Content uploaded by Sheng Yan
Author content
All content in this area was uploaded by Sheng Yan
Content may be subject to copyright.
Eur Respir J 1998; 11: 416–421
DOI: 10.1183/09031936.98.11020416
Printed in UK - all rights reserved
Copyright ©ERS Journals Ltd 1998
European Respiratory Journal
ISSN 0903 - 1936
Partitioning of the elastic work of inspiration in
patients with COPD during exercise
P. Sliwinski*, D. Kaminski*, J. Zielinski*, S. Yan+
aa
During exercise, the elastic work of inspiration (Wi) in
healthy subjects increases mainly due to an increase in
tidal volume (VT). This is partly compensated for by the
fall in end-expiratory lung volume (EELV) as a result of
exercise-induced recruitment of expiratory muscles [1–3].
In patients with chronic obstructive pulmonary disease
(COPD), the increase in Wi is more than would be pre-
dicted from the increase in VT during exercise [4], and
contributes significantly to the sense of breathing effort
and reduced exercise capacity. Two additional factors
play a crucial role in increasing Wi in these patients. Firs-
tly, due to expiratory airflow limitation, dynamic hyper-
inflation is almost inevitable during exercise, leading to
an increased EELV and intrinsic positive end-expiratory
pressure (PEEPi) [5–8] in spite of expiratory muscle ac-
tivity. PEEPi behaves as an inspiratory threshold load and
needs to be overcome before inspiratory flow can be gen-
erated [9], therefore, it significantly increases Wi for a
given ventilation. Secondly, the increased frequency of
breathing during exercise results in a decreased dynamic
lung compliance (CL,dyn) in patients with COPD [10, 11],
which also contributes to the increase in Wi under dyna-
mic conditions.
How PEEPi and changes in CL,dyn in patients with
COPD quantitatively contribute to an increase in Wi dur-
ing exercise has not been described in detail. The purposes
of the present study were, therefore, to quantify, in a group
of clinically stable patients with COPD, the increase in Wi
during exercise, and to partition it into the contributions
from the increased PEEPi and from the decreased CL,dyn,
in order to better understand inspiratory mechanics during
exercise in this patient population.
Methods
Patients
A total of 10 patients with severe COPD (nine males
and one female) in clinically stable conditions voluntee-
red for the experiment. Clinically stable COPD is defined
as having the medical history and signs of chronic air-
way obstruction but without acute exacerbation within 1
month prior to the study. Patients with evidence of car-
diovascular disease, systemic disease and neuromuscular
disease which are likely to affect exercise performance
Partitioning of the elastic work of inspiration in patients with COPD during exercise. P.
Sliwinski, D. Kaminski, J. Zielinski, S. Yan. ©ERS Journals Ltd 1998.
ABSTRACT: During exercise, dynamic hyperinflation-induced intrinsic positive end-
expiratory pressure (PEEPi) and decreased dynamic lung compliance (CL,dyn) of pa-
tients with chronic obstructive pulmonary disease (COPD) increase the elastic work
of inspiration (Wi) more than would be predicted from the increase in tidal volume
(VT). This contributes significantly to their exertional breathlessness.
In 10 stable patients with COPD, the dynamic Wi was measured during incremen-
tal bicycle exercise to exhaustion. The total Wi was then partitioned into the portion
required to overcome PEEPi (Wi,PEEPi) and nonPEEPi elastic load (Wi,nonPEEPi).
The latter is used to overcome the increase in the total respiratory system elastance
during inflation.
From resting breathing to peak exercise, Wi more than doubled (p<0.001). This
increase was largely due to Wi,PEEPi, which significantly rose from 1.7±0.3 to 5.3±
0.8 L·cmH2O-1 (p<0.001). In comparison, Wi,nonPEEPi increased from only 3.0±0.4 to
5.1±0.5 L·cmH2O-1 (p<0.01). Consequently, Wi,PEEPi as a fraction of total Wi incre-
ased from 35.5±5.6 to 51.0±3.3% (p<0.02). In addition, the measured Wi,nonPEEPi at
peak exercise, when expressed as a percentage of its value during resting breath-
ing, was 25% more than that predicted from the increase in VT alone. Assuming a con-
stant chest wall compliance, this can be attributed to the exercise-induced decrease in
CL,dyn, which was 0.27±0.04 and 0.17±0.02 L·cmH2O-1 (p<0.01), respectively, during
resting breathing and peak exercise.
