- Email: [email protected]
The effect of continuous positive airway pressure versus positive end-expiratory pressure on the diaphragm
The Effect of Continuous Positive Airway Pressure Versus Positive End-Expiratory Pressure on the Diaphragm J.D. Road, A.M. Leevers,
and A. Grassino
Continuous positive airway pressure (CPAP) and positive end-expiratory pressure (PEEP) both increase lung volume and hence may compromise diaphragm function. However, the effects of these two positive airway pressure modalities on inspiratory work of breathing are conflicting. In this study, we compared the effect of CPAP versus PEEP on diaphragm function in spontaneously breathing anesthetized dogs. Eight sodium pentobarbital-anesthetized dogs were randomly exposed to various levels of CPAP and PEEP. Measurements of diaphragmatic shortening, transdiaphragmatic pressure swings, and diaphragmatic electromyogram (EMG) were made. The change in lung volume and diaphragm length was similar at equivalent airway pressures during PEEP or CPAP. Therefore, expiratory muscle recruitment in the two conditions was equivalent.
However, tidal diaphragmatic EMG and transdiaphragmatic pressure swings increased markedly during PEEP compared with CPAP. At a PEEP of 18 cm H,O, crural and costal EMG activities were 185% r 16% and 163% ? 8% of control, respectively, whereas during CPAP the EMG activity was 66% 2 11% of control for both the costal and the crural diaphragms (&SE). During PEEP, the duration of neural inspiration (T,,,) was greater than the duration of inspiration as measured by airflow (TJ. On the other hand, during CPAP, T ,EMo was less than T,. We conclude that although expiratory muscle recruitment is comparable and tidal volume greater during CPAP, the inspiratory activation of the diaphragm decreases with CPAP but increases markedly with PEEP. Copyright o 1991 by W. B. Saunders Company
B
PEEP and CPAP and the work of breathing in these two studies, however, have been complicated by the different changes in EELV produced by the two modalities. We therefore wished to define the effect of these two positive airway pressure modalities on the function of the main inspiratory muscle, the diaphragm. The work of breathing as estimated by pressure and volume ignores the work done by the respiratory muscles on distortion of the chest wall. Direct measurements of respiratory muscle contraction can therefore provide information not available from analysis of pressure and volume change. Therefore, using the techniques of sonomicrometry4 and electromyography (EMG) to directly measure muscle contraction and activation, we compared the effects of PEEP and CPAP on the diaphragm in anesthetized dogs.
OTH POSITIVE end-expiratory pressure (PEEP) and continuous positive airway pressure (CPAP) have been used to increase functional residual capacity (FRC) and improve gas exchange. Both positive airway pressure modalities may increase the work of breathing; however, the increase has been shown to be less with CPAP compared with PEEP.’ The main mechanism leading to a relative reduction in work of breathing during CPAP was attributed to the reduction in inspiratory work at the expense of an increase in expiratory work. Increased expiratory work prevents the expected increase in end-expiratory lung volume (EELV) and, with CPAP but not with PEEP, allows an inspiratory assist. Schlobohm et al,’ however, reported that oxygen consumption was not different with either modality. The inspiratory assist produced by CPAP3 should have reduced the inspiratory work of breathing and hence reduced inspiratory muscle work and oxygen consumption. Comparisons between From the Department of Medicine, University of British Columbia, Vancouver, BC; and Notre Dame Hospital and The Meakins-Christie Laboratories, Montreal, Quebec. Received December 19, 1990; May 14,199l. Dr Road was a scholar of the British Columbia Health Care Research Foundation. Address reprint requests to J.D. Road, MD, Department of Medicine, University Hospital (UBC), 221 I Wesbrook Mall, Vancouver, BC V6T 1 W5. Copyright 0 1991 by W.B. Saunders Company 0883-9441191/0603-0008$05.00/O 136
METHODS To compare PEEP versus CPAP, eight pentobarbital sodium (25 to 30 mgikg intravenously) anesthetized dogs (mean weight, 24 f 3 kg [*SD]) were studied. The dogs were intubated with a no. 9 endotracheal tube and then placed in the supine position. Supplemental doses of pentobarbital sodium were given throughout the experiment to suppress the cornea1 reflex. Pairs of piezoelectric transducers (length transducers) were then inserted into the crural and costal parts of the diaphragm via a midline laparotomy, as previously described.? Bipolar stainless steel hook electrodes were placed in the crural and costal diaphragms close to the transducers to measure the diaphragmatic EMG. The abdominal inciJournalof
Critical
Care, Vol6,
No 3 (September),
1991: pp 136-142
DIAPHRAGMATIC
FUNCTION
DURING
137
CPAP L’ PEEP
sion was then tightly sutured to prevent air leaks and free air was aspirated from the abdominal cavity through a catheter. Two 5-mL balloons connected to PE 200 tubing were positioned for the measurement of transdiaphragmatic pressure. One balloon was positioned in the midesophagus and tilled with 0.5 mL of air, and the abdominal balloon was placed between the cranial surface of the liver and the central tendon of the diaphragm. The abdominal balloon was filled with 1 mL of air. The catheters were connected to a differential pressure transducer (Validyne MP 45 2 100 cm H,O, Engineering Corp, Northridge, CA). The output from this transducer summed gastric and pleural pressure and gave transdiaphragmatic pressure. Airflow was measured by a pneumotachograph (Fleisch #l; Whittaker Medical Manufacturing, Richmond, VA) attached to the endotracheal tube. Volume was obtained by integrating the flow signal obtained from the pneumotachograph. A twoway valve (#2600; Hans-Rudolph Inc, Kansas City, MO) was placed distal to the pneumotachograph. This valve separated inspiratory and expiratory airflow. Airway pressure (Pao) was recorded from a port in the middle chamber of the two-way valve. The raw diaphragmatic EMG was amplified and filtered (Grass model P511; band width, 40 to 3,000 Hz with a 60-Hz notch filter; Grass Instrument Co, Quincy, MA) and then fully rectified and integrated (resistance capacitance filter; time constant, 100 milliseconds). Electromyographic activity was determined by measuring the surface area under the integrated EMG signal. Diaphragmatic length measurements were made by connecting the piezoelectric transducers to a sonomicrometer (Sonomicrometer 120; Triton Technology, San Diego, CA). We have previously described the use of this technique to measure diaphragmatic length.4 The technique has been used extensively for measurement of cardiac dimensions and it can provide accurate measurements of diaphragmatic and other skeletal muscle length change. In brief, one transducer emits an ultrasonic burst that is propagated through the diaphragm to the receiving transducer. Measurement of the transit time is made and, given that the conduction velocity in the muscle is a constant, a calculation of the intervening distance is made. This measurement is made at a frequency of 1,537 Hz. The resultant analog signal can be used to determine the velocity and amount of diaphragmatic shortening. Diaphragmatic length change is expressed as a percentage change from the length at functional residual capacity (%L&. Diaphragmatic length and EMG activity were recorded from both the costal and crural parts of the diaphragm because each part has been shown to have separate innervation and function.’ Continuous positive airway pressure was produced by increasing airflow on the inspired side of the two-way valve and regulated by varying the resistance to aitiow (by partially clamping the expiratoty line) on the expiratory side of the two-way valve. Airway pressure during this maneuver was recorded from the port in the middle chamber of the two-way valve. Airway pressure did not fluctuate by more than 1 cm during inspiration and 2 cm during expiration using this technique. Positive end-expiratory pressure was applied by placing a tube connected to the expiratory side of the two-way valve under varying levels of water. The level of PEEP was again recorded from the port in the two-way
valve. Airway pressure during expiration oscillated in early expiration, but was always within 1 cm of the desired level prior to inhalation, when a regular breathing pattern was established. The duration of inspiration (T,,,) was recorded from points of zero flow on the flow signal. Neural T, was recorded from the crural EMG and termed T,,,,. Neural T, was the time between the initial upward deflection of the integrated EMG to the peak of the integrated EMG.
Protocol Control parameters of the dependent variables were made during spontaneous breathing. The dependent variables included air flow, tidal volume, APdi, costal and crural diaphragmatic EMG, and costal and crural diaphragmatic length change. These animals were randomly assigned to receive either PEEP or CPAP first, then varying levels of positive airway pressure were applied in a random order. The levels of aitway pressure used were 2,6. 10, 14, and 1X cm H,O. Each level of CPAP or PEEP was applied for a period of 4 to 6 minutes to allow stabilization of breathing. Between each series of measurements, a rest period of 5 minutes allowed all signals to return to baseline. The dependent variables were recorded at each level of PEEP or CPAP. The change in end-expiratory lung volume. EELV produced by PEEP, or CPAP was recorded by measuring the amount of air expired at the end of each CPAP or PEEP application. The change in diaphragmatic initial length of the costal and crural diaphragms was also recorded at that time. At the completion of the experiment, the passive relationship between lung volume and Pao was measured. After the administration of pancuronium bromide (0.05 mgikg intravenously), two inflations to total lung capacity (TLC) were performed. Lung volume and Pao were then recorded during a slow inflation to TLC to obtain the passive values. A mean of six breaths was analyzed for each parameter, and the mean values were compared with the control values at the start of the experiment by analysis of variance with repeated measurements. Comparisons between CPAP and PEEP at any given airway pressure were made by Student’s paired c-test. Significance was accepted at a level off’ .C .05. Values reported are mean ? standard error unless otherwise
stated.
RESULTS
In Table 1, a comparison of the breathing patterns between CPAP and PEEP is shown. During both PEEP and CPAP, the expired volume over 1 minute (V,) decreased compared with control at Pao > 10 cm H,O. The decrease in V, was mainly due to a reduction in breathing frequency. In both conditions, respiratory frequency, as expressed by the total duration of inspiration and expiration (TTOT)from the flow tracing, decreased significantly compared with control. This decrease was similar during CPAP and PEEP except at airway pressures of 14 and 18 cm H,O, at which breathing frequency de-
138
ROAD,
Table
1. The Effect
of Continuous
Positive
on Respiratory Airway PreSStIre
Ahway
Timing
Pressure
and Functional
and Positive Residual
End-Expiratory
AND
GRASSINO
Pressure
Capacity 10
6
LEEVERS,
0
2
6.4 2 0.7 6.1 t 0.4
8.2 -c 1 7.6 r 0.8
11.1 + 1.1 9.7 2 1.4
12.7 + 1.3 10.3 f 1.7
14
18
T TOT* CPAP (set) PEEP
13.1 + 2.2 8.3 + 1.6
13.4 k 2.6 8.6 k 1.7
Lt CPAP (set) PEEP T, EM* CPAP (set) PEEP AEELVS
1.6 + 0.1
1.64 + 0.1
1.72 + 0.16
1.55 2 0.1
1.48 + 0.1
1.49 2 0.11
1.64 + 0.17 1.41 2 0.13
1.6 + 0.17 1.3 + 0.13
1.43 -c 0.14 1.19 + 0.12
1.6 2 0.1 1.55 + 0.08
1.5 * 0.1 1.62 + 0.1
1.5 2 0.15 1.80 + 0.1
1.5 k 0.15 1.81 + 0.1
1.3 + 0.14 1.75 2 0.17
1.2 5 0.13 1.7 r 0.18
-
77 2 23
138 2 24
148 + 23
212 2 31
319 + 58
-
42 r 6
104 + 16
54+
206 ‘- 34
153 r 16 409 2 63
181 + 27 623 2 71
223 + 28 850 + 96
2.1 -c 1.0 2.3 2 1.0
352
CPAP (ml) PEEP ALung Vol(l
0
IO
V, CPAP (L/min) PEEP
3.9 + 2.1 3.9 2 1.4
*T,,, (set) is the total duration flow signal.
3.3 t 1.7 3.2 2 1.2
of the respiratory
cycle
2.6 L 1.4 2.4 +- 1.0 and increased
compared
with control
2 k 0.6
2 -t 0.6 2.8 -c 2.3
at Pao of 26 cm H,O measured
from
the
tT,, (set) is the duration of inspiration measured between points of zero airflow. of inspiration measured from the crural diaphragmatic EMG (see Methods for further ST, EMOor neural T, (set) is the duration description). T, IMG was greater during PEEP than control at Pao 2 6 cm H,O and less than control at CPAP 14 and 18 cm H,O. T,, was greater
than T, EMGat all levels
of PEEP > 2 cm H,O (paired
f-test).
§AEELV (m) is the change the two conditions.
in lung volume
produced
by the various
/JALung Vol is the change
in lung volume
produced
by passive
creased more during CPAP than PEEP (paired t-test). However, minute ventilation was similar in the two conditions at all airway pressures because tidal volume (V,) increased significantly during CPAP compared with PEEP (Fig 1). Tidal volume was reduced compared with control at a PEEP of 18 cm H,O, but was similar to control at all other levels of PEEP. Tidal volume did not change with respect to control during CPAP.
3001 0
2
6
Airway Pressure
10
14
16
(cm H20)
Fig 1. Tidal volume is shown as a function of increasing airway pressure (Pao) during PEEP (solid circles) or CPAP (open circles). Tidal volume was greater with CPAP compared with PEEP at all levels of Pao 1 6 cm H,O (paired t-test).
levels
inflation
of CPAP or PEEP. There after muscle
was no difference
in AEELV between
relaxant.
T,, as measured by airflow (T,,), was unchanged at the various levels of Pao during PEEP and CPAP (Table 1). Although T,+ was unchanged during CPAP, neural T,,,, progressively decreased from 1.6 to 1.2 seconds with increasing CPAP levels. Interestingly, during PEEP T,+ was less than T,,,, and during CPAP TIYwas greater than T,,,, (Table 1). As has been previously reported,6 the increase in EELV at each airway pressure during PEEP or CPAP was similar in this preparation. The increases in EELV were relatively small (Table 1); indeed, compared with the passive values at the same airway pressure, the lung volume change was markedly attenