION
higher (P < 0.01) during exercise (0.026 ± 0.006 nmol/L) than Both CF groups were well matched in lung disease severity.
during rest (0.015 ± 0.005 nmol/L) and recovery (0.014 ± 0.003 There were no significant differences between groups in anthro-
nmol/L). Within the CFDM group, epinephrine was significantly pometric measures, body composition, or nutritional status. In
higher (P < 0.01) during exercise (5.11 ± 3.16 pmol/L) than dur- the recovery phase, EE was significantly higher in the CF and
ing rest (2.97 ± 1.77 pmol/L) and recovery (3.17 ± 1.41 pmol/L), CFDM groups than in the control group. This difference was
· · ·
and norepinephrine was significantly higher (P < 0.01) during supported by higher VE, VT, and VO2 values in the CF and CFDM
exercise (0.026 ± 0.011 nmol/L) than during rest (0.017 ± 0.005 groups than in the control group.
nmol/L) and recovery (0.017 ± 0.006 nmol/L). Within the con- The respiratory muscles of patients with CF, regardless of the
trol group, epinephrine was significantly higher (P < 0.001) dur- presence of diabetes, work harder. The respiratory muscles must
ing exercise (4.57 ± 4.10 pmol/L) than during rest (2.54 ± 2.63 generate more force to provide sufficient oxygen and gas
 CF, DIABETES MELLITUS, AND ENERGY EXPENDITURE 917

TABLE 3 differences between CF patients and matched control subjects in

Substrate utilization in the 3 study groups during rest, exercise, and incremental increases in energy from resting to each of several
recovery1 activities of daily living, including 2 levels of exercise. Although
CF group CFDM group Control group EE was higher (NS) during rest and exercise in the CF and
Variable (n = 10) (n = 7) (n = 10) CFDM groups than in the control group, the CF and CFDM
Rest
groups were able to compensate and maintain EE values within
RQ 0.85 ± 0.04 0.85 ± 0.062 0.83 ± 0.03 the range of those of the control group. Patients with CF and
NPRQ 0.87 ± 0.06 0.90 ± 0.153 0.83 ± 0.04 significant pulmonary involvement have an enlarged physiologic
Carbohydrates (%) 45.0 ± 13.0 39.0 ± 23.3 38.0 ± 11.6 dead space and hypoxemia at rest (21). Physiologic dead space
Lipids (%) 34.5 ± 16.8 33.5 ± 22.3 42.1 ± 13.3 is enlarged because any ventilation to scarred, infected, or
Protein (%) 20.5 ± 12.0 27.5 ± 23.7 19.9 ± 7.4 blocked portions of the lungs is wasted. During exercise in per-
Exercise sons without respiratory distress, the physiologic dead space nat-
RQ 0.85 ± 0.05 0.86 ± 0.034 0.84 ± 0.03 urally decreases (22). If, however, physiologic dead space (not
NPRQ 0.85 ± 0.05 0.86 ± 0.045 0.84 ± 0.04 measured in this study) also decreases during exercise, it might
Carbohydrates (%) 50.6 ± 17.96 52.7 ± 13.57 47.5 ± 11.36
explain why EE variables were not significantly different
Lipids (%) 43.0 ± 18.0 38.7 ± 9.8 46.7 ± 12.6
Protein (%) 6.4 ± 3.78 8.6 ± 6.48 5.8 ± 3.08
between the CF patients and the control subjects during exercise,
Recovery but were significantly different during recovery.
RQ 0.83 ± 0.05 0.82 ± 0.07 0.83 ± 0.05 An elevated REE has been documented in patients with CF (21,
NPRQ 0.84 ± 0.07 0.82 ± 0.09 0.84 ± 0.07 23–27). The REE in normal adults varies greatly, depending on
Carbohydrates (%) 37.7 ± 16.2 38.2 ± 21.4 35.0 ± 16.1 many factors, particularly the amount of metabolically active tis-
Lipids (%) 42.6 ± 20.8 33.3 ± 14.7 44.0 ± 15.1 sue. The range of daily expended energy in severely ill non-CF
Protein (%) 19.7 ± 12.6 28.5 ± 23.1 21.0 ± 9.5 patients varies as well (28). In CF patients, REE appears to
1–
x ± SD. CF, cystic fibrosis; CFDM, CF plus diabetes mellitus; RQ, res- increase as lung function declines (29). In our study, the CF and
piratory quotient; NPRQ, nonprotein respiratory quotient. CFDM groups had subjects with predicted REE values that were
2,5,6
Significantly different from recovery: 2 P < 0.04, 5 P < 0.03, 6 P < 0.01. as high as 130% and 137%, respectively, whereas values in the
3
Significantly different from exercise and recovery, P < 0.03. control group remained within normal limits, the highest predicted
4,7,8
Significantly different from rest and recovery: 4 P < 0.04, 7 P < 0.01, value being 110%. Despite very low pulmonary function, the REE
8
P < 0.001.