In conclusion, the dynamic hyperinflation-induced intrinsic positive end-expira-
tory pressure is more important than the increase in tidal volume in raising the work
of inspiration during exercise in patients with chronic obstructive pulmonary disease;
the decrease in dynamic lung compliance plays a definite but less important role.
Eur Respir J 1998; 11: 416–421.
*Institute of Tuberculosis and Lung Dis-
eases, Warsaw, Poland. +Meakins-Christie
Laboratories, McGill University and Mon-
treal Chest Institute, Royal Victoria Hospi-
tal, Montreal, Canada
Correspondence: S. Yan
Meakins-Christie Laboratories
McGill University and Montreal Chest
Institute
3626 St. Urbain Street
Montreal
Quebec
Canada H2X 2P2
Fax: 1 514 398 7483
Keywords: Dynamic hyperinflation
dynamic lung compliance
elastic work of breathing
intrinsic positive end-expiratory pressure
Received: April 7 1997
Accepted after revision December 1 1997
The study was supported by the Polish
State Research Committee and Montreal
Chest Institute
ELASTIC WORK OF BREATHING IN COPD PATIENTS 417
were excluded. Patient characteristics and lung function
test results are given in table 1. All patients gave their in-
formed consent to the protocol which had been explain-
ed to them in detail necessary for their understanding. The
experimental protocol was approved by an appropriate in-
stitutional research ethics committee.
Measurements
Respiratory flow was measured by a Fleisch No. 2 pneu-
motachograph (Fleisch, Lausanne, Switzerland) and a dif-
ferential pressure transducer. Mouth pressure (Pm) was
measured by a differential pressure transducer which was
connected to the mouthpiece. Oesophageal and gastric pres-
sures (Poes and Pga, respectively) were measured conven-
tionally by two additional differential pressure transducers
attached to balloon-tipped catheters. One was placed in the
lower third area of the oesophagus and the other in the
stomach, to reflect the changes in pleural and abdomin-
al pressures (Ppl and Pab), respectively. The oesophageal
balloon was inflated with 0.5 mL of air and its position
was carefully adjusted so that the amplitudes of Poes and Pm
were equal during an airway occlusion test and car- dio-
genic artifacts were minimized. The gastric balloon was
inflated with 1.0 mL of air. End-tidal carbon dioxide ten-
sion (PET,CO2) was measured using a CO2 analyser whose
inlet was inserted into the mouthpiece. During exercise,
arterial oxygen saturation (Sa,O2) and electrocardiogram
(ECG) measurements were continuously monitored. Arte-
rial blood pressure was measured at the end of each work-
load. Mot-ions of the rib cage and abdomen during
breathing were measured by two respitrace bands placed on
the upper rib cage and the abdomen. These signals were
used for estima-ting expiratory muscle contribution to Poes
baseline during quiet breathing [12, 13].
Protocol
The patients were seated on an electronically braked er-
gometer (Ergo-Metrics 800; Ergo-Line GmbH + Co KG,
Bitz, Germany) with their body tilted slightly forward and
their arms supported by the handle of the ergometer in
order to maintain an unchanged posture throughout the
experiment. The patients breathed through a mouthpiece
while wearing a noseclip. No patient took long-lasting
bronchodilators and all refrained from taking coffee or
bronchodilators for at least 4 h before the experiment. At
the beginning of the experiment, patients were instruct-
ed to breathe quietly on the bicycle without pedalling for
~5 min to become accustomed to the breathing circuit. In
the last 2 min of this adaptation, quiet breathing data were
recorded. The patients then performed an incremental bi-
cycle exercise starting at a workload of 10 or 20 W. The
workload was progressively increased every minute by 10
or 20 W until symptom limitation. At the end of quiet bre-
athing and the end of each level of exercise, an inspiratory
capacity (IC) inspiration was performed by "making a fur-
ther maximal effort on top of a maximal inspiration" [6,
14].
