Determinants of Aerobic and Anaerobic Exercise

Performance in Cystic Fibrosis

ASHISH R. SHAH, DAVID GOZAL, and THOMAS G. KEENS
Division of Pediatric Pulmonology, Childrens Hospital Los Angeles; Department of Pediatrics, University of Southern California
School of Medicine, Los Angeles, California; and Constance S. Kaufman Pediatric Pulmonary Research Laboratory,
Departments of Pediatrics and Physiology, Tulane University School of Medicine, New Orleans, Louisiana

We examined aerobic and anaerobic exercise performance in 17 subjects with cystic fibrosis (CF) (age
25 6 10 [SD] yr; 47% females; FEV1 62 6 21% pred) and 17 age- and sex-matched control subjects
(age 25 6 8 [SD] yr; 41% females; FEV1 112 6 15% pred) in relation to pulmonary function and nutri-
.
tional status. Aerobic capacity was determined as maximal oxygen consumption (VO2max) (ml/kg/min)
.
and anaerobic threshold (AT; ml Vo2/kg/min) from a graded exercise stress test on an electronically
braked bicycle ergometer. Anaerobic performance was assessed from the average work of two bouts
.
of pedaling to exhaustion at a load corresponding to 130% V o2max from graded exercise. Both aero-
bic and anaerobic performances were decreased in subjects with CF (p , 0.001). The duration of
anaerobic exercise in subjects with CF was similar to control subjects. In control subjects, pulmonary
function did not correlate to aerobic or anaerobic exercise. In subjects with CF significant relation-
ships between FEV1, vital capacity, and FEF25–75% to AT were found, suggesting the pulmonary limita-
tion to aerobic capacity. In both patients with CF and control subjects, lean body mass and arm muscle
.
area significantly correlated with anaerobic performance but not with Vo2max or AT. We conclude
that nutritional status, rather than pulmonary function, is the major determinant of anaerobic exer-
cise capacity in CF. The preserved duration of anaerobic exercise at equivalent workloads (correspond-
.
ing to 130% of V o2max from graded exercise) suggests that readily available energy stores in muscle
may be similar in CF and normal individuals. Shah AR, Gozal D, Keens TG. Determinants of aero-
bic and anaerobic exercise performance in cystic fibrosis. AM J RESPIR CRIT CARE MED 1998;157:1145–1150.

Aerobic exercise in cystic fibrosis (CF) is limited by the inabil-
ity of the cardiorespiratory system to compensate for the in-
crease in metabolic demands inherent to sustained effort.
Regular exercise in patients with CF has been associated with
improved aerobic exercise endurance and quality of life (1–3).
Several training schedules have attempted to improve pulmo-
nary function and exercise tolerance in patients with CF with
varying success (2, 4–6). In a recent study, Nixon and col-
leagues demonstrated that patients with CF with greater aero-
bic exercise tolerance also had improved survival (7).
However, the type of exercise (aerobic versus anaerobic)
employed during training may have different effects on im-
provement of exercise tolerance. Aerobic exercise training
may not always improve baseline pulmonary function or
weight gain in patients with CF (2). Anaerobic exercise train-
ing, such as weight lifting, achieved significant improvements
in body weight and muscle strength (6). Such effects are not
surprising because interval training, a technique that employs
repeated bouts of anaerobic exercise, ultimately leads to sig-
(Received in original form May 9, 1997 and in revised form December 17, 1997)
Correspondence and requests for reprints should be addressed to Thomas G.
Keens, M.D., Division of Pediatric Pulmonology, Childrens Hospital Los Angeles,
MS#83, 4650 Sunset Blvd., Los Angeles, CA 90027.
Am J Respir Crit Care Med Vol 157. pp 1145–1150, 1998
nificant improvements in aerobic exercise performance in
both athletes and in individuals with obstructive lung disease
(8). Thus, improvements in anaerobic exercise tolerance may
lead to improvement of aerobic capacity in CF.
Daily activity frequently requires repeated bouts of anaer-
obic exercise (climbing a flight of stairs or short sprints to
catch a bus) rather than performance of sustained, aerobic ef-
forts. Therefore, understanding of anaerobic exercise perfor-
mance may be of more practical importance in the daily life of
a patient with CF than assessment of the more classic concept
of aerobic exercise endurance. Despite extensive research on
anaerobic exercise, it remains unclear whether patients with
obstructive lung disease, such as in CF, have limited anaerobic
capacity. The primary components that could account for such
a decline also remain unclear, although Boas and colleagues
recently reported a strong positive correlation in adolescent
male patients with CF between the Wingate Anaerobic Test
performance and body mass index (9).
We hypothesized that anaerobic exercise is compromised
in patients with CF. However, since anaerobic exercise in-
volves short and intense bursts of activity, we suspected that
malnutrition, rather than the severity of lung disease, will af-
fect anaerobic exercise capacity. To determine this, we com-
pared graded exercise tests (aerobic) to anaerobic exercise in
17 CF and 17 matched control subjects and attempted to es-
tablish the influence of pulmonary function and nutritional
status on aerobic and anaerobic exercise.
1146 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE
METHODS
Subjects
Seventeen subjects with CF and 17 age- and sex-matched control sub-
jects were studied. Subjects with CF were recruited from the Division
of Pediatric Pulmonology at Childrens Hospital Los Angeles. All sub-
jects had pulmonary symptoms, signs, and radiologic changes consis-
tent with CF and were diagnosed by pilocarpine iontophoretic sweat
chloride tests. All subjects with CF were treated with pancreatic en-
zymes, vitamins, and bronchodilators. Some subjects with CF were
also receiving antibiotics at the time of the study. Subjects with CF
were studied when clinically stable, either as outpatients or at the end
of a 2- to 3-wk hospital stay.
Control subjects were healthy adults recruited from members of
the hospital staff and their families. No control subjects smoked or
had cardiopulmonary disease.
Informed consent was obtained from each subject. The study was
approved by the Institutional Review Board of Childrens Hospital
Los Angeles.
Pulmonary Function Testing
All subjects underwent pulmonary function testing in the Pulmonary
Function Laboratory at the Childrens Hospital Los Angeles, located at
sea level (mean atmospheric pressure, 751 mm Hg). All measurements
for each subject were performed on the same day. The vital capacity
(VC) and its subdivisions were measured from a slow exhalation with
a wedge spirometer (Model 3000; Medscience, St. Louis, MO). The
best forced vital capacity, forced expiratory volume in 1 s (FEV1), and
mean forced expiratory flow during the middle half of forced vital ca-
pacity (FEF25–75%) obtained from forced expiration into the wedge
spirometer were selected and corrected for body temperature, pres-
sure saturated (BTPS). Functional residual capacity was measured with
a body pressure plethysmograph (2800 Autobox; Sensormedics, Yorba
Linda, CA) by the method of Dubois and coworkers (10). Residual
volume (RV) and total lung capacity (TLC) were calculated, and the
RV/TLC ratios were determined from the actual values. Individual
test results were analyzed and considered abnormal if they were
greater than 6 2 SD from available reference values appropriate for
age, height, and gender (11–14).
Nutritional Assessment
Nutritional assessment, consisting of anthropometric measurements
and the calculation of lean body mass, was performed. Anthropome-
try included the triceps, biceps, anterior superior iliac, and subscapu-
lar skinfolds and were measured using Lange skinfold calipers. Six
measurements were taken at each site, and the average of the six was
used in calculations. Midarm circumference, height, and weight were
also measured. Lean body mass (LBM) and arm muscle area (AMA)
were derived from standard equations (15–18).
TABLE 1
CHARACTERISTICS OF STUDY POPULATION*
Age (% females)
Height, m
Weight, kg
VC, % predicted
FEV1, % predicted
FEF25–75%, % predicted
RV/TLC
Resting SpO2
Arm muscle area, mm2
Lean body mass, % total body weight
Definition of abbreviations: VC 5 vital capacity; FEV1 5 forced expiratory volume in 1 s; FEF25–75% 5 forced expiratory flow from 25% to
75% of vital capacity; RV 5 residual volume; TLC 5 total lung capacity.
* Values are mean 6 SD; n 5 17 in both CF and control subjects.
†
p , 0.05.
‡
p , 0.01.
§
p , 0.001.
VOL 157 1998
Exercise Testing
During all exercise tests subjects breathed through a mouthpiece from
which inspired and expired gas concentrations were continuously ana-
lyzed with a rapid response zirconium O2 and infrared CO2 analyzers
using a computerized breath-by-breath exercise system (Sensormedics
4400; Sensormedics). Inhaled and exhaled tidal volumes were mea-
sured with a turbine digital volume transducer (Sensormedics 4400;
Sensormedics). From these, the following gas exchange parameters
·
were measured on a breath-by-breath basis: minute ventilation ( V E),
· ·
oxygen consumption ( V O2), carbon dioxide production ( V CO2), respi-
· ·
ratory exchange ratio (RER 5 V CO2/ V O2), and ventilatory equiva-
· · · ·
lents for oxygen ( V E/V O2) and carbon dioxide ( V E/ V CO2). Heart rate
was continuously monitored by electrocardiogram and oxygen satura-
tion (SpO2) determined by pulse oximeter (Nellcor Inc., Hayward,
CA). Breathing reserve was calculated from the equation: breathing
·
reserve 5 maximal voluntary ventilation (MVV) 2 peak V E. MVV
was estimated as FEV1 3 35.
