J Appl Physiol 99: 1402–1411, 2005.
First published June 16, 2005; doi:10.1152/japplphysiol.01165.2004.
Efficiency of the normal human diaphragm with hyperinflation
Kevin E. Finucane,1,2 Janine A. Panizza,1,2 and Bhajan Singh1,2,3
1
Department of Pulmonary Physiology, Sir Charles Gairdner Hospital, 2West
Australian Sleep Disorders Research Institute, QEII Medical Centre, and 3Department
of Physiology, University of Western Australia, Nedlands, Western Australia, Australia
Submitted 15 October 2004; accepted in final form 10 June 2005
in vivo diaphragm length; power output; electromyogram activity;
diaphragm contractile function and efficiency; fluoroscopy
striated muscle segments, including the diaphragm, is characterized by the instantaneous relationships between force, length, and velocity of shortening (4,
19). The power output of a muscle is the product of force
developed and velocity of shortening. The efficiency of a
muscle determines the power output achieved for a given rate
of oxygen consumption (V̇O2) and further defines the contractile function of muscle (69). In humans, the diaphragm accounts for a substantial fraction of inspiratory work (44, 45);
however, its efficiency is unknown. Measurement of diaphragm efficiency (Effdi) could be more useful in recognizing
and quantifying contractile dysfunction of the diaphragm than
measurements of strength or electromechanical effectiveness
measured, respectively, by maximum transdiaphragmatic pressure (Pdi; Pdimax) and the ratio of Pdi to the amount of
THE CONTRACTILE FUNCTION OF
Address for reprint requests and other correspondence: K. E. Finucane, Dept. of
Pulmonary Physiology, Sir Charles Gairdner Hospital, Hospital Ave., Nedlands,
Western Australia 6009, Australia (E-mail: [email protected]).
1402
electrical activity of the diaphragm (3, 31, 53). This hypothesis
is suggested by the mechanical behavior of muscle in vitro.
First, the mechanical output of a muscle during shortening is
defined by its power output and cannot be predicted from
maximum force alone (32). Second, during shortening, the
power output of a muscle is determined by muscle length and
level of activation, decreasing approximately linearly with both
length and activation (4, 11, 16, 19, 33). Third, muscle efficiency relates power output to the energy consumed, including
costs of activation and cross-bridge cycling (62, 69).
Assessment of the efficiency of the intact human diaphragm
(Effdi) requires breath-by-breath measurement of the inspiratory power output and V̇O2 of the diaphragm (V̇O2 di) and,
because efficiency and power output relative to activation vary
with muscle length (4, 11, 19, 33, 62), measurement of diaphragm length at end expiration (Ldi ee). Power output can be
estimated as the product of the average change during inspiration of Pdi (⌬Pdimean) and the rate of volume displaced by
the diaphragm (⌬Vdi䡠TI⫺1, where ⌬Vdi is the volume displaced by the diaphragm and TI is inspiratory duration). ⌬Vdi
has been estimated from motion of the chest wall (2, 17, 44),
change in the length of the diaphragm apposed to the rib cage
measured with ultrasound (30), and from analysis of radiographic images of diaphragm motion (58, 59). A fluoroscopic
method, developed and validated by us (59), allows breath-bybreath measurement of ⌬Vdi and of Ldi ee, enabling the effect
of diaphragm length on Effdi to be examined. V̇O2 di cannot be
measured in intact humans. However, studies in dogs (50) and
lambs (61) have shown a linear relationship between V̇O2 di and
the amount of diaphragm electrical activity. Beck et al. (7–10)
and Sinderby et al. (54, 56, 57) have developed and validated
methods for accurately quantifying the amount of diaphragm
electrical activity using multiple electrodes positioned in the
esophagus to span the crural diaphragm. This method controls
for diaphragm position (8, 10, 57) and signal contamination
(54, 56); electrical activity is quantified by the root mean
square of the electromyogram (EMG) signals (RMSdi) and is
an accurate index of global diaphragm activation (7, 60). We
propose that the inspiratory power output of the diaphragm
relative to the amount of electrical activity is an index of Effdi
Eff di ⫽ ⌬Pdimean 䡠 ⌬Vdi 䡠 TI⫺1 䡠 RMSdi⫺1
(1)
Interpretation of the proposed measurement of Effdi is complicated by the uncertain relationship between electrical activation and V̇O2 of the human diaphragm in vivo and by data
showing that the distribution of tension within (13), and activation of, the diaphragm (14, 22, 26) are inhomogeneous. In
The costs of publication of this article were defrayed in part by the payment
of page charges. The article must therefore be hereby marked “advertisement”
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
8750-7587/05 $8.00 Copyright © 2005 the American Physiological Society
http://www. jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
Finucane, Kevin E., Janine A. Panizza, and Bhajan Singh.
Efficiency of the normal human diaphragm with hyperinflation. J Appl
Physiol 99: 1402–1411, 2005. First published June 16, 2005;
doi:10.1152/japplphysiol.01165.2004.—We evaluated an index of diaphragm efficiency (Effdi), diaphragm power output (Ẇdi) relative to
electrical activation, in five healthy adults during tidal breathing at
usual end-expiratory lung volume (EELV) and diaphragm length
(Ldi ee) and at shorter Ldi ee during hyperinflation with expiratory
positive airway pressure (EPAP). Measurements were repeated with
an inspiratory threshold (7.5 cmH2O) plus resistive (6.5
cmH2O 䡠 l⫺1 䡠 s) load. Ẇdi was the product of mean inspiratory transdiaphragmatic pressure (⌬Pdimean), diaphragm volume displacement
measured fluoroscopically, and 1/inspiratory duration (TI⫺1). Diaphragm activation, measured with esophageal electrodes, was quantified by computing root-mean-square values (RMSdi). With EPAP, 1)
EELV increased [mean r2 ⫽ 0.91 (SD 0.01)]; 2) in four subjects, Ldi ee
decreased [mean r2 ⫽ 0.85 (SD 0.07)] and mean Effdi decreased 34%
per 10% decrease in Ldi ee (P ⬍ 0.001); and 3) in one subject, gastric
pressure at EELV increased two- to threefold, Ldi ee was unchanged or
increased, and Effdi increased at two of four levels of EPAP (P ⱕ
0.006, ANOVA). Inspiratory loading increased Ẇdi (P ⫽ 0.003) and
RMSdi (P ⫽ 0.004) with no change in Effdi (P ⫽ 0.63) or its
relationship with Ldi ee. Effdi was more accurate in defining changes in
Ldi ee [(true positives ⫹ true negatives)/total ⫽ 0.78 (SD 0.13)] than
⌬Pdimean 䡠 RMSdi⫺1, RMSdi, or ⌬Pdimean 䡠 TI (all ⬍0.7, P ⱕ 0.05,
without load). Thus Effdi was principally a function of Ldi ee independent of inspiratory loading, behavior consistent with muscle forcelength-velocity properties. We conclude that Effdi, measured during
tidal breathing and in the absence of expiratory muscle activity at
EELV, is a valid and accurate measure of diaphragm contractile
function.
EFFICIENCY OF THE HUMAN DIAPHRAGM
METHODS
Five healthy, nonsmoking men were studied. Subjects gave written,
informed consent. The Committee for Human Rights, University of
Western Australia, approved the study.
Measurements
Figure 1 summarizes the methods of measurement. Esophageal
(Pes), gastric (Pg), and mouth pressures and flow were measured, as
previously described (27). Signals were digitally recorded and displayed with computed Pdi (Pdi ⫽ Pg ⫺ Pes) and volume change at the
mouth (VT) (PowerLab, version 6, ADI Instruments). ⌬Vdi and Ldi ee
were measured from lateral fluoroscopic images of the diaphragm and
adjacent chest wall stored on videotape and analyzed, as previously
described (59). Radiation exposure was varied to optimize contrast of
the diaphragm silhouette and bony landmarks; the cumulative duration of fluoroscopy varied between 7 and 10 min, with radiation
exposures up to 2.6 mSv. Images were subsequently digitized by
using a monochrome video capture card (National Instruments PCI
1409) and LabVIEW software (National Instruments, version 6 with
IMAQ toolbox). Magnification and distortion of images in each
subject were defined from images of a grid of precise squares formed
by horizontal and vertical radiopaque lines at 1-cm intervals positioned at the same distance from the image intensifier (II) as the right
midclavicular line of the subject during fluoroscopy. This distance
was measured as the sum of the distances between the II and left chest
wall and, from posteroanterior chest X-rays at functional residual
capacity (FRC), the thickness of the left lateral chest wall and
three-fourths of the internal coronal diameter of the rib cage at the
level of the right costophrenic angle. The right midclavicular line was
assumed to define the plane of the most cephalad part of the right
hemidiaphragm producing the silhouette on fluoroscopy (67). RMSdi
was measured with the methods of Beck et al. (7, 8, 10) and Sinderby
et al. (54, 56). Eight stainless steel electrodes 2 mm wide and mounted
1 cm apart on a multicore Silastic catheter, 3.5 mm outer diameter,
were arranged as six bipolar pairs by using an overlapping array i.e.,
1 vs. 3, 2 vs. 4, 3 vs. 5, 4 vs. 6, 5 vs. 7, and 6 vs. 8 (Fig. 1). The
catheter was positioned in the esophagus so that the electrically active
center of the diaphragm (EACdi) (56) was located at electrode pair 4
vs. 6 at FRC during relaxed tidal breathing. The EACdi was identified
as the electrode pair where the EMG signals from the electrode pairs
above and below were opposite in polarity (see Fig. 1). EMG signals
were amplified (Grass Wideband A.C. amplifier, model 17P3C,
Quincy, MA), band-pass filtered between 10 and 1,000 Hz, digitized
at a sampling rate of 2 kHz, and stored (LAB View Version 6) for
subsequent analysis.
Protocol
Measurements were performed during tidal breathing at four to five
lung volumes, with and without a moderate inspiratory load. Subjects
sat erect with the left shoulder as close as possible to the II, with arms
elevated and hands resting on the head. The subject was positioned to
minimize rotation of the thorax relative to the II and to keep the image
Fig. 1. Diaphragm efficiency was calculated from simultaneous measurement of transdiaphragmatic pressure (Pdi) measured as the pressure difference between gastric (Pg) and
esophageal pressures (Pes), diaphragm volume displacement
(⌬Vdi) measured using lateral fluoroscopy, duration of inspiration and diaphragm electromyogram (EMG) measured with 6
overlapping bipolar electrode pairs (see inset), mounted on a
catheter positioned in the lower esophagus. The EMG was
quantified from the root mean square (RMS) of the signals. In
the illustration, the electrically active center of the diaphragm is
located at the 4th electrode pair. Subjects were seated erect, and
measurements were made during tidal breathing at functional
residual capacity (FRC) and during hyperinflation induced by
expiratory positive airway pressure (EPAP). Measurements
were repeated with an inspiratory threshold (7.5 cmH2O) plus
resistive (6.5 cmH2O 䡠 l⫺1 䡠 s) load. The displayed image of the
lower thorax has a 94° clockwise rotation. See text for details.
CF, center frequency.
J Appl Physiol • VOL
99 • OCTOBER 2005 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
addition, factors other than the contractile properties of muscle
may affect the efficiency of the intact diaphragm. Principal
among these are the geometry and mechanical advantage of the
diaphragm and the interaction between the diaphragm and
other respiratory muscles. Diaphragm mechanical advantage
decreases with hyperinflation (68) tending to decrease power
output relative to V̇O2. Contraction of abdominal muscles
during expiration and their relaxation during inspiration can
increase diaphragm power output (Ẇdi) during subsequent
inspiration (2, 45) and could increase efficiency.
The purpose of this study was to examine the validity and
utility of the proposed measurement of Effdi by comparing, in
healthy subjects, measurements obtained during tidal breathing
at usual end-expiratory lung volume (EELV) with those obtained at shorter diaphragm lengths during hyperinflation. We
hypothesized that any decrease in Ldi ee with hyperinflation
would be associated with a linear decrease in Effdi, consistent
with the force-length-velocity properties of muscle. Because
the electrical activity of the intact, human gastrocnemius muscle increases proportionately with load at a given velocity of
shortening (11), we hypothesized further that measured Effdi
would be independent of an inspiratory load, which did not
decrease the rate of ⌬Vdi.
1403
1404
EFFICIENCY OF THE HUMAN DIAPHRAGM
Data Analysis
For each breath, the TI was measured from the onset of inspiratory
flow or of a sustained decrease in Pes to the onset of expiratory flow.
The end-expiratory and inspiratory changes (⌬) of Pg, Pes, and Pdi
were measured, and ⌬Pdimean and diaphragm pressure-time product
per breath (⌬Pdimean 䡠 TI) were computed, taking Pdi at end expiration
(Pdiee) as baseline. EELV during EPAP was obtained by subtracting
measured IC from the vital capacity (VC) and expressed as a percentage of VC. The fluoroscopic images were digitized, and ⌬Vdi and
Ldi ee of each breath were measured by using the methods described
previously (58, 59) and an interactive analysis program written in
MATLAB (version 6.1.0.450). The first to fifth tidal breath and
subsequent inspiration to TLC of each condition of measurement were
identified, and the images at end expiration and end inspiration of each
breath were superimposed. Any mismatching between these images
was corrected by using as landmarks adjacent vertebral bodies and the
images of two ball bearings adherent to the posterior chest wall in the
midline at the level of the 10th thoracic vertebra. One image was
shifted relative to the other until cross-correlation of these markers
approximated 1. The images of the dome of the right hemidiaphragm
at end expiration and end inspiration, the posterior chest wall, and the
anterior border of the spinal column swept by the diaphragm were
segmented interactively by using a mouse-controlled cursor to identify multiple points along each of these structures. The anterior and
posterior insertions of the diaphragm at end expiration and end
inspiration, identified from bony landmarks adjacent to the costophrenic angles at TLC (15), were defined as single points. Each point
of the segmented images was corrected for distortion and magnification, applying data obtained from the grid of precise squares. The grid
image showed rotational, pin cushion radial, and decentering distortions, which were corrected by using a look-up table comprising the
coordinates of all cross points on the grid (52). After correcting for
distortion and magnification, the average error per calibration point
was 4.3%. Each point used to outline the diaphragm, spine, and chest
wall was converted to its correct position in space using the look-up
table, the four closest calibration points, and a least squares minimization technique (37). The points defining each structure were fitted
with lines using polynomial expressions. The square of the average
error of these fitted lines was typically 0.2 to 1 mm2. ⌬Vdi of each
inspiration and the lengths of the right hemidiaphragm in the sagittal
plane at end expiration and end inspiration were computed from the
J Appl Physiol • VOL
segmented, reconstructed images by using the methods and equations
described previously (59). The EMG signals of each inspiration were
analyzed in the time and frequency domains, as previously described
(60), by using the methods of Sinderby et al. (54, 56). Briefly, the
digitized EMG signals of each breath were displayed, and a 300- to
500-ms segment, including the maximum amplitude EMG, but avoiding any period containing an ECG complex, was selected for analysis.
Power spectra were evaluated for signal quality, accepting only those
data that met the criteria for uncontaminated signals (54). The EACdi
was defined by cross-correlation analysis of the signals from alternate
electrode pairs, and the RMSdi value was obtained by subtracting the
signals above and below the electrical center of the diaphragm (56).
This analysis yielded a single value of RMSdi and of center frequency
for each tidal inspiration. RMSdi was also measured during each
inspiration to TLC, and the maximum value (RMSdi max) was the
highest value obtained during any inspiration to TLC. RMSdi values
were corrected for changes in power spectrum, as may occur with
muscle fatigue, using the diaphragm center frequency of each breath
and the method of Beck et al. (7). Ẇdi (⌬Pdimean 䡠 ⌬Vdi 䡠 TI⫺1) and
Effdi of each breath were calculated.
Statistical Analysis
All variables were normalized by dividing the computed variables
of each breath by the mean value of that variable at FRC in each
subject. RMSdi was also normalized by expressing it as a percentage
of the highest value obtained during inspirations to TLC
(RMSdi %max). Linear regression was used to examine the relationships between 1) EPAP and EELV; 2) normalized Ldi ee and EELV; 3)
normalized Ldi ee and normalized Effdi and its components; and 4)
normalized Pg at end expiration (Pgee) and normalized Effdi. Multiple
linear regression analysis with stepwise elimination was used to
determine the relationships between the normalized dependent variables ⌬Vdi and Effdi and the normalized independent variables Ldi ee,
Pgee, Pdiee, TI, and VT 䡠 TI⫺1. One-way ANOVA was used to compare
Effdi with and without EPAP. The effect of inspiratory loading on the
measured and computed variables and on the relationships between
Effdi and the independent variables Ldi ee, Pgee, Pdiee, TI, and VT 䡠 TI⫺1
was examined with paired t-tests and multiple regression analysis,
respectively. The relative utility of the different measures of diaphragm function, i.e., efficiency, effectiveness, activation, and the
pressure-time product, were examined by comparing the degree of
association of each with normalized Ldi ee using Pearson’s correlation
coefficients and by calculating their accuracy in detecting a change in
normalized Ldi ee with hyperinflation. The accuracy of each variable
was calculated for each subject separately, as the number of true
positives plus true negatives divided by the total number of data
points. The accuracy of each variable for the group was taken as the
mean of the individual values. All results were analyzed by using
SigmaStat statistical software (version 2, SPSS, Chicago, IL) and are
reported as means and standard deviations (SD), unless otherwise
stated. Significance was defined as P ⬍ 0.05.
RESULTS
Mean age, body mass index, height, VC, and IC of the
subjects were 52 yr (SD 10), 26 kg/m2 (SD 1), 1.78 m (SD
0.05), 4.9 liters (SD 0.5), and 3.1 liters (SD 0.4), respectively.
Hyperinflation Without Load
Volumes and pressures. EELV increased with EPAP in all
subjects [mean r2 ⫽ 0.91 (SD 0.01)] (Fig. 2). Relative to FRC,
Pgee and ⌬Pes increased with EPAP in all subjects (Fig. 3).
Pgee was greatest in subject 5, increasing two- to threefold at
three of four levels of EPAP. Inspiratory ⌬Pg with EPAP was
variable; relative to FRC, ⌬Pg changed little in three subjects,
decreased in subject 2, and increased twofold in subject 5.
99 • OCTOBER 2005 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
of the dome of the diaphragm and adjacent chest wall within the
fluoroscopic field over the volume range FRC to total lung capacity
(TLC). Subjects breathed through a pneumotachograph and lowresistance two-way valve. The expiratory port was either open to
atmosphere or connected to a positive airway pressure unit (BiPAP S,
Respironics). The inspiratory port was either open to atmosphere or
connected to an inspiratory threshold valve (7.5 cmH2O) in series with
a resistance (6.5 cmH2O 䡠 l⫺1 䡠 s). Expiratory positive airway pressure
(EPAP) levels were adjusted between measurements to achieve predetermined levels of hyperinflation approximating 25, 50, and 75% of
the inspiratory capacity (IC) at FRC. Measurements with fluoroscopy
were made during five tidal breaths and an inspiration to TLC initiated
from end expiration. Conditions of measurement were randomized.
Before each measurement, EMG signals were reviewed to ensure that
the EACdi was at the fourth electrode pair. During measurement,
subjects were encouraged to relax during expiration; no other instructions were given regarding pattern of breathing. Where the level of
hyperinflation achieved differed substantially from the target levels or
the data appeared unsatisfactory, the measurements were repeated.
Subjects had one to three fully instrumented studies before fluoroscopy to familiarize them with the protocol, define the levels of EPAP
required to achieve targeted hyperinflation, and establish the range of
movement, with hyperinflation, of the EACdi in relation to electrode
pairs.
EFFICIENCY OF THE HUMAN DIAPHRAGM
Diaphragm length and volume change. At FRC, mean Ldi ee
was 27.0 cm (SD 2.5) (Table 1). Relative to its length at FRC,
Ldi ee decreased with hyperinflation in four subjects [mean r2 ⫽
0.85 (SD 0.07)] and did not change in subject 5, despite a 28%
increase in EELV (Fig. 4). Fractional shortening of the diaphragm between FRC and TLC was 0.28 (SD 0.1). ⌬Vdi
relative to VT was 55% (SD 14) at FRC and tended to decrease
with Ldi ee, being 35% (SD 12) (n ⫽ 4) at the shortest Ldi ee
achieved (P ⫽ 0.06). Multiple regression analysis showed that
⌬Vdi was positively correlated with Ldi ee (P ⬍ 0.001) and TI
(P ⫽ 0.02) and not correlated with Pdiee (P ⫽ 0.46) or Pgee
(P ⫽ 0.14). Where the diaphragm shortened with EPAP (n ⫽
4), ⌬Vdi was positively correlated with Ldi ee (P ⫽ 0.003) and
TI (P ⫽ 0.001) and negatively correlated with Pgee (P ⫽
0.006).
EMG. At each condition of measurement, there was at least
one and usually two electrode pairs distal to the EACdi. Mean
RMSdi %max was 15.9% (SD 3.0) at FRC and 66.5% (SD 6.2)
at the shortest Ldi ee (P ⬍ 0.001) (Table 1). Where Ldi ee
decreased with hyperinflation, RMSdi %max increased (r2 ⫽
0.44, P ⬍ 0.001).
Efficiency. A total of 126 tidal breaths and 28 inspirations to
TLC, distributed between 25 separate conditions of measurement among the five subjects, were imaged. Eighteen breaths
could not be analyzed because of loss or distortion of radiographic images (10 breaths), Pes artifact (4 breaths), and/or
EMG signal quality (5 breaths). A minimum of three tidal
breaths and the inspiration to TLC could be analyzed for each
condition of measurement in all subjects. Absolute values of
Effdi and its components at FRC and maximum hyperinflation
for all subjects are summarized in Table 1. The relationships
between normalized Ldi ee and normalized ⌬Pdimean,
⌬Vdi䡠TI⫺1, RMSdi, and Effdi are shown for subject 3 in Fig. 5.