Data analysis
The respiratory flow, Poes, Pga, respitrace signals, Sa,O2,
and PET,CO2 were all amplified, digitized at 100 Hz, and
stored in a computer. The actual sampling frequency
was sometimes increased up to 1,000 Hz when necessary
by linear extrapolation, in order to analyse fast events,
such as the development of Poes during the initial part
of inspiration. The signals for the last five breaths at a
given workload were averaged for further analysis. VT,
breathing frequency, minute ventilation (V'E), and duty
cycle were derived from the flow signal. Change in EELV
(∆EELV) was measured by calculating the magnitude of
IC, assuming the decrease in IC equals the increase in
EELV during exercise [5–8]. The change in end-inspira-
tory lung volume (∆EILV) was calculated as ∆EELV + VT.
CL,dyn was measured by dividing VT by the difference in
Poes values at the beginning and end of inspiratory flow.
Figure 1 shows schematically the method used to calcu-
late and partition the elastic work of inspiration. In this
figure, the static volume-pressure relationship of the chest
wall (the straight line passing points F, E, D) is shown. It
was assumed that this relationship is linear [15] and its
slope in patients with COPD is comparable to that ob-
tained in healthy subjects [16–18]. Thus, the slope of
Table 1. – Patient characteristics and lung function test results
Pt
No. Age
yr W
kg H
cm FEV1
LFEV1/FVC
%Raw
cmH2O·L-1·s-1 TLC
LFRC
LRV
LPa,O2
kPa Pa,CO2
kPa
1
2
3
4
5
6
7
8
9
10
Mean
SEM
58
62
67
64
57
56
41
47
62
59
57
8
76
68
68
78
87
55
54
65
83
70
70
3
160
168
164
180
169
158
169
170
172
170
168
2
0.76 (33.3)
0.92 (31.0)
0.44 (16.0)
1.32 (38.6)
1.10 (34.0)
0.84 (37.1)
0.80 (32.0)
0.88 (25.6)
0.96 (30.9)
0.68 (21.9)
0.87 (30.0)
0.07 (2.2)
0.42
0.37
0.25
0.46
0.31
0.35
0.30
0.29
0.25
0.25
0.32
0.02
10.1
12.8
7.2
4.6
3.6
7.0
7.7
8.6
7.7
9.4
7.9
0.8
4.37 (90.4)
8.50 (134.0)
7.89 (131.0)
5.60 (76.7)
5.12 (73.3)
6.41 (138.1)
9.22 (143.6)
7.80 (121.4)
8.40 (126.1)
8.87 (136.4)
7.22 (117.1)
0.54 (8.4)
3.66 (137.4)
6.49 (191.4)
6.72 (200.1)
4.27 (115.7)
2.25 (67.8)
4.45 (171.4)
7.86 (242.4)
6.30 (192.1)
6.24 (178.5)
7.33 (214.4)
5.56 (171.1)
0.56 (16.2)
2.72 (146.6)
5.56 (240.2)
5.40 (223.7)
2.69 (106.9)
1.55 (85.8)
3.69 (211.8)
6.65 (348.7)
5.08 (254.7)
4.71 (197.3)
6.23 (271.6)
4.43 (208.7)
0.53 (25.0)
10.2
8.0
10.1
7.7
10.8
8.2
8.2
7.2
9.0
7.7
8.8
0.4
5.9
6.7
5.9
6.9
5.6
5.6
5.6
5.9
5.9
6.3
6.0
0.1
Values are presented as absolute values with percentages of predicted values in parenthesis. Pt: patient; W: body weight; H: height;
FEV1: forced expiratory volume in one second; FVC: forced vital capacity; Raw: airway resistance; TLC: total lung capacity; FRC:
functional residual capacity; RV: residual volume. Pa,O2: arterial oxygen tension; Pa,CO2: arterial carbon dioxide tension. Lung func-
tions were measured in a body plethysmograph. Patients No. 4 and 5 also have restrictive pulmonary changes. 1 kPa=7.52 mmHg.
418 P. SLIWINSKI ET AL.
the static volume-pressure relationship of the chest wall
was obtained from the literature, taking the age and sex
of each patient into consideration [19]. In order to posi-
tion the static volume-pressure relationship of the chest
wall properly, we first determined the end-expiratory Poes
value during quiet breathing, to represent the chest wall
elastic recoil pressure, shown at point F in figure 1. Then,
we positioned the static volume-pressure relationship of
the chest wall by passing it through point F. It needs to be
pointed out that: firstly, since most of our patients with
COPD had dynamic hyperinflation and PEEPi during qu-
iet breathing, Poes started to fall before inspiratory flow
began [20, 21]. End-expiration, therefore, was defined as
the point at which Poes began to fall for the next inspira-
tory effort rather than the beginning of inspiratory flow;
secondly, due to dynamic hyperinflation during quiet
breathing, the volume at point F was often already above
the relaxation volume (Vr); and thirdly, expiratory muscles
were active at end-expiration during quiet breathing in
some patients with COPD, resulting in elevation of end-
expiratory Poes independent of chest wall elasticity [13].