Graded Maximal Aerobic Exercise Testing
This initial test was designed to establish peak oxygen uptake
· ·
(V O2max) of each subject, anaerobic threshold (AT), and peak V E, and
allow determination of the exercise load to be applied during anaerobic
exercise. After assessment of baseline cardiorespiratory measures at
rest, the test consisted of pedaling an electronically braked bicycle er-
gometer starting at zero load with increases in load of 8 watts (50 ki-
lopondmeters [kpm]/min) every 30 s until the subject was unable to
sustain a pedaling frequency of > 40 rotations per minute (rpm) for
the assigned load (19–21). The load at which the test was discontinued
·
was recorded. The highest V O2 achieved during the last 30 s of exer-
·
cise was recorded as the individual’s V O2max. AT was determined
·
from cardiorespiratory records as the point at which V CO2 increased
·
out of proportion to V O2 (22). Subjects were allowed to rest for a min-
imum of 2 h prior to beginning any anaerobic exercise testing.
Anaerobic Exercise Testing
Subjects underwent two bouts of supramaximal exercise tests at a
·
workload approximately 130% of the workload recorded at V O2max
during the graded exercise test. In each bout, baseline cardiorespira-
tory parameters were monitored while subjects sat immobile on the
ergocycle, breathing room air. When values became stable over a pe-
riod of 60 s, zero load pedaling was allowed for 1 min, after which sub-
·
jects pedaled at the designated load (130% of V O2max). Subjects were
encouraged to maintain a pedaling frequency of around 70 rpm. The
test was discontinued when the subject could no longer maintain a
pedaling frequency of > 40 rpm.
Within 2 wk of completion of this exercise protocol, subjects un-
derwent a second identical anaerobic exercise test. The average of the
two anaerobic bouts was used to determine the duration of anaerobic
exercise, anaerobic performance (anaerobic performance 5 workload
CF Control
25 6 10 (47%) 25 6 8 (41%)
1.68 6 0.1 1.70 6 0.1
55 6 14 70 6 11‡
83 6 14 110 6 15§
62 6 21 112 6 15§
31 6 24 93 6 26§
0.44 6 0.02 0.19 6 0.01§
97 6 2 99 6 1†
3,555 6 1,019 4,891 6 1,469‡
44 6 10 53 6 10†
Shah, Gozal, and Keens: Aerobic and Anaerobic Exercise in CF 1147

TABLE 2 markedly reduced. Resting SpO2 was slightly lower in subjects

COMPARISON OF PEAK EXERCISE VALUES IN SUBJECTS with CF. Clinical characteristics, pulmonary function values,
WITH CF AND CONTROL SUBJECTS AT END OF and anthropometric data are shown in Table 1.
GRADED EXERCISE TEST*
Aerobic Exercise
CF Control ·
· Subjects with CF showed significantly lower peak VE,
VE 51 6 18 93 6 31† ·
·
V o2max, ml/min/kg 24.6 6 6.0 35.5 6 8.5†
VO2max, heart rate (HR), exercise duration, and peak work-
· · loads compared with control subjects. Significant oxygen de-
V E/ V o2 43 6 8 38 6 8
·
V co2, ml/min 1,563 6 584 3,259 6 962† saturations, defined as a change . 5%, did not occur during
· ·
V E/ V co2 35 6 4 28 6 5† aerobic exercise, and mean SpO2 at exhaustion were similar.
RER 1.27 6 0.08 1.31 6 0.10 Mean end-tidal oxygen pressure (PETO2) and end-tidal carbon
·
AT, ml V o2/min/kg 13.0 6 4.5 21.5 6 6.3†
· dioxide pressure (PETCO2) at peak exercise were also ·similar.
AT/ Vo2max 53 6 14 60 6 9
Breathing reserve, L 21 6 31 28 6 38
When·
AT was expressed as a function of individual VO2max
PETO2 112 6 3 113 6 6 (% VO2max), no significant differences occurred in the two
PETCO2 37 6 3 40 6 6 study groups (Table 2).
HR 160 6 14 181 6 13†
SpO2 96 6 3 97 6 2 Anaerobic Exercise
Time exercise, min 7.6 6 2.0 12.1 6 3.4†
Peak workload, kpm/min 715 6 200 1,185 6 360†
Subjects with CF had significantly lower minute ventilation at
the end of anaerobic exercise
·
than did control subjects. In sub-
· ·
Definition of abbreviations: V E 5 minute ventilation; V o2max 5 maximal oxygen con- jects with CF peak V E at the end of anaerobic exercise was
· · · ·
sumption; V E/ V o2 5 ventilatory equivalent for oxygen; V co2 5 carbon dioxide pro-
· ·
duction; V E/ V co2 5 ventilatory equivalent for carbon dioxide; RER 5 respiratory
similar to peak VE at the end of aerobic exercise (p 5 NS). ·
In
quotient; AT 5 anaerobic threshold; PETO2 5 end-tidal oxygen tension; PETCO2 5 end- contrast, control subjects had significantly lower peak VE at
tidal carbon dioxide tension; HR 5 heart rate; SpO2 5 oxygen saturation. the end of anaerobic exercise when compared with their peak
* Values are mean 6 SD. ·
†
p , 0.001.
VE at the end of aerobic exercise. During anaerobic exercise,
neither the CF group nor the control group developed signifi-
·
cant oxygen desaturation, although SpO2 was lower in CF at
in kpm/min 3 average time of exercise in minutes), maximal V E, and the end of anaerobic exercise. Anaerobic performance was
peak heart rate. significantly lower in CF subjects compared with control sub-
jects (p , 0.005), although bout duration was similar in both
Data Analysis
study groups (p 5 NS). Comparison of peak anaerobic values
Data are presented as mean 6 SD. Two-tailed paired t tests were used are shown in Table 3.
to evaluate potential differences within each of the two groups (aero-
bic versus anaerobic in CF or controls). Unpaired t tests were used to Correlation of Pulmonary Function and
compare differences between the two groups. Unpaired t tests were Nutritional Status to Exercise
used to compare clinical characteristics, pulmonary function values,
and anthropometric data between the two groups. When multiple Potential correlations between pulmonary function and nutri-
comparisons were made, a Bonferroni adjustment to alpha was incor- tional status to aerobic and anaerobic exercise were explored
porated into the analysis.
·
(Table 4). ·
The relationships of aerobic and anaerobic variables (V O2max, Aerobic performance ( VO2max and AT) in control subjects
AT, anaerobic performance), pulmonary function values (FEV1, VC, did not exhibit significant correlations with any pulmonary
FEF25–75%, RV/TLC), and anthropometric data (AMA, LBM) were function measures. In subjects with CF, however, AT dis-
assessed by linear regression analysis and calculation of Pearson cor-
played a significant positive relationship to FEV1, suggesting
relation coefficients. In all tests, p , 0.05 was considered significant.
that pulmonary function contributes to some degree to aero-
bic performance.
RESULTS
Anaerobic
·
performance was significantly correlated to AT
Seventeen patients with CF and 17 normal subjects were stud- and VO2max in control subjects. In· subjects with CF, anaero-
ied. Most subjects with CF showed moderate obstructive dis- bic performance correlated with VO2max, but not with AT,
ease by pulmonary function testing, and all had some degree suggesting that aerobic and anaerobic exercise tolerance do
of hyperinflation. Control subjects had normal pulmonary not correlate as well in CF (Figure 1). In subjects with CF pre-
function tests. In subjects with CF, AMA and LBM were all diction· of anaerobic exercise tolerance from aerobic parame-
ters ( VO2 and AT) was not as good as in controls.
TABLE 3 Aerobic performance was not correlated to LBM or AMA
COMPARISON OF PARAMETERS AT END OF ANAEROBIC EXERCISE* in either subjects with CF or control subjects. Anaerobic per-
formance, however, correlated significantly to both LBM and
CF Control AMA in both CF and control subjects (Figure 2).
·
VE 53 6 18 80 6 25†
Breathing reserve, L 19 6 25 38 6 29 DISCUSSION
PETO2, mm Hg 107 6 5 108 6 8
PETCO2, mm Hg 40 6 5 44 6 6 In this study, we demonstrated that aerobic and anaerobic ex-
HR 153 6 10 163 6 10† ercise performance are decreased in CF when compared with
SpO2 96 6 2 98 6 1† control subjects. In control subjects, however,· there was signif-
Time, min 1.25 6 0.47 1.25 6 0.22 icant correlation between aerobic capacity ( VO2max and AT)
AnP, kpm 1,176 6 568 1,983 6 798†
and anaerobic performance. In subjects with CF anaerobic ex-
·
Definition of abbreviations: V E 5 minute ventilation; PETO2 5 end-tidal oxygen ten- ercise performance did not correlate as well to aerobic exer-
sion; PETCO2 5 end-tidal carbon dioxide tension; HR 5 heart rate; AnP 5 anaerobic per- cise performance. This may be because aerobic exercise in CF
formance.
* Average of two anaerobic bouts, peak values. Values are mean 6 SD. is limited by both cardiovascular and pulmonary mechanisms.
†
p , 0.05. Our analysis has shown muscle mass to be the factor with the
1148 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE VOL 157 1998

TABLE 4
CORRELATION VALUES

CF Control
· ·
V o2max AT AnP V o2max AT AnP

FEV1 0.22 0.65* 20.23 20.04 0.07 0.05

VC 0.42 0.67* 0.14 0.36 0.57 0.39
FEF25–75% 0.28 0.67* 20.4 0.03 0.003 20.31
RV/TLC 20.13 20.54 20.04 0.1 20.06 20.26
AMA, mm2 0.5 0.19 0.80† 0.51 0.56 0.78†
LBM, % total body weight 0.34 20.02 0.69* 0.49 0.56 0.85†
·
V o2max 0.65* 0.78†
AT 0.19 0.68*

Definition of abbreviations: FEV1 5 forced expiratory volume in 1 s; VC 5 vital capacity; FEF25–75% 5 forced expiratory flow from 25%–75%
·
of vital capacity; RV 5 residual volume; TLC 5 total lung capacity; AMA 5 arm muscle area; LBM 5 lean body mass; V o2max 5 maximal ox-
ygen consumption; AT 5 anaerobic threshold; AnP 5 anaerobic performance.
* p , 0.01.
†
p , 0.001.

highest correlation coefficient with anaerobic exercise perfor-
mance in our subjects with CF, whereas pulmonary function
did not correlate. This suggests that pulmonary function has
little or no effect on the ability to perform high intensity, short
duration exercise.
Aerobic Exercise
It is well established that in chronic lung disease, aerobic exer-
cise capacity is decreased and is limited not only by pulmonary
mechanisms, but also by poor nutritional status (23–28). The
pulmonary limitation to aerobic · exercise
·
may involve an in-
crease in PETCO2, an increase in VE/ VO2 ratio, or arterial oxy-
gen desaturation. Coates and coworkers (26) have previously
demonstrated CO2 retention associated with severe pulmo-
nary mechanical impairment and increased dead ·
space to tidal
volume ratios in subjects with CF. Also, peak VE during aero-
bic exercise in an individual with chronic lung disease is usu-
ally very close to maximal voluntary ventilation, suggesting
that these patients have little or no ventilatory reserve.
Individuals with advanced lung disease have a decreased
FEV1, which is thought to contribute
·
to their exercise limitation.
However, the decrease in VO2max in chronic lung disease may
only be partially explained by the decrease in FEV1 (23). In
adults with chronic obstructive pulmonary disease (COPD),
Baurle and Younes (23) found that decreased inspiratory ca-
pacity correlated to aerobic exercise performance. They noted
that in subjects with COPD of similar severity, · there ·
was a
variability in ventilatory equivalent for oxygen ( VE/ VO2) that
correlated with aerobic exercise limitation (23). Also, Coates
and colleagues have previously suggested that poor nutritional
status and muscle wasting may lead to limitation of exercise
tolerance in children with CF (27).

·
Figure 1. Correlation of V O2max to anaerobic performance in sub-
jects with CF (closed squares) and control subjects (open circles). In
subjects with CF linear regression correlation coefficient (r) 5 0.65
(p , 0.01). In control subjects linear regression correlation coeffi-
cient (r) 5 0.78 (p , 0.001).
Figure 2. Correlation of lean body mass to anaerobic performance
in CF subjects (closed squares) and control subjects (open circles). In
subjects with CF linear regression correlation coefficient (r) 5 0.69
(p , 0.01). In control subjects linear regression correlation coeffi-
cient (r) 5 0.85 (p , 0.001).
Shah, Gozal, and Keens: Aerobic and Anaerobic Exercise in CF
·
In this study, we found that VO2max, AT, duration of exer-
cise, peak HR, and maximum workload were all decreased in
subjects with CF compared with control subjects. Neither the
CF group nor the control group had significant oxygen desatu-
ration during aerobic exercise. Resting arterial oxygen satura-
tion was also not markedly different between the control group
and the CF group, reflecting the mild to moderate lung involve-
ment of our subjects. Subjects with more severe pulmonary in-
volvement would probably have displayed lower arterial oxy-
gen saturation at rest and during exercise (29). Indeed, Henke
and Orenstein found that subjects with CF with lower FEV1
(FEV1 , 50%VC) were more likely to have a significant (. 5%)
decrease in oxygen saturation during exercise (30).