of the CF and CFDM groups was not as high as expected. Within
each group, regression analysis of the percentage of predicted
FEV1 was compared with the following indexes: percentage ideal
exchange because of airway obstruction and resistance, such as body weight, percentage of predicted REE, and percentage body
decreased lung compliance. This additional work requires addi- fat. The only significant correlation found was between FEV1 and
·
tional energy. Furthermore, VE was higher in the CF and CFDM percentage body fat in the CF group (r = 0.67, P < 0.05). No
groups to compensate for the ventilation that is wasted because adjustments were made for multiplicity of planned comparisons.
it never reaches the blood (ventilation-to-perfusion mismatch). The most important function of glucagon is to increase blood
No incremental changes in EE were observed between the CF glucose concentrations, mainly via hepatic glycogenolysis and
and CFDM groups and the control group. This finding was con- increased hepatic gluconeogenesis. During rest, the blood
sistent with that of Grunow et al (20), who found no significant glucagon concentration of the CFDM group was significantly

TABLE 4
Ventilatory indexes in the 3 groups during rest, exercise, and recovery1
Variable CF group (n = 10) CFDM group (n = 7) Control group (n = 10)
Rest
·
VE (L/min) 8.5 ± 1.6 8.5 ± 1.5 6.4 ± 1.12
·
VT (mL) 480.0 ± 116.8 464.0 ± 111.3 431.0 ± 142.8
RR (breaths/min) 18.6 ± 5.1 18.8 ± 3.7 15.7 ± 3.6
·
VO2 (mL ? kg21 ? min21) 3.8 ± 0.61 3.9 ± 0.73 3.5 ± 0.37
Exercise
·
VE (L/min) 23.3 ± 4.9 23.7 ± 4.1 18.2 ± 3.02
·
VT (mL) 819.4 ± 197.4 840.4 ± 138.3 762.1 ± 205.8
RR (breaths/min) 29.5 ± 9.0 28.5 ± 5.1 25.1 ± 6.2
·
VO2 (mL ? kg21 ? min21) 12.6 ± 2.5 12.7 ± 1.6 12.5 ± 2.6
Recovery
·
VE (L/min) 8.8 ± 2.0 9.4 ± 1.2 6.0 ± 1.32
·
VT (mL) 472.4 ± 119.1 498.7 ± 100.63 360.6 ± 88.9
RR (breaths/min) 19.5 ± 4.8 19.3 ± 3.6 16.9 ± 2.8
·
VO2 (mL ? kg21 ? min21) 4.1 ± 0.56 4.5 ± 0.78 3.2 ± 0.472
1– · · ·
x ± SD. Normal values based on clinical standards are as follows: VE, 6.0–7.5 L/min; VT, 500 mL; RR, 12–15 breaths/min. VE, minute ventilation;
· ·
VT, tidal volume; RR, respiratory rate; VO2, oxygen consumption; CF, cystic fibrosis; CFDM, CF plus diabetes mellitus.
2
Significantly different from the CF and CFDM groups, P < 0.01 (post hoc Tukey-Kramer’s test).
3
Significantly different from the control group, P < 0.02 (post hoc Tukey-Kramer’s test).
918 WARD ET AL

higher than that of the CF group. Increased glucagon secretions ments to compensate for a ventilation-to-perfusion mismatch.
have been reported in type 1 and type 2 diabetic patients (3) and This additional energy requirement should be considered by CF
in patients with severe diabetic ketoacidosis (30). In patients and CFDM patients who choose to exercise regularly and should
with CFDM, glucagon concentrations have been described as be incorporated into nutritional counseling goals intended for
normal or reduced (3, 30). Both hypoglucagonemia and hyper- these patients. Some patients express concern about weight loss
glucagonemia have been reported in other studies (31). Kien et from exercise. Accurate assessments of energy use patterns in CF
al (32) found elevated hepatic glucose production in children patients, which may differ from those in persons without CF, will
with CF. The CFDM group in our study had a higher glucagon allow patients to exercise comfortably, receive health benefits
concentration than the CF and control groups during rest (P < associated with regular exercise, and adjust their daily energy
0.05), exercise (NS), and recovery (NS). The blood glucagon intake to compensate for the additional energy requirement dur-
concentration in the CFDM group was possibly higher during ing exercise recovery. Further studies of the energy requirements
rest to guard against hypoglycemia because neither decreased of patients with CF or CFDM during recovery periods postexer-
plasma insulin concentrations nor increased fat oxidation were cise, with larger sample sizes and more in-depth blood analyses
observed after the overnight fast. The latter condition indicates a (eg, of fatty acids, glycerol, lactate, glucagon, and epinephrine
low blood glucose concentration or a shift in substrate utiliza- concentrations), are needed to advance the understanding of EE
tion, which would stimulate the release of glucagon. However, in these patients.
because of the large number of variables measured in the blood
profile and the small sample size, conclusions based on glucagon We thank Lisa Davis for her dedicated assistance and Kevin B Stansberry
data must be made with caution. for statistical assistance.
The CFDM group tended to have a higher epinephrine con-
centration than the other 2 groups at the end of each study REFERENCES
period. In addition, epinephrine concentrations in the CF and 1. FitzSimmons S. The changing epidemiology of cystic fibrosis.
CFDM groups remained higher than baseline values at the end of J Pediatr 1993;122:1–9.
2. Corey M, McLaughlin F, Williams M, Levison H. A comparison of
the recovery period, whereas the control group had an epineph- survival, growth and pulmonary function in patients with cystic
rine concentration that was lower than baseline at this time point. fibrosis in Boston and Toronto. J Clin Epidemiol 1988;41:588–91.
Despite these trends, no significant differences were found in 3. Dodge J, Morrison G. Diabetes mellitus in cystic fibrosis: a review.
epinephrine concentrations between the 3 groups and there was J R Soc Med 1992;85(suppl):25–8.
great variability in the catecholamine data. Catecholamines may 4. Reisman J, Corey M, Canny G, Levison H. Diabetes mellitus in
serve as indicators of stress and are known to increase the basal patients with cystic fibrosis: effect on survival. Pediatrics
metabolic rate by 7–15%, increase glucagon, decrease insulin 1990;86:374–7.
5. Finkelstein S, Wielinski C, Elliott G, et al. Diabetes mellitus asso-
production, increase glucose production by the liver (gluconeo- ciated with cystic fibrosis. J Pediatr 1988;112:373–7.
genesis), increase glycogenolysis, and increase lipolysis. In 6. Lanng S, Thorsteinsson B, Erichsen G, Nerup J, Koch C. Glucose
addition, chronically elevated epinephrine concentrations result tolerance in cystic fibrosis. Arch Dis Child 1991;66:612–6.
in hyperglycemia and fat store depletion. In this study, we were 7. Cystic Fibrosis Foundation. Consensus conference on CF-related
unable to conclude that circulating catecholamines in the CFDM diabetes mellitus. Consens Conf Concepts Care 1990;1:1–8.
group contributed to the increased EE found during recovery. In 8. Durie PR, Pencharz PB. Nutrition. Br Med Bull 1992;48:823–47.
9. Durnin J, Rahaman M. The assessment of the amount of fat in the
one study, adult patients with CF (without DM) were found to
human body from measurements of skinfold thickness. Br J Nutr
have significantly higher epinephrine concentrations than con- 1967;21:681–9.
trol subjects (33). We hypothesized that chronically elevated epi- 10. Siri W. Body composition from fluid spaces and density. In: Brozek
nephrine concentrations in patients with CF and CFDM may J, Henschell A, eds. Techniques for measuring body composition.
result in increased EE, hyperglycemia, increased fat store deple- Washington, DC: National Academy of Science, 1961:223–44.
tion, and poor nutritional status. We believe these observed 11. Blackburn G, Bistrian B, Maini B, Schlamm H, Smith M. Nutri-
trends in epinephrine need further study. tional and metabolic assessment of the hospitalized patient. JPEN J
Parenter Enteral Nutr 1977;1:11–22.
This study was not designed specifically with the power to
12. Ramsey B, Farrell P, Pencharz P. Nutritional assessment and man-
detect group differences in any particular index; thus, inferences agement in cystic fibrosis: a consensus report. The Consensus Com-
about nonsignificant comparisons are particularly susceptible to mittee. Am J Clin Nutr 1992;55:108–16.
type II errors. For instance, mean within-group EEs at rest and 13. Gettman L. Fitness testing. In: Durstine JL, King AC, Painter PL,
during exercise were not significantly different in the present Roitman JL, Zwiren LD, eds. American College of Sports Medicine
study; however, these results were subject to type II error and resource manual for guidelines for exercise testing and prescription.
thus should be interpreted with caution. On the other hand, these Philadelphia: Lea & Febiger, 1993:229–46.
14. Bullock J, Boyle J, Wang M. Physiology. 2nd ed. Baltimore: Williams
same comparisons may indicate trends that should be investi- & Wilkins, 1991.
gated in larger-scale studies not susceptible to type II error. 15. Cryer P. Catecholamines and metabolism. Introduction. Am J Phys-
In this study, patients with CF and CFDM were able to main- iol 1984;247:E1–3.
tain EE values within the normal range of the control group dur- 16. FAO/WHO/UNU. Energy and protein requirements. Report of a
· joint FAO/WHO/UNU Expert Consultation. World Health Organ
ing rest and exercise periods, despite a significantly higher VE.
However, during recovery from exercise, the CF and CFDM Tech Rep Ser 1985;724:1–206.
17. Ravussin E, Bogardus C. Relationship of genetics, age, and physi-
groups used more energy than the control group and not only
· · cal fitness to daily energy expenditure and fuel utilization. Am J
maintained a higher VE, but a higher VO2 as well. Our data sug- Clin Nutr 1989;49:968–75.
gest that the elevated EE in both the CF and CFDM patients, dur- 18. Wootton S, Murphy J, Bond S, Ellis J, Jackson. A Energy balance
ing periods of recovery from mild exercise or activity, was due to and growth in cystic fibrosis. J R Soc Med 1991;84(suppl):22–7.
increased breathing resulting from higher ventilatory require- 19. Pencharz P, Hill R, Archibald E, Levy L, Newth C. Energy needs
 CF, DIABETES MELLITUS, AND ENERGY EXPENDITURE 919

and nutritional rehabilitation in undernourished adolescents and Pediatr Res 1994;35:451–60.

young adult patients with cystic fibrosis. J Pediatr Gastroenterol 27. Thomson M, Wilmott R, Wainwright C, Masters B, Francis P, Shepherd
Nutr 1984;3(suppl):S147–53. R. Resting energy expenditure, pulmonary inflammation, and genotype
20. Grunow J, Azcue M, Berall G, Pencharz P. Energy expenditure in cys- in the early course of cystic fibrosis. J Pediatr 1996;129:367–73.
tic fibrosis during activities of daily living. J Pediatr 1993;122:243–5. 28. Goldstein S, Elwyn D. The effects of injury and sepsis on fuel uti-
21. Vaisman N, Pencharz P, Corey M, Canny G, Hahn E. Energy expen- lization. Annu Rev Nutr 1989;9:445–73.
diture of patients with cystic fibrosis. J Pediatr 1987;111:496–500. 29. Fried M, Durie P, Tsue L-C, Corey M, Levison H, Pencharz P. The
22. Wasserman K, Van Kessel A, Burton G. Interactions of physiologi- cystic fibrosis gene and resting energy expenditure. J Pediatr
cal mechanisms during exercise. J Appl Physiol 1967;22:71–85. 1991;119:913–6.
23. Buchdahl R, Cox M, Fulleylove C, et al. Increased resting energy 30. Strutchfield P, O’Halloran S, Teale J, Isherwood D, Smith C, Heaf
expenditure in cystic fibrosis. J Appl Physiol 1988;64:1810–6. D. Glycosylated haemoglobin and glucose intolerance in cystic
24. Hirsch J, Zhang S-P, Rudnick M, Cerny F, Cropp G. Resting oxygen fibrosis. Arch Dis Child 1987;62:805–10.
consumption and ventilation in cystic fibrosis. Pediatr Pulmonol 31. Zipf W. Non-traditional management of diabetes related to cystic
1989;6:19–26. fibrosis. Pediatr Pulmonol 1991;56:211–4.
25. O’Rawe A, McIntosh I, Dodge J, et al. Increased energy expenditure 32. Kien C, Horswill C, Zipf W, McCoy K, O’Dorisio T. Elevated
in cystic fibrosis is associated with specific mutations. Clin Sci hepatic glucose production in children with cystic fibrosis. Pediatr
1992;82:71–6. Res 1995;37:600–5.
26. Tomezsko J, Stallings V, Kawchak D, Goin J, Diamond G, Scanlin 33. Elborn J, Cordon S, Western P, McDonald J, Shale D. Tumour
T. Energy expenditure and genotype of children with cystic fibrosis. necrosis factor[hyphen]a, resting energy expenditure and cachexia