This was confirmed by an expiratory rise in Pga converted
to an abrupt inspiratory fall with outward displacement of
the abdomen [13]. To circumvent this problem, the Poes
value at point F was carefully corrected by subtracting the
expiratory rise in Pga whenever applicable, as recently
suggested by LESSARD et al. [22]. Figure 2 shows the tra-
cings of expiratory muscle recruitment during quiet bre-
athing and defines the method used to correct the effect of
this phenomenon on end-expiratory Poes estimation. Five
of our patients displayed the pattern shown in figure 2
during quiet breathing. The mean correction for these
patients was 2.1 cmH2O (range: 0.8–3.6 cmH2O).
As shown in figure 1, a clockwise dynamic volume-
pressure breathing loop during exercise is superimposed on
the static volume-pressure relationship of the chest wall.
The EELV of this breath, as given by the level of A and E,
is above the volume at point F, suggesting dynamic hyper-
inflation or further dynamic hyperinflation. Points A and B
represent the beginning and end of inspiratory flow (zero
flow points), respectively. If we assume that inspiratory
effort begins from point E, the pressure difference between
points E and A represents dynamic PEEPi [23]. The pres-
sure difference between points D and B represents the total
inspiratory pressure needed for the whole inspiration. The
straight line passing points A and C (dashed line) is made
parallel to the static volume-pressure relationship of the
chest wall. Consequently, the total Wi is given by the area
enclosed by points A, B, D and E. However, Wi can be
partitioned into the portion required to overcome PEEPi
(Wi,PEEPi) as expressed by the area enclosed by points A,
C, D and E, and the portion required to overcome the non-
PEEPi elastic load (Wi,nonPEEPi) as given by the area en-
closed by points A, B and C. The latter is used to overcome
the increase in the total respiratory system elastance during
inflation.
The Wi,nonPEEPi measured as described above was fur-
ther compared with the calculated Wi,nonPEEPi predicted
from the increase in VT alone during exercise [24]. As
shown in figure 1, Wi,nonPEEPi is given by the triangle
ABC and can be expressed as 1/2·VT·BC. Since BC repre-
sents the increase in total respiratory system elastic recoil
–+
Oesophageal pressure
+
Lung volume L
BCD
E
F
A
Fig. 1. – Diagrammatic presentation of the partitioning of the elastic
work of inspiration during dynamic hyperinflation. Please see text fo
r
further explanation.
1
0
-1
0
-5
-10
-15
25
20
15
10 43210
Pga cmH2OPoes cmH2OFlow L·s-1
Time s
Fig. 2. – Example of flow, oesophageal and gastric pressure (Poes and
Pga, respectively) tracings with expiratory muscle recruitment during
quiet breathing. Pga is biphasic with a large rise in inspiration and a
small rise in expiration. The latter suggests active expiration. The
amount of the expiratory increase in Pga (as shown between the arrows)
is subtracted from the Poes baseline, to correct the effect of expiratory
muscle contribution.
ELASTIC WORK OF BREATHING IN COPD PATIENTS 419
pressure during inspiratory inflation, it can be expressed
by VT/Crs where Crs is the compliance of the respiratory
system. Thus, assuming linearity, Wi,nonPEEPi can be cal-
culated as 1/2·VT·VT/Crs or VT2/2·Crs. If Crs is constant, the
increase in Wi,nonPEEPi during exercise will be proportio-
nal to the increase in VT2. Any discrepancy between the
measured Wi,nonPEEPi and the Wi,nonPEEPi predicted from
the increase in VT alone will therefore indicate the contri-
bution of changes in Crs to Wi,nonPEEPi.
Results are presented as mean±SEM. Analysis of variance
and paired t-test were used, to compare the results under
different experimental conditions. A p-value of <0.05 was
considered to be statistically significant.
Results
At the symptom limitation, the exercise workload ac-
complished by the patients was 28±2.5% pred. At this le-
vel of exercise, V'E was 27.9±1.8 L·min-1.