Anaerobic Exercise
Anaerobic performance is decreased in subjects with CF when
compared with normal subjects. Peak heart rate at the end of
anaerobic exercise was also significantly lower in the CF
group than in the control group. As in aerobic exercise, no sig-
nificant oxygen desaturation was found during anaerobic ex-
ercise in either the CF group or the control group.
Little is known about anaerobic exercise in individuals with
chronic lung disease. It is difficult to analyze anaerobic
·
exer-
·
cise
·
ability,
· ·
since parameters of gas exchange (VO2, AT, VE/
VO2, VE/ VCO2) are difficult to measure and interpret during
anaerobic exercise bouts. One must rely on such parameters
as heart rate, time of exercise, minute ventilation, and work-
load in an attempt to quantitate anaerobic exercise. As men-
tioned above, only subjects with CF with mild or moderate
lung disease entered the study. Including a more severe group
might have revealed further information during correlation
analysis of anaerobic performance. However, in such subjects
the extent of disease might have had a negative impact on per-
formance variability and reproducibility.
Relationship of Pulmonary Function to Exercise
No significant relationships were identified between pulmo-
nary function and anaerobic exercise in CF or control subjects.
Cabrera and colleagues have suggested that reduced anaero-
bic performance in CF is associated with diminished pulmo-
nary function (31). Our data, however, are similar to that of
Boas and coworkers who found no correlation between pul-
monary function and anaerobic performance (9).
In healthy individuals cardiovascular, rather than pulmo-
nary, mechanisms are the limiting factor in aerobic exercise,
explaining the lack of correlation of aerobic exercise to pul-
monary function. Significant pulmonary reserve at peak aero-
bic exercise is present, such that minute ventilation during
maximal aerobic exercise rarely exceeds 60–80% of maximal
voluntary ventilation. In addition, peak minute ventilation at
the end of anaerobic exercise was lower than at peak aerobic
exercise in control subjects, suggesting that pulmonary me-
chanics do not contribute to anaerobic performance in healthy
individuals. In contrast, subjects with CF had similar peak
minute ventilation at the end of anaerobic exercise when com-
pared with peak minute ventilation at the end of aerobic exer-
cise. A significant contribution of pulmonary function to aero-
bic exercise does exist in CF, explaining the aerobic exercise
limitation. However, there does not appear to be any associa-
tion of pulmonary function with anaerobic exercise in CF sub-
jects. Anaerobic exercise in CF appears to be limited by fac-
tors other than those which limit aerobic exercise.
Relationship of Muscle Mass to Exercise
Although the total amount of anaerobic work was decreased
in the CF group, the total length of time that the CF group
1149
could perform anaerobic exercise was the same as for the con-
trol group. The ability of muscle to perform anaerobically may
be viewed as a function of two factors, the efficiency of muscle
contractions and the length of time a particular amount of
work can be sustained.
In both the CF and control groups, anaerobic performance
correlated highly to lean body mass and arm muscle area. In
analogy to the fact that poor aerobic performance in CF is
multifactorial, and not only secondary to decreased pulmo-
nary function, the decrease in anaerobic performance in CF
may not just be secondary to poor nutrition, but also to altered
muscle metabolism. This issue has been controversial in CF.
Lands and colleagues have previously reported that subjects
with CF performed similarly to control subjects during sprint
work when corrected for lean body mass (32). These investiga-
tions suggested that the nutrition-related decrease in exercise
performance in CF may reflect a loss of muscle mass, not a
loss of performance of the existing muscle (32). When com-
paring anaerobic exercise to lean body mass, Boas and co-
workers suggested that CF muscle does not perform as well as
in healthy subjects (9). Further delineation of this loss of per-
formance may lead to better understanding of muscle metabo-
lism during anaerobic exercise.
Boas and colleagues (9) used the Wingate Anaerobic Test
to study anaerobic exercise. This test involved 30-s bouts of
all-out exercise. In our study, the load for anaerobic exercise
was determined from the graded exercise test in order to bet-
ter standardize across individuals with different levels of daily
activity of physical training. Using this method, we have shown
that subjects with CF could sustain an anaerobic bout of exer-
cise for a similar length of time.
These data raise several issues regarding the performance
and energy production of skeletal muscle in CF during anaer-
obic exercise. Medbo and Tabata have demonstrated that ap-
proximately 2 min of exercise to exhaustion are required to
fully use anaerobic energy sources (33). It has also been sug-
gested, however, that aerobic energy production via oxidative
phosphorylation may provide some energy during short exer-
cise bouts in healthy muscle (34). Thus, the relative energy
contribution from aerobic and anaerobic energy sources would
vary depending on the subject (CF versus control) and type of
anaerobic exercise employed. Since anaerobic exercise is of short
duration, it would be reasonable to assume that the majority of
energy for anaerobic exercise comes from sources already stored
in resting muscle. The duration of anaerobic exercise probably
depends minimally on the ability of the cardiorespiratory sys-
tem to deliver oxygen to working muscle. Therefore, at compa-
rable intensities, subjects with CF in our study were able to
sustain pedaling for similar lengths of time as control subjects.