In the four subjects in whom Ldi ee decreased with hyperinflation, diaphragm shortening was associated with an increase in
RMSdi (P ⬍ 0.001), a decrease in Effdi (P ⬍ 0.001), and no
change in Ẇdi and ⌬Vdi䡠TI⫺1. In subject 5, where Ldi ee
changed ⬍10% with hyperinflation, there was no significant
change in ⌬Pdimean, ⌬Vdi䡠TI⫺1, RMSdi, or Effdi with Ldi ee.
The relationship between normalized Ldi ee and normalized
Effdi for all breaths in all subjects is shown in Fig. 6. In subject
5, Effdi was highly variable and greater than that at FRC at two
of four levels of EPAP, i.e., 6 (P ⫽ 0.006) and 13 cmH2O (P ⫽
0.003, ANOVA) (Fig. 6). Multiple regression analysis (n ⫽ 5)
showed that normalized Effdi was highly correlated with Ldi ee
(P ⬍ 0.001) and not correlated with Pdiee (P ⫽ 0.204) or Pgee
(P ⫽ 0.193). Where Ldi ee decreased with hyperinflation (n ⫽
4), normalized Effdi was positively correlated with Ldi ee (P ⬍
0.001) and Pdiee (P ⫽ 0.01) and negatively correlated with Pgee
(P ⫽ 0.002); the latter accounted for ⬃5% of the variance
of Effdi.
Hyperinflation with Load
There were 20 conditions of measurement where all data
were obtained with and without inspiratory loading, including
three to four levels of EPAP in each subject. Considering
paired data at each level of EPAP, loading was associated with
an increase in ⌬Pdimean (P ⬍ 0.001), Ẇdi (P ⫽ 0.003), and
RMSdi (P ⫽ 0.004). Load had no effect on Ldi ee (Fig. 4B),
EELV, ⌬Vdi䡠TI⫺1, Pgee, Pdiee, or Effdi (Fig. 7B). The relationships between normalized Ldi ee and Effdi with and without load
are compared in Fig. 7. There was no difference in the slope or
Fig. 3. Effect of EPAP on end-expiratory Pes (Pesee), peak
inspiratory Pes (Pespeak), end-expiratory Pg (Pgee), and
peak inspiratory Pg (Pgpeak) during tidal breathing in each
subject. Data are means (SD) for each condition of measurement.
J Appl Physiol • VOL
99 • OCTOBER 2005 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
Fig. 2. Shown are end-expiratory lung volume (EELV) as percent vital
capacity (%VC) at rest, EPAP ⫽ 0, and each level of EPAP in each subject
(n ⫽ 5). The regression line is the mean slope and intercept of the individual
regressions.
1405
1406
EFFICIENCY OF THE HUMAN DIAPHRAGM
Table 1. Diaphragm efficiency and its components at FRC and maximum hyperinflation
Subject 2
Subject 3
Subject 4
Subjects 1– 4
27.1 (0.2)
18.8 (0.4)
29.2 (0.6)
22.3 (1.1)
28.3 (0.9)
22.0 (0.2)
22.7 (0.2)
19.5 (0.5)
27.0 (2.5)
21.0 (1.6)†
27.9 (0.6)
28.5 (0.7)
10.3 (1.2)
14.6 (1.6)
5.2 (0.6)
7.0 (0.9)
6.4 (0.3)
18.8 (2.8)
11.3 (1.5)
16.9 (2.9)
8.3 (3.0)
14.3 (5.1)*
7.5 (0.6)
24.8 (1.3)†
0.26 (0.04)
0.19 (0.02)
0.79 (0.09)
0.35 (0.09)
0.58 (0.03)
0.73 (0.75)
0.49 (0.04)
0.15 (0.04)
0.53 (0.22)
0.36 (0.26)
0.35 (0.07)
0.43 (0.07)
0.21 (0.0)
0.23 (0.03)
0.47 (0.06)
0.17 (0.05)
0.39 (0.09)
0.41 (0.20)
0.21 (0.03)
0.15 (0.04)
0.32 (0.13)
0.24 (0.12)
0.24 (0.04)
0.30 (0.06)
0.21 (0.04)
0.32 (0.07)
0.24 (0.05)
0.12 (0.04)
0.17 (0.01)
0.72 (0.1)
0.21 (0.02)
0.30 (0.12)
0.24 (0.03)
0.37 (0.26)
0.18 (0.03)
0.95 (0.26)†
15.8 (3.7)
67.0 (4.5)
17.0 (3.3)
63.0 (6.6)
11.0 (2.0)
61.0 (25.8)
15.5 (0.9)
75.0 (8.4)
15.9 (3.0)
66.5 (6.2)†
19.8 (3.2)
50.5 (7.1)†
10.4 (1.1)
7.75 (1.6)
4.24 (0.5)
1.3 (0.6)
5.48 (3.5)
2.1 (2.03)†
2.35 (0.4)
4.80 (1.2)‡
4.72 (1.4)
1.55 (0.3)
1.6 (0.3)
0.4 (0.15)
Subject 5
Values are means (SD). Ldi ee, diaphragm length at end expiration; FRC, functional residual capacity; EELV, end-expiratory lung volume; ⌬Pdimean, mean
inspiratory transdiaphragmatic pressure; ⌬Vdi, volume displaced by diaphragm motion; ⌬Vdi/TI, mean inspiratory flow rate from diaphragm, where TI is
inspiratory duration; Power, diaphragm power output; RMSdi%max, electrical activity of the diaphragm as a percent maximum; Effdi, diaphragm efficiency; au,
arbitrary units. Significant difference from FRC: *P ⫽ 0.01, ‡P ⫽ 0.009, †P⬍0.001 (paired t-test).
intercept of the mean regression for all subjects or for the four
subjects in whom Ldi ee decreased with hyperinflation. Multiple
regression analysis in the four subjects in whom Effdi decreased with Ldi ee showed that Effdi was highly correlated with
Ldi ee (P ⬍ 0.001), positively correlated with Pdiee (P ⫽ 0.037),
and not correlated with Pgee (P ⫽ 0.27).
Comparison of measures of diaphragm function. See Table
2. Normalized Ldi ee was more highly correlated with normalized Effdi than with other measures of diaphragm function,
independent of inspiratory loading. Effdi more accurately reflected changes in Ldi ee than other measures of diaphragm
function; however, with inspiratory loading, the differences
were not significant due to an increased variance of the
measurements.
DISCUSSION
In this study, diaphragm shortening, induced by hyperinflation, and inspiratory loading were used to evaluate a novel, in
vivo measure of diaphragm function, the ratio of inspiratory
power output to neural activation. This index of Effdi was
principally a function of diaphragm length and was independent of a modest inspiratory load. These findings are consistent
with the in vitro force-length-velocity and power output-neural
activation behavior of muscle (4, 11, 19) and suggest that the
measurement reflects the contractile properties of the diaphragm.
Critique of Methods and Assumptions
The efficiency of respiratory muscles is usually measured by
relating power output to the rate of respiratory muscle V̇O2 (18,
21, 43, 48). V̇O2 di is not measurable in intact humans but is
likely to be linearly related to the amount of electrical activity
of the diaphragm, independent of diaphragm length. First, in
humans, during submaximal rates of positive (miometric)
J Appl Physiol • VOL
work, electrical activity and V̇O2 of the quadriceps muscle are
linearly related (12). Second, in dogs (50) and lambs (61),
diaphragm electrical activity increased linearly with V̇O2 di
over an approximately sixfold increase in V̇O2 di. Third, the
relationship between V̇O2 and maximum isometric tension of
vascular smooth muscle is independent of muscle length (47).