Figure 3 shows the relationship between lung volume
(VL) and Poes at the beginning (open circles) and end of
inspiratory flow (closed circles). Therefore, the level of
VL corresponding to each open and closed circle repre-
sents EELV and EILV, respectively for a given condition.
Values were obtained during quiet breathing and at differ-
ent intensities of exercise. In this figure, zero VL repre-
sents the level of EELV during quiet breathing. During
exercise, all patients demonstrated increased EELV sug-
gesting dynamic hyperinflation. Most patients reached the
highest EELV and EILV at maximal exercise. From quiet
breathing to peak exercise, EELV increased by 0.35±0.07
L. This resulted in an increase in dynamic PEEPi from
2.5±0.4–6.0±0.7 cmH2O (p<0.0001). As shown in figure
3, the curvilinear aspect of the dynamic VL-Poes relation-
ship at zero flow likely reflected the combined effects of
increasing curvature of the volume-pressure curve of the
lung itself due to increasing volume and decreasing CL,dyn
during exercise. From quiet breathing to peak exercise,
CL,dyn significantly decreased from 0.27±0.04–0.17±0.02
L·cmH2O-1 (p<0.01).
Figure 4 shows Wi, Wi,PEEPi, and Wi,nonPEEPi during
quiet breathing and at the first, the middle, and the peak
exercise intensity (E1, E2 and E3, respectively). Although
all three parameters progressively increased during exer-
cise, the rates of increase in Wi,PEEPi, and Wi,nonPEEPi dif-
fered. During quiet breathing, Wi,PEEPi, and Wi,nonPEEPi
were 1.68±0.28 and 3.02±0.43 L·cmH2O-1, respectively,
with the former significantly less than the latter (p<0.05).
During exercise, Wi,PEEPi increased more than Wi,nonPEEPi,
so that at peak exercise, their values reach 5.34±0.78 and
5.10±0.47 L·cmH2O-1, respectively, and were no longer sig-
nificantly different (p>0.75). Consequently, when Wi,PEEPi
is expressed as a fraction of Wi, the ratio progressively in-
creased with increasing exercise intensities from 35.5±
5.6% during quiet breathing to 51.0±3.3% at peak exercise
(p<0.05).
Figure 5 plots the measured Wi,nonPEEPi as a function of
the predicted Wi,nonPEEPi calculated from the increase in
VT alone during exercise. Both are expressed as a percen-
tage of the value obtained during quiet breathing. The
measured Wi,nonPEEPi was slightly but consistently higher
than the predicted Wi,nonPEEPi. At peak exercise (E3), the
discrepancy was 25%.
2.0
1.5
1.0
0.5
0.0
-0.5
V
L
L
2.0
1.5
1.0
0.5
0.0
-0.5
V
L
L
2.0
1.5
1.0
0.5
0.0
-0.5
V
L
L
2.0
1.5
1.0
0.5
0.0
-0.5
V
L
L
2.0
1.5
1.0
0.5
0.0
-0.5
V
L
L
-15 -10 -5 05
Poes cmH2O
-15 -10 -5 05
Fig. 3. – The relationship between lung volume (VL) and oesophageal
pressure (Poes) at the beginning ( ) and end (●) of inspiratory flow
during quiet breathing and at different intensities of exercise in individ-
ual patients. Zero VL represents the end-expiratory lung volume during
quiet breathing.
420 P. SLIWINSKI ET AL.
Discussion
We measured the dynamic Wi in patients with severe
COPD during incremental bicycle exercise, until symp-
tom limitation. We partitioned the total Wi into Wi,PEEPi
and Wi,nonPEEPi which corresponded to the work per-
formed to overcome the PEEPi and nonPEEPi elastic load
of the respiratory system, respectively. We found that: 1)
Wi,PEEPi became an increasingly important contributor to
Wi as exercise proceeded to the maximum. At exhaustion,
Wi,PEEPi accounted for over 50% of total Wi in these pa-
tients; and 2) at peak exercise, the measured Wi,nonPEEPi
was 25% greater than the value predicted from the in-
crease in VT alone. This can be explained by a significant
decrease in CL,dyn associated with exercise.
We used the method described by FLEURY et al. [17] to
calculate Wi while taking into consideration of ∆EELV
and the PEEPi-imposed inspiratory threshold load [25].