CONCLUSION
Decreased aerobic exercise capacity in chronic lung disease
has previously been documented. However, the study of anaer-
obic exercise has been relatively scarce in respiratory disease
and particularly in CF. Our study demonstrates significant re-
ductions in anaerobic exercise capacity in subjects with CF,
which seem to be primarily determined by muscular factors
rather than pulmonary function. We speculate that improving
nutritional status (muscle mass) will improve anaerobic exer-
cise performance in CF.
Daily activity frequently involves bursts of anaerobic exer-
cise and reduced pulmonary function may not be as detrimen-
tal to daily activities as previously thought. Subjects with CF
could sustain anaerobic exercise for similar durations as nor-
mal subjects. We postulate that the efficacy of an exercise
1150 AMERICAN JOURNAL OF RESPIRATORY AND CRITICAL CARE MEDICINE
training program in CF, directed toward improving daily activ-
ity requirements, could be substantially improved by empha-
sizing anaerobic, rather than aerobic, routines.
Acknowledgment : The authors thank the technicians of the Pulmonary Phys-
iology Laboratory at Childrens Hospital Los Angeles for technical support.
References
1. Kaplan, T. A., J. D. ZeBranek, and R. M. McKey. 1991. Use of exercise
in the management of cystic fibrosis. Pediatr. Pulmonol. 10:205–207.
2. Orenstein, D. M., B. A. Franklin, C. F. Doershuk, H. K. Hellerstein, K. J.
Germann, and J. G. Horowitz. 1981. Exercise conditioning and car-
diopulmonary fitness in cystic fibrosis. Chest 80:392–398.
3. Orenstein, D. M., P. A. Nixon, E. A. Ross, and R. M. Kaplan. 1989. The
quality of well-being in cystic fibrosis. Chest 95:344–347.
4. Keens, T. G., I. R. B. Krastins, E. M. Wannamaker, H. Levison, D. N.
Crozier, and C. A. Bryan. 1977. Ventilatory muscle endurance train-
ing in normal subjects and patients with cystic fibrosis. Am. Rev. Re-
spir. Dis. 116:853–860.
5. Sawyer, E. H., and T. L. Clanton. 1993. Improved pulmonary function
and exercise tolerance with inspiratory muscle conditioning in chil-
dren with cystic fibrosis. Chest 104:1490–1497.
6. Strauss, D. G., A. Osher, C. Wang, E. Goodrich, F. Gold, W. Colman,
M. Stabile, A. Dobrenchuk, and T. G. Keens. 1987. Variable weight
training in cystic fibrosis. Chest 92:273–276.
7. Nixon, P. A., D. M. Orenstein, S. F. Kelsey, and C. F. Doershuk. 1992.
The prognostic value of exercise testing in patients with cystic fibrosis.
N. Engl. J. Med. 327:1785–1788.
8. Simpson, K., K. Killian, N. MacCartney, D. G. Stubbing, and N. L.
Jones. 1992. Randomized controlled trial of weightlifting exercise in
patients with chronic airflow limitation. Thorax 47:70–75.
9. Boas, S. R., M. L. Joswiak, P. A. Nixon, J. A. Fulton, and D. M. Oren-
stein. 1996. Factors limiting anaerobic performance in adolescent
males with cystic fibrosis. Med. Sci. Sports Exerc. 28:291–298.
10. DuBois, A. B., S. W. Botello, and J. H. Comroe. 1956. A new method for
measuring airways resistance in man using body plethysmography:
values in normal subjects and in patients with respiratory disease. J.
Clin. Invest. 35:327–335.
11. Platzker, A. C. G., and T. G. Keens. 1985. Pulmonary function testing in
pediatric patients. In A. F. Wilson, editor. Pulmonary Function Test-
ing: Indications and Interpretations. Grune and Stratton, New York.
275–292.
12. Goldman, H. I., and M. R. Becklake. 1959. Respiratory function tests.
Am. Rev. Tuberc. 79:457.
13. Lindal, A., A. Medina, and J. T. Grismer. 1967. A re-evaluation of nor-
mal pulmonary function measurements in the adult. Am. Rev. Respir.
Dis. 95:1061.