Fourth, the V̇O2 associated with isotonic contraction of the
intact dog gastrocnemius muscle performing maximal work for
a short duration is a function of the number of nerve impulses
delivered to the muscle (28). Finally, the diaphragm obtains its
energy by oxidative metabolism over a wide range of work
outputs (5, 34, 41, 49, 51).
The amount of diaphragm electrical activity during each
inspiration was quantified by using the methods of Beck et al.
(7–10) and Sinderby et al. (54, 56, 57). These studies are the
basis of several assumptions that are central to our analysis of
Effdi. Namely, that the RMSdi of the crural diaphragm measured with these methods 1) accurately reflects activation of the
entire diaphragm (7); 2) is linearly related to diaphragm activation (7); 3) is not subject to artifact related to changes of lung
volume (7) or chest wall configuration (9); and 4) minimizes
artifact related to changes in the muscle to electrode distance
(8, 10).
The RMS of the crural diaphragm may not accurately define
activation of the entire diaphragm because regional activation
and shortening of the diaphragm during breathing are not
uniform (13, 22, 26, 64, 68). Furthermore, RMSdi is not linear
with activation, being proportional to the square root of the
number and firing frequency of motor units (7), thus tending to
underestimate higher levels of activation. Studies in both dogs
and humans suggest that regional activation within individual
respiratory muscles is directly proportional to regional mechanical advantage (23, 25, 39, 40). In the dog, diaphragm
mechanical advantage and blood flow are inhomogeneous but
99 • OCTOBER 2005 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
Ldi ee, cm
@FRC
@Maximum EELV
⌬Pdimean, cmH2O
@FRC
@Maximum EELV
⌬Vdi, liter
@FRC
@Maximum EELV
⌬Vdi/TI, l/s
@FRC
@Maximum EELV
Power, W
@FRC
@Maximum EELV
RMSdi%max
@FRC
@Maximum EELV
Effdi, au
@FRC
@Maximum EELV
Subject 1
EFFICIENCY OF THE HUMAN DIAPHRAGM
highly and positively correlated: they differ little between
adjacent costal and crural regions, and, with exercise, the
regional distribution of blood flow does not change (35). These
findings suggest a direct relationship between crural and global
diaphragm EMG that is independent of the level of activation.
In humans, the RMS of the crural diaphragm is linearly related
to a measure of global activation of the diaphragm, the ratio of
Pdi to Pdimax, up to 75% of Pdimax independent of diaphragm
length (7). We have shown that, with hypercapnia in healthy
subjects, RMSdi increases linearly with end-tidal PCO2 up to at
least 60% RMSdi max, suggesting a linear relationship between
neural drive to the diaphragm and the RMS of the crural
diaphragm (60). Thus available evidence suggests that the
RMS of the crural diaphragm is linearly related to activation of
the entire diaphragm, up to ⬃75% maximum. In the present
study, all measurements without load and the majority with
load were within the range at which RMSdi and global diaphragm activation are linearly related (7, 60). RMSdi exceeded
75% of RMSdi max during inspiratory loading at the highest
J Appl Physiol • VOL
level of EPAP in two subjects, potentially underestimating
activation and overestimating efficiency. However, in keeping
with results at lower levels of EPAP and diaphragm activation,
Effdi of these measurements was not different from those
without load.
Hyperinflation with EPAP causes an inspiratory threshold
load that would have contributed to an increase of RMSdi with
hyperinflation, independent of diaphragm shortening. However, we found that a moderate inspiratory threshold plus
resistive load was associated with approximately proportional
changes in Ẇdi and RMSdi and was not associated with a
change in Effdi or in the relationship between efficiency and
diaphragm length. This result is in keeping with the in vivo
relationships between force, velocity of shortening, and
amount of muscle electrical activity of human gastrocnemius
where, at a given muscle length and velocity of shortening, the
amount of electrical activity increased proportionately with
load (11). In healthy subjects breathing against resistive loads,
respiratory efficiency is constant over a wide range of power
outputs (21), and, in dogs, Effdi is constant with progressive
resistive loading (48). Taken together, these results suggest
that, in the intact human diaphragm, inspiratory loading associated with the levels of EPAP used in this study was not
responsible for the decrease in measured Effdi with diaphragm
shortening.
⌬Vdi was derived as previously described (58, 59); inaccuracy associated with the computer-assisted method of image
analysis was small (see METHODS) and did not add significantly
to the errors of measurement of Ldi or ⌬Vdi. Mean ⌬Vdi䡠VT⫺1
at FRC and fractional shortening of the diaphragm between
FRC and TLC were not different from those in the 10 healthy
subjects reported previously (59). Measured ⌬Vdi does not
include the contribution to inspired lung volume resulting from
the action of the costal diaphragm in expanding and elevating
the rib cage. Although this component of ⌬Vdi is believed to
be small (36, 66), any effect would tend to diminish with
hyperinflation so that consequent underestimates of efficiency
would be greatest at FRC, thereby reducing the change in Effdi
with decreasing Ldi ee. In our study, RMSdi %max during tidal
breathing at FRC was 1.9 times greater than that reported for
healthy subjects by Sinderby et al. (53). This difference is
attributable to a lower respiratory rate and larger (1.8 times) VT
in our subjects relative to those of Sinderby et al.
Implications
In our study, the decrease in Effdi with hyperinflation was
due principally to a decrease in diaphragm length and was
independent of modest inspiratory loading. These findings are
consistent with the interdependence of force, velocity of shortening, length, and electrical activity of muscle (4, 11, 19) and
suggest that measured Effdi reflects the contractile properties of
diaphragm muscle. Additionally, the results suggest that other
factors can affect in vivo measurements of Effdi. First, activity
of expiratory muscles at end expiration, as in subject 5, can
increase Effdi. To the extent that Ẇdi at a given level of
activation and muscle length is approximately constant and
independent of load (11), only passive recoil of the diaphragm
would act to increase efficiency. However, this contribution
could be modulated by the degree of relaxation of abdominal
muscles during inspiration, perhaps explaining the variable
99 • OCTOBER 2005 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
Fig. 4. A: effect of pulmonary hyperinflation, without inspiratory loading, on
diaphragm length at end expiration (Ldi ee) normalized for diaphragm length at
FRC. EELV is expressed as %VC. Data are means (SD) at each condition of
measurement. The regression line is the mean slope and intercept of the
individual regressions in the 4 subjects where the diaphragm shortened with
hyperinflation [mean r2 ⫽ 0.85 (0.07)]. B: Bland and Altman comparison of
the mean Ldi ee at each level of EPAP (0 –14 cmH2O) in the 5 subjects. The
solid line is the mean difference between load (L) and no load (NL) (P ⫽ 0.67),
and the dashed lines are limits of the 95% confidence intervals.
1407
1408
EFFICIENCY OF THE HUMAN DIAPHRAGM
Fig. 5. Results in subject 3 without inspiratory loading. Relationships between Ldi ee
and mean inspiratory Pdi (⌬Pdimean; A),
mean inspiratory flow rate from the diaphragm (⌬Vdi 䡠 TI⫺1, where TI is inspiratory
duration; B), electrical activity of the diaphragm (RMSdi; C), and diaphragm efficiency (Effdi; D) for each tidal breath. All
data are normalized to the mean value of
each variable at FRC. Linear regressions
and their coefficients of determination (r2)
and P values are displayed.
Fig. 6. Relationship between normalized Effdi and normalized Ldi ee for each
breath in all subjects without load. Subject 5 is represented by open squares.
The regression line is the mean regression and intercept of the individual
regressions in the 5 subjects.
J Appl Physiol • VOL
with exercise (1), hypercapnia (65), inspiratory loading (42),
and chronic obstructive pulmonary disease (COPD) (46). Its
occurrence has the potential to increase Effdi relative to diaphragm length and confound interpretation of the measurement
because, in this event, Effdi reflects both the contractile properties of the diaphragm and the work performed on the diaphragm by expiratory muscles. In these circumstances, a measurement of Effdi, which principally reflects contractile function relative to diaphragm length, could be obtained by
incremental measurement during a continuous inspiration from
relaxed FRC to near TLC.
The results are qualitatively in accord with previous human
studies showing a reduced efficiency of respiratory muscles
with increasing lung volume (20, 43). Previous in vitro studies
show a substantial, although variable, effect of muscle length
on the efficiency or maximum power output of striated muscle
segments during cyclical activation (4, 33, 38, 62). The efficiency of frog and maximum power of rat ventricular muscle
and of fast and slow skeletal muscles of mice decreased
linearly by 54, 56, 39, and 32% of maximum, respectively, per
10% decrease in muscle length below 90% of the length at
maximum isometric force. In our study, efficiency decreased
by ⬃34% for each 10% decrease in diaphragm length. The
similar effects of muscle shortening on efficiency or maximum
power output in vitro and in the intact human diaphragm
support the concept that diaphragm inspiratory power output
relative to neural activation is an index of the efficiency of the
diaphragm.
In emphysema with hyperinflation, we found that Ldi ee at
FRC was ⬃25% less than in healthy controls (58). The data
from the present study (Fig. 6) suggest that a 25% decrease in
Ldi ee would, in the absence of remodeling of the diaphragm
(63), be associated with an ⬃80% decrease in Effdi and require
an approximately fourfold increase in activation to maintain
Ẇdi at normal levels (Table 2). This finding is in general
accord with previous results in subjects with COPD in whom
99 • OCTOBER 2005 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
increase in Effdi in subject 5. Second, a decrease in diaphragm
mechanical advantage is suggested by the small decrease in
Effdi with Pgee in the four subjects in whom Ldi ee decreased
with hyperinflation and Pgee increased. Passive hyperinflation
increases Pg and decreases the mechanical advantage of the
diaphragm in dogs (68). A decreased mechanical advantage
would be expected to decrease Effdi because, at a given
diaphragm length, mechanical advantage determines power
output relative to diaphragm oxygen uptake (35). Third, the
efficiency of muscle in vitro decreases at low and high velocities of shortening (32). In our study, mean diaphragm flow rate
changed little with hyperinflation (Table 1) or loading and was
not systematically related to Effdi. Thus, in our study, flowrelated changes in Effdi were not observed. Expiratory muscle
activity at end expiration is not a usual response to continuous
positive airway pressure in supine humans (65) but is usual
1409
EFFICIENCY OF THE HUMAN DIAPHRAGM
Table 2. Correlations between and accuracy of measures of diaphragm function with respect to diaphragm
length at various lung volumes
Correlation Coefficients
Measures of
Diaphragm Function
Definition
Efficiency
˙ di/RMS
W
di
Effectiveness
⌬Pdimean/RMSdi
Neural activation
RMSdi%max
Pressure ⫻ time
⌬Pdimean䡠TI
Accuracy
No Load
Load
No Load
Load
0.64
P⬍0.001
0.61
P⬍0.001
0.52
P⬍0.001
0.10
P ⫽ 0.31
0.64
P⬍0.001
0.58
P⬍0.001
0.45
P⬍0.001
0.42
P⬍0.001
0.78 (0.13)
0.69 (0.14)
P ⫽ 0.029
0.67 (0.20)
P ⫽ 0.048
0.45 (0.36)
P ⫽ 0.025
0.89 (0.10)
0.64 (0.37)
P ⫽ 0.076
0.67 (0.35)
P ⫽ 0.145
0.61 (0.34)
P ⫽ 0.109
Pearson’s correlations are shown between measures of diaphragm function and Ldi ee for all subjects (n ⫽ 5). Ẇdi/RMSdi, inspiratory diaphragm power
output/amount of inspiratory diaphragm electromyogram (EMG); ⌬Pdimean/RMSdi, mean transdiaphragmatic pressure/amount of inspiratory diaphragm EMG;
RMSdi%max, amount of inspiratory diaphragm EMG normalized to the maximum value obtained during inspirations to total lung capacity; ⌬Pdimean䡠TI, product
of mean transdiaphragmatic pressure and inspiratory time. Measures of diaphragm function and Ldi ee were normalized by dividing the value obtained for each
breath by the mean value at FRC in each subject. Accuracy of each measure of diaphragm function was calculated as the mean (SD) of the individual values.
Under Accuracy, the P values define the significance of the difference from diaphragm efficiency (paired t-test).
J Appl Physiol • VOL
99 • OCTOBER 2005 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
Fig. 7. A: normalized Effdi relative to normalized Ldi ee for each breath without
(solid symbols) and with (open symbols) inspiratory loading. Subject 5 is
represented by squares. The regression lines are the means of the individual
regressions in the 5 subjects without (dashed line) and with (solid line)
inspiratory loading. The mean slopes and intercepts were not different (P ⫽
0.61 and 0.42, respectively). B: Bland and Altman comparison of the mean
Effdi (arbitrary units, au) at each level of EPAP (0 –14 cmH2O) in the 5
subjects. The solid line is the mean difference between L and NL (P ⫽ 0.63),
and the dashed lines are limits of the 95% confidence interval.
RMSdi %max during quiet breathing was approximately five
times that in healthy subjects (53). In COPD, during quiet
breathing (24, 53) and progressive exercise (55), diaphragm
electrical activity is markedly increased, implying either decreased Effdi, or increased respiratory loads, or both. However,
without measurement of the rate of ⌬Vdi and power output, the
Effdi in COPD remains undefined.
Effdi was more closely correlated with diaphragm length and
more accurate in defining changes of Ldi ee than diaphragm
effectiveness (⌬Pdimean 䡠RMSdi⫺1). Although the differences
were small and not significant with inspiratory loading (Table
2), the results imply that assessment of diaphragm contractile
function is incomplete without considering both the pressure
developed by the diaphragm and its rate of ⌬Vdi. This conclusion differs from that of Beck et al. (6), who showed no
decrease in ⌬Pdimean 䡠RMSdi⫺1 with increasing inspiratory
flows up to 1.4 l/s, suggesting no measurable effect of diaphragm velocity of shortening. However, the need to consider
the rate of volume displacement is emphasized by the diaphragm pressure-velocity behavior during progressive exercise
in healthy subjects (2). In this latter study, ⌬Pdi increased less
than twofold, whereas estimated velocity of diaphragm shortening increased more than sixfold between rest and exercise to
70% maximum workload. In this case, diaphragm effectiveness
would underestimate efficiency by a factor of 6. Similarly,
measured efficiency was more accurate than RMSdi %max and
⌬Pdimean 䡠TI in defining changes in Ldi ee, implying that indexes
of diaphragm activation (24, 53) or of V̇O2 di (29) provide a less
complete assessment of diaphragm function than one which
considers both power output and neural activation.
Measurement of Effdi should allow an estimate of the maximum power output and reserve capacity of the diaphragm in
individuals. For example, in the four subjects in whom Ldi ee
decreased with hyperinflation, mean RMSdi %max was 67% (SD
6) at the shortest Ldi ee achieved without inspiratory loading.
Assuming that efficiency does not change with power output, this finding implies a maximum power output reserve of
the diaphragm of ⬃33% at diaphragm lengths commonly
found in COPD.
In conclusion, the results of this study suggest that the power
output of the diaphragm relative to the amount of electrical
activity during tidal inspirations reflects the efficiency and
1410
EFFICIENCY OF THE HUMAN DIAPHRAGM
contractile function of the human diaphragm. Expiratory muscle activity at end expiration can maintain diaphragm length
near that at FRC, despite hyperinflation, and increase efficiency. However, this effect was variable, indicating that the
interactions between activity of respiratory muscles, Ldi ee, and
passive recoil of the diaphragm require systematic evaluation.
18.
19.
20.
ACKNOWLEDGMENTS
The authors thank Y. M. Lam and Kai Shen for developing, respectively,
the EMG and ⌬Vdi analysis software, Karen Wills for assistance with data
analysis and preparation of the manuscript, Y. M. Lam and Satvinder Dhaliwal
for statistical advice, N. Hicks for performing fluoroscopy, and the Department
of Radiology, Sir Charles Gairdner Hospital, for access to equipment and
materials.
21.
22.
23.
GRANTS
24.
REFERENCES
26.
1. Abraham KA, Feingold H, Fuller DD, Jenkins M, Mateika JH, and
Fregosi RF. Respiratory-related activation of human abdominal muscles
during exercise. J Physiol 541: 653– 663, 2002.
2. Aliverti A, Cala SJ, Duranti R, Ferrigno G, Kenyon CM, Pedotti A,
Scano G, Sliwinski P, Macklem PT, and Yan S. Human respiratory
muscle actions and control during exercise. J Appl Physiol 83: 1256 –
1269, 1997.
3. American Thoracic Society/European Respiratory Society. ATS/ERS
statement on respiratory muscle testing. Am J Respir Crit Care Med 166:
518 – 624, 2002.
4. Bahler AS, Fales JT, and Zierler KL. The dynamic properties of
mammalian skeletal muscle. J Gen Physiol 51: 369 –384, 1968.
5. Bazzy AR, Pang LM, Akabas SR, and Haddad GG. O2 metabolism of
the sheep diaphragm during flow resistive loaded breathing. J Appl Physiol
66: 2305–2311, 1989.
6. Beck J, Sinderby C, Lindstrom L, and Grassino A. Crural diaphragm
activation during dynamic contractions at various inspiratory flow rates.
J Appl Physiol 85: 451– 458, 1998.
7. Beck J, Sinderby C, Lindstrom L, and Grassino A. Effects of lung
volume on diaphragm EMG signal strength during voluntary contractions.
J Appl Physiol 85: 1123–1134, 1998.
8. Beck J, Sinderby C, Lindstrom L, and Grassino A. Influence of bipolar
esophageal electrode positioning on measurements of human crural diaphragm electromyography. J Appl Physiol 81: 1434 –1449, 1996.
9. Beck J, Sinderby C, Lindstrom L, and Grassino G. Diaphragm interference pattern EMG and compound muscle action potentials: effects of
chest wall configuration. J Appl Physiol 82: 520 –530, 1997.
10. Beck J, Sinderby C, Weinberg J, and Grassino A. Effects of muscleto-electrode distance on the human diaphragm electromyogram. J Appl
Physiol 79: 975–985, 1995.
11. Bigland B and Lippold OC. The relation between force, velocity and
integrated electrical activity in human muscles. J Physiol 123: 214 –224,
1954.
12. Bigland-Ritchie B and Woods JJ. Integrated EMG and oxygen uptake
during dynamic contractions of human muscles. J Appl Physiol 36:
475– 479, 1974.
13. Boriek AM, Rodarte JR, and Reid MB. Shape and tension distribution
of the passive rat diaphragm. Am J Physiol Regul Integr Comp Physiol
280: R33–R41, 2001.
14. Brancatisano A, Kelly SM, Tully A, Loring SH, and Engel LA.
Postural changes in spontaneous and evoked regional diaphragmatic activity in dogs. J Appl Physiol 66: 1699 –1705, 1989.
15. Braun NM, Arora NS, and Rochester DF. Force-length relationship of
the normal human diaphragm. J Appl Physiol 53: 405– 412, 1982.
16. Brooks SV, Faulkner JA, and McCubbrey DA. Power outputs of slow
and fast skeletal muscles of mice. J Appl Physiol 68: 1282–1285, 1990.
17. Cala SJ, Kenyon CM, Ferrigno G, Carnevali P, Aliverti A, Pedotti A,
Macklem PT, and Rochester DF. Chest wall and lung volume estimation
J Appl Physiol • VOL
25.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
99 • OCTOBER 2005 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
This study was supported by a grant from the Medical Research Fund of
Western Australia and Sir Charles Gairdner Hospital Research Fund. B. Singh
is the recipient of an Australian Lung Foundation/Boehringer Ingelheim
Chronic Airflow Limitation Research Fellowship.
by optical reflectance motion analysis. J Appl Physiol 81: 2680 –2689,
1996.
Campbell EJM, Westlake EK, and Cherniack RM. The oxygen consumption and efficiency of the respiratory muscles of young male subjects.
Clin Sci (Lond) 18: 55– 64, 1959.
Coirault C, Riou B, Bard M, Suard I, and Lecarpentier Y. Contraction,
relaxation, and economy of force generation in isolated human diaphragm
muscle. Am J Respir Crit Care Med 152: 1275–1283, 1995.
Collett PW and Engel LA. Influence of lung volume on oxygen cost of
resistive breathing. J Appl Physiol 61: 16 –24, 1986.
Collett PW, Perry C, and Engel LA. Pressure-time product, flow, and
oxygen cost of resistive breathing in humans. J Appl Physiol 58: 1263–
1272, 1985.
Darian GB, DiMarco AF, Kelsen SG, Supinski GS, and Gottfried SB.
Effects of progressive hypoxia on parasternal, costal, and crural diaphragm
activation. J Appl Physiol 66: 2579 –2584, 1989.
De Troyer A, Gorman RB, and Gandevia SC. Distribution of inspiratory drive to the external intercostal muscles in humans. J Physiol 546:
943–954, 2003.
De Troyer A, Leeper JB, McKenzie DK, and Gandevia SC. Neural
drive to the diaphragm in patients with severe COPD. Am J Respir Crit
Care Med 155: 1335–1340, 1997.
De Troyer A and Legrand A. Inhomogeneous activation of the parasternal intercostals during breathing. J Appl Physiol 79: 55– 62, 1995.
Easton PA, Fitting JW, Arnoux R, Guerraty A, and Grassino AE.
Costal and crural diaphragm function during CO2 rebreathing in awake
dogs. J Appl Physiol 74: 1406 –1418, 1993.
Eastwood PR, Hillman DR, and Finucane KE. Ventilatory responses to
inspiratory threshold loading and role of muscle fatigue in task failure.
J Appl Physiol 76: 185–195, 1994.
Fales JT, Heisey R, and Zierler KL. Dependency of oxygen consumption of skeletal muscle on number of stimuli during work in the dog. Am J
Physiol 198: 1333–1342, 1960.
Field S, Sanci S, and Grassino A. Respiratory muscle oxygen consumption estimated by the diaphragm pressure-time index. J Appl Physiol 57:
44 –51, 1984.
Gorman RB, McKenzie DK, Pride NB, Tolman JF, and Gandevia SC.
Diaphragm length during tidal breathing in patients with chronic obstructive pulmonary disease. Am J Respir Crit Care Med 166: 1461–1469,
2002.
Grassino A, Goldman MD, Mead J, and Sears TA. Mechanics of the
human diaphragm during voluntary contraction: statics. J Appl Physiol 44:
829 – 839, 1978.