We used the end-expiratory Poes baseline to represent the
end-expiratory elastic recoil pressure of the chest wall
during quiet breathing, and positioned the predicted static
pressure-volume relationship of the chest wall at that
point. This procedure has been used by many previous re-
searchers [17, 25, 26]. In addition, we used the method
described by LESSARD et al. [22] to correct the effect of
expiratory muscle recruitment on determination of end-
expiratory chest wall elastic recoil pressure during quiet
breathing. It can be seen in figure 1 that once the static
chest wall volume-pressure relationship was positioned,
all subsequent measurements were referenced to this rela-
tionship and additional adjustments for expiratory muscle
contribution were not necessary during exercise. However,
there are still a number of concerns that need to be
addressed. Firstly, end-expiratory Poes could have chang-
ed due to changes in flow resistance independent of
inspiratory muscle contribution, which would have influ-
enced the reliability of point F determination in figure 1.
After carefully checking our signals, we found that at
end-expiration, the fall in Poes almost always coincided
with the rise in transdiaphragmatic pressure for the subse-
quent in-spiratory effort during quiet breathing, suggest-
ing that the resistive pressure loss at end-expiration would
be reasonably small. This is consistent with recent find-
ings in pati-ents with COPD that during spontaneous
breathing, the rapid decrease in Poes at end-expiration was
in phase with the start of the electric activities of both the
diaphragm and accessory muscles [22]. Secondly, since
we used the predicted volume-pressure relationship of the
chest wall as reference of the measurements for each indi-
vidual, a certain degree of random error was unavoidable
due to intersubject variability. We hope that this would be
minimized by only reporting and interpreting the group
results. Thirdly, it is obvious that the effect of any error of
the slope or position of the chest wall volume-pressure
curve on calculation of the Wi will be amplified by
increasing VT (fig. 1). However, VT only increased from
0.72 L during quiet breathing to 0.88 L at maximal exer-
cise. This makes any error of Wi or its components to be
relatively constant and independent of exercise, and there-
fore less likely to have significantly influenced the major
interpretations of our results based on comparisons
between rest and at different levels of exercise.
As has been shown previously [6], in spite of severe ex-
piratory airflow limitation, the increase in EELV in pati-
ents with COPD during exercise was modest. At symptom
limitation, we found an average increase in EELV of 0.35
L. This, with the small increase in VT of 0.16 L, however,
leads to a reduction of the inspiratory reserve volume by
0.51 L or 54%. Meanwhile, Wi,PEEPi more than tripled
between quiet breathing and peak exercise and accounted
for more than 50% of Wi at exhaustion, while Wi,nonPEEPi
only increased moderately (fig. 4). These results suggest
that, in patients with COPD, the increase in Wi during
exercise comes mainly from the increase in the magnitude
of dynamic hyperinflation. Accordingly, our present work
has added evidence to support the notion that dynamic
hyperinflation is probably the most important factor con-
tributing to exercise limitation in patients with COPD. In
fact, even during quiet breathing, the effect of dynamic
hyperinflation could be more significant than ever thou-
ght. We observed a dynamic PEEPi of 2.5 cmH2O in our
patients during quiet breathing, a magnitude consistent
with other studies [20, 21]. However, 36% of Wi was re-
quired to overcome this "mild" PEEPi. Since respiratory
controllers are able to detect very small increases in elas-
tic load and respond to it to adjust ventilatory output [27],
we would postulate that the PEEPi we observed at rest
12
10
8
6
4
2
0QB E1 E2 E3
Exercise intensity
W
i
L·cmH
2
O
-1
Fig. 4. – Elastic work of inspiration (Wi) during exercise. ●: total Wi; :
the work required to overcome intrinsic positive end-expiratory pressure
(PEEPi) (Wi,PEEPi); ◆ : the work required to overcome nonPEEPi elas-
tic load (Wi,nonPEEPi). QB: quiet breathing; E1, E2 and E3: first, middle
and peak exercise intensity.
100
150
200
250
100 150 200 250
Predicted W
i,nonPEEPi
% QB
Measured W
i,nonPEEPi
% QB
*
*
Fig. 5. – The measured elastic work of inspiration (Wi) required to
overcome the nonintrinsic positive end-expiratory pressure (nonPEEPi)
elastic load (Wi,nonPEEPi) plotted against the predicted Wi,nonPEEPi.