14. Cherniack, R. M., and M. B. Raber. 1972. Normal standards for ventila-
tory function using an automated wedge spirometer. Am. Rev. Respir.
Dis. 106:38–46.
15. Brook, C. G. D. 1971. Determination of body composition of children
from skinfold measurements. Arch. Dis. Child. 46:182–184.
16. Hammer, L. D., H. C. Kraemer, D. M. Wilson, P. L. Ritter, and S. M.
Dornbusch. 1991. Standardized percentile curves of body mass index
for children and adolescents. Am. J. Dis. Child. 145:259–263.
VOL 157 1998
17. Leiko, N. S., C. Stawskic, and K. Benkov. 1987. The nutritional assess-
ment of the pediatric patient. In R. Grand, J. L. Sutphen, and W. H.
Dietz, editors. Pediatric Nutrition: Therapy and Practice. Butter-
worth, Newton, MA. 404–407.
18. Slaughter, M. H., T. G. Lohman, R. A. Boileau, C. A. Horswill, R. J.
Tillman, M. D. Van Loan, and D. A. Bemben. 1988. Skinfold equa-
tions for estimation of body fatness in children and youth. Human Bi-
ology 60:709–723.
19. Godfrey, S. 1974. Exercise Testing in Children. W. B. Saunders, Phila-
delphia. 30.
20. Jones, N. L. 1988. Clinical Exercise Testing. W. B. Saunders, Philadel-
phia.
21. Nixon, P. A., and D. M. Orenstein. 1988. Exercise testing in children. Pe-
diatr. Pulmonol. 5:107–122.
22. Wasserman, K., J. E. Hansen, D. Y. Sue, B. J. Whipp, and R. Casaburi.
1994. Principles of Exercise Testing and Interpretation. Lea & Fe-
biger, Philadelphia. 62–64.
23. Bauerle, O., and M. Younes. 1995. Role of ventilatory response to exer-
cise in determining exercise capacity in COPD. J. Appl. Physiol. 79:
1870–1877.
24. Marcotte, J. E., R. K. Grisdale, H. Levison, A. L. Coates, and G. J.
Canny. 1986. Multiple factors limit exercise capacity in cystic fibrosis.
Pediatr. Pulmonol. 2:274–281.
25. Orenstein, D. M. 1988. Exercise tolerance and exercise conditioning in
children with chronic lung disease. J. Pediatr. 112:1043–1047.
26. Coates, A. L., G. Canny, R. Zinman, R. Grisdale, K. Desmond, D. Rou-
meliotis, and H. Levison. 1988. The effects of chronic airflow limita-
tion, increased dead space, and the pattern of ventilation on gas ex-
change during maximal exercise in advanced cystic fibrosis. Am. Rev.
Respir. Dis. 138:1524–1531.
27. Coates, A. L., P. Boyce, D. Muller, M. Mearns, and S. Godfrey. 1980.
The role of nutritional status, airway obstruction, hypoxia, and abnor-
malities in serum lipid composition in limiting exercise tolerance in
children with cystic fibrosis. Acta Paediatr. Scand. 69:353–358.
28. Marcotte, J. E., G. J. Canny, R. Grisdale, K. Desmond, M. Corey, R.
Zinman, H. Levinson, and A. L. Coates. 1986. Effects of nutritional
status on exercise performance in advanced cystic fibrosis. Chest 90:
375–379.
29. Marcus, C. L., D. Bader, M. W. Stabile, C. Wang, A. B. Osher, and T. G.
Keens. 1992. Supplemental oxygen and exercise performance in pa-
tients with cystic fibrosis with severe pulmonary disease. Chest 101:
52–57.
30. Henke, K. G., and D. M. Orenstein. 1984. Oxygen saturation during ex-
ercise in cystic fibrosis. Am. Rev. Respir. Dis. 129:708–711.
31. Cabrera, M. E., M. D. Lough, C. F. Doershuk, and G. A. Derivera. 1993.
Anaerobic performance assessed by the Wingate test in patients with
cystic fibrosis. Pediatr. Exerc. Sci. 5:78–87.
32. Lands, L. C., J. F. Heigenhauser, and N. L. Jones. 1993. Analysis of fac-
tors limiting maximal exercise performance in cystic fibrosis. Clin. Sci.
83:391–397.
33. Medbo, J. I., and I. Tabata. 1993. Anaerobic energy release in working
muscle during 30s to 3 minutes of exhausting bicycling. J. Appl. Phys-
iol. 75:1654–1660.
34. Medbo, J. I., and I. Tabata. 1989. Relative importance of aerobic and
anaerobic energy release during short-lasting exhausting bicycle exer-
cise. J. Appl. Physiol. 67:1881–1886.