Hill AV. The heat of shortening and the dynamic constants of muscle.
Proc R Soc London Ser B 126: 136 –195, 1938.
James RS, Altringham JD, and Goldspink DF. The mechanical properties of fast and slow skeletal muscles of the mouse in relation to their
locomotary function. J Exp Biol 198: 491–502, 1995.
Jardim J, Farkas G, Prefaut D, Thomas D, Macklem PT, and Roussos
C. The failing inspiratory muscles under normoxic and hypoxic conditions. Am Rev Respir Dis 124: 274 –279, 1981.
Johnson RL Jr, Hsia CC, Takeda S, Wait JL, and Glenny RW.
Efficient design of the diaphragm: distribution of blood flow relative to
mechanical advantage. J Appl Physiol 93: 925–930, 2002.
Kenyon CM, Cala SJ, Yan S, Aliverti A, Scano G, Duranti R, Pedotti
A, and Macklem PT. Rib cage mechanics during quiet breathing and
exercise in humans. J Appl Physiol 83: 1242–1255, 1997.
Lagarias JC, Reeds JA, Wright MH, and Wright PE. Convergence
properties of the Nelder-Mead method in low dimensions. SIAM J Optimization 9: 112–147, 1998.
Layland J, Young IS, and Altringham JD. The length dependence of
work production in rat papillary muscle in vitro. J Exp Biol 198: 2491–
2499, 1995.
Legrand A, Brancatisano A, Decramer M, and De Troyer A. Rostrocaudal gradient of electrical activation in the parasternal intercostal muscles of the dog. J Physiol 495: 247–254, 1996.
Legrand A and De Troyer A. Spatial distribution of external and internal
intercostal activity in dogs. J Physiol 518: 291–300, 1999.
Manohar M and Hassan AS. Diaphragmatic energetics during prolonged
exhaustive exercise. Am Rev Respir Dis 144: 415– 418, 1991.
Martin JG and De Troyer A. The behaviour of the abdominal muscles
during inspiratory mechanical loading. Respir Physiol 50: 63–73, 1982.
EFFICIENCY OF THE HUMAN DIAPHRAGM
J Appl Physiol • VOL
56. Sinderby CA, Beck JC, Lindstrom LH, and Grassino AE. Enhancement of signal quality in esophageal recordings of diaphragm EMG.
J Appl Physiol 82: 1370 –1377, 1997.
57. Sinderby CA, Comtois AS, Thomson RG, and Grassino AE. Influence
of the bipolar electrode transfer function on the electromyogram power
spectrum. Muscle Nerve 19: 290 –301, 1996.
58. Singh B, Eastwood PR, and Finucane KE. Volume displaced by diaphragm motion in emphysema. J Appl Physiol 91: 1913–1923, 2001.
59. Singh B, Panizza JA, and Finucane KE. Breath-by-breath measurement
of the volume displaced by diaphragm motion. J Appl Physiol 94:
1084 –1091, 2003.
60. Singh B, Panizza JA, and Finucane KE. Diaphragm electromyogram
root mean square response to hypercapnoea and its intersubject and
day-to-day variation. J Appl Physiol 98: 274 –281, 2005.
61. Soust M, Walker AM, and Berger PJ. Diaphragm V̇O2, diaphragm
EMG, pressure-time product and calculated ventilation in newborn lambs
during hypercapnic hyperpnoea. Respir Physiol 76: 107–117, 1989.
62. Syme DA. The efficiency of frog ventricular muscle. J Exp Biol 197:
143–164, 1994.
63. Thomas AJ, Supinski GS, and Kelsen SG. Changes in chest wall
structure and elasticity in elastase-induced emphysema. J Appl Physiol 61:
1821–1829, 1986.
64. Wakai Y, Leevers AM, and Road JD. Regional diaphragm shortening
measured by sonomicrometry. J Appl Physiol 77: 2791–2796, 1994.
65. Wakai Y, Welsh MM, Leevers AM, and Road JD. Expiratory muscle
activity in the awake and sleeping human during lung inflation and
hypercapnia. J Appl Physiol 72: 881– 887, 1992.
66. Ward ME, Ward JW, and Macklem PT. Analysis of human chest wall
motion using a two-compartment rib cage model. J Appl Physiol 72:
1338 –1347, 1992.
67. Whitelaw WA. Shape and size of the human diaphragm in vivo. J Appl
Physiol 62: 180 –186, 1987.
68. Wilson TA, Boriek AM, and Rodarte JR. Mechanical advantage of the
canine diaphragm. J Appl Physiol 85: 2284 –2290, 1998.
69. Woledge RC, Curtin NA, and Homsher E. Energetic aspects of muscle
contraction. Monogr Physiol Soc 41: 1–357, 1985.
99 • OCTOBER 2005 •
www.jap.org
Downloaded from http://jap.physiology.org/ by 10.220.32.247 on June 17, 2017
43. McCool FD, Tzelepis GE, Leith DE, and Hoppin FG Jr. Oxygen cost
of breathing during fatiguing inspiratory resistive loads. J Appl Physiol 66:
2045–2055, 1989.
44. Mead J and Loring SH. Analysis of volume displacement and length
changes of the diaphragm during breathing. J Appl Physiol 53: 750 –755,
1982.
45. Mead J, Loring SH, and Smith JC. Volume displacements of the chest
wall and their mechanical significance. In: The Thorax (2nd ed.), edited by
Roussos C. New York: Dekker, 1995, p. 565–586.
46. Ninane V, Rypens F, Yernault JC, and De Troyer A. Abdominal
muscle use during breathing in patients with chronic airflow obstruction.
Am Rev Respir Dis 146: 16 –21, 1992.
47. Paul RJ and Peterson JW. Relation between length, isometric force, and
O2 consumption rate in vascular smooth muscle. Am J Physiol 228:
915–922, 1975.
48. Reid MB and Johnson RLJ. Efficiency, maximal blood flow, and aerobic
work capacity of the canine diaphragm. J Appl Physiol 54: 763–772, 1983.
49. Robertson CHJ, Foster GH, and Johnson RLJ. The relationship of
respiratory failure to the oxygen consumption of, lactic acid by, and
distribution of blood flow among respiratory muscles during increasing
inspiratory resistance. J Clin Invest 59: 31– 42, 1977.
50. Rochester DF and Bettini G. Diaphragmatic blood flow and energy
expenditure in the dog. Effects of inspiratory airflow resistance and
hypercapnia. J Clin Invest 57: 661– 672, 1976.
51. Rochester DF and Briscoe AM. Metabolism of the working diaphragm.
Am Rev Respir Dis 119: 101–106, 1979.
52. Shen K. Interactive Computerised Functional Analysis of Lateral Fluoroscopy Images to Quantify the Volume Displaced by Motion of the
Diaphragm (Honours Thesis). Perth: University of Western Australia,
2002.
53. Sinderby C, Beck J, Spahija J, Weinberg J, and Grassino A. Voluntary
activation of the human diaphragm in health and disease. J Appl Physiol
85: 2146 –2158, 1998.
54. Sinderby C, Lindstrom L, and Grassino AE. Automatic assessment of
electromyogram quality. J Appl Physiol 79: 1803–1815, 1995.
55. Sinderby C, Spahija J, Beck J, Kaminski D, Yan S, Comtois N, and
Sliwinski P. Diaphragm activation during exercise in chronic obstructive
pulmonary disease. Am J Respir Crit Care Med 163: 1637–1641, 2001.
1411