Both are expressed as percentage of quiet breathing (QB). – – – : the
line of identity. : QB; ■, and ▲ : first, middle and peak exercise
intensity (E1, E2 and E3, respectively). *: measured Wi,nonPEEPi sig-
nificantly greater than predicted.
ELASTIC WORK OF BREATHING IN COPD PATIENTS 421
might be sufficient to affect the control of spontaneous
quiet breathing in patients with stable COPD.
In this study, we assumed that dynamic compliance of
the chest wall (Cw,dyn ) during exercise is constant and
equal to its static compliance. This assumption may not be
true. Our calculation of Wi,nonPEEPi did not include influ-
ences from a possible change in Cw,dyn [28], because we
could not measure it in the present study. In other words,
our results, showing that part of the increase in Wi,nonPEEPi
during exercise cannot be explained by the increase in VT
alone (fig. 5), were entirely attributable to changes in lung
mechanics. As has been described previously [10], we
sho-wed a significant decrease in CL,dyn in our patients
during exercise. This may be largely due to a significant
frequ-ency dependence of CL,dyn in patients with COPD
[11] and, probably to a less extent, due to a decrease in the
static lung compliance (CL,stat) at increased operating lung
volume during dynamic hyperinflation. To what extent the
latter contributed to the observed decrease in CL,dyn in our
patients is unknown. MACKLEM and BECKLAKE [14] and STUBBING
et al. [10] have shown that in patients with airway obstruc-
tion, the static pressure-volume relationship of the lung
(CL,stat) was linear over a large part of the entire inspira-
tory capacity because of the loss of lung elastic recoil.
Therefore, the contribution of the fall in CL,stat to the
observed decrease in CL,dyn under our experimental condi-
tions would have been small. Since the predicted
Wi,nonPEEPi was calculated by assuming a constant respi-
ratory system compliance, and the measured Wi,nonPEEPi
was calculated by assuming a constant chest wall compli-
ance, the discrepancy between the measured and predicted
Wi,nonPEEPi (fig. 5) was thus the direct consequence of the
observed decrease in CL,dyn during exercise.
In summary, in patients with chronic obstructive pul-
monary disease exercising to symptom limitation, dyna-
mic hyperinflation is a major factor contributing to the
increase in the elastic work of inspiration, while increas-
ing tidal volume and decreasing the dynamic compliance
of the lung during exercise plays a less important role.
Therefore, without appropriate intervention for dynamic
hyperinflation, the exercise limitation of these patients can
hardly be improved.
References
1. Younes M, Kivinen G. Respiratory mechanics and bre-
athing pattern during and following maximal exercise. J
Appl Physiol 1984; 57: 1773–1782.
2. Cha EJ, Sedlock D, Yamashiro SM. Changes in lung vol-
ume and breathing pattern during exercise and CO2 inha-
lation in humans. J Appl Physiol 1987; 62: 1544–1550.
3. Henke KG, Sharratt M, Pegelow D, Dempsey JA. Regula-
tion of end-expiratory lung volume during exercise. J
Appl Physiol 1988; 64: 135–146.
4. Younes M. Determinants of thoracic excursions during
exercise. In: Whipp BJ, Wasserman K, eds. Exercise: Pul-
monary Physiology and Pathophysiology (Lung Biology
in Health and Disease, Vol. 52), 1st ed., New York, Mar-
cel Dekker, Inc., 1991; pp. 1–65.
5. Dodd DS, Brancatisano T, Engel LA. Chest wall mechan-
ics during exercise in patients with severe chronic airflow
obstruction. Am Rev Respir Dis 1984; 129: 33–38.
6. O'Donnell DE, Webb KA. Exertional breathlessness in
patients with chronic airflow limitation. Am Rev Respir
Dis 1993; 148: 1351–1357.
7. Belman MJ, Botnick WC, Shin JM. Inhaled bronchodila-
tors reduce dynamic hyperinflation during exercise in
patients with chronic obstructive pulmonary disease. Am
J Respir Crit Care Med 1996; 153: 967–975.
8. O'Donnell DE, Bertley JC, Chau LKL, Webb KA. Quali-
tative aspects of exertional breathlessness in chronic airflow
limitation: pathophysiologic mechanisms. Am J Respir Crit
Care Med 1997; 155: 109–115.
9. Rossi A, Polese G, Brandi G. Dynamic hyperinflation. In:
Marini JJ, Roussos C, eds: Ventilatory Failure. Vol. 15,
Berlin, Springer-Verlag, 1991; pp. 199–218.
10. Stubbing DG, Pengelly LD, Morse JC, Jones NL. Pulmo-
nary mechanics during exercise in subjects with chronic
airflow obstruction. J Appl Physiol 1980; 49: 511–515.
11. Pride NB, Milic-Emili J. Lung mechanics. In: Calverly P,
Pride N, eds. Chronic Obstructive Pulmonary Disease. 1st
ed. London, Chapman & Hall, 1995; pp. 135–160.
12. Ninane V, Rypens F, Yernault JC, De Troyer A. Abdomi-
nal muscle use during breathing in patients with chronic
airflow obstruction. Am Rev Respir Dis 1992; 146: 16–
21.
13. Ninane V, Yernault JC, De Troyer A. Intrinsic PEEP in
patients with chronic obstructive pulmonary disease. Am
Rev Respir Dis 1993; 148: 1037–1042.
14. Macklem PT, Becklake MR. The relationship between
the mechanical and diffusing properties of the lung in
health and disease. Am Rev Respir Dis 1963; 87: 47–56.
15. Campbell EJM. The respiratory muscles and the mechan-
ics of breathing. London, Lloyd-Luke, 1958.
16. Sharp JT, Van Lith P, Nuchprayoon CV, Briney R, John-
son FN. The thorax in chronic obstructive lung disease.
Am J Med 1968; 44: 39–46.
17. Fleury B, Murciano D, Talamo C, Aubier M, Pariente R,
Milic-Emili J. Work of breathing in patients with chronic
obstructive pulmonary disease in acute respiratory fail-
ure. Am Rev RespirDis 1985; 131: 822–827.
18. Guerin C, Coussa ML, Eissa NT, et al. Lung and chest
wall mechanics in mechanically ventilated COPD pati-
ents. J Appl Physiol 1993; 74: 1570–1580.
19. Estenne M, Yernault JC, De Troyer A. Rib cage and dia-
phragm-abdomen compliance in humans: effects of age
and posture. J Appl Physiol 1985; 59: 1842–1848.
20. Haluszka J, Chartrand DA, Grassino AE, Milic-Emili J.
Intrinsic PEEP and arterial PCO2 in stable patients with
chronic obstructive pulmonary disease. Am Rev Respir
Dis 1990; 141: 1194–1197.
21. Vecchio LD, Polese G, Rossi R, Rossi A. "Intrinsic" positive
end-expiratory pressure in stable patients with chronic
obstructive pulmonary disease. Eur Respir J 1990; 3: 74–80.
22. Lessard MR, Lofaso F, Brochard L. Expiratory muscle
activity increases intrinsic positive end-expiratory pres-
sure independently of dynamic hyperinflation in mechan-
ically ventilated patients. Am J Respir Crit Care Med
1995; 151: 562–569.
23. Yan S, Kayser B, Tobiasz M, Sliwinski P. Comparison of
static and dynamic intrinsic positive end-expiratory pres-
sure using the Campbell diagram. Am J Respir Crit Care
Med 1996; 154: 938–944.
24. Yan S, Sliwinski P, Gauthier AP, Lichros I, Zakynthinos
S, Macklem PT. Effect of global inspiratory muscle fati-
gue on ventilatory and respiratory muscle responses to
CO2. J Appl Physiol 1993; 75: 1371–1377.
25. Lougheed MD, Webb KA, O'Donnell DE. Breathlessness
during induced lung hyperinflation in asthma: the role of
the inspiratory threshold load. Am J Respir Crit Care
Med 1995; 152: 911–920.
26. Martin JG, Shore S, Engel LA. Effect of continuous posi-
tive airway pressure on respiratory mechanics and pattern
of breathing in induced asthma. Am Rev Respir Dis 1982;
126: 812–817.
27. Puddy A, Younes M. Effect of slowly increasing elastic
load on breathing in conscious humans. J Appl Physiol
1991; 70: 1277–1283.
28. D'Angelo E, Robatto FM, Calderini E, et al. Pulmonary
and chest wall mechanics in anesthetized paralyzed hum-
ans. J Appl Physiol 1991; 70: 2602–2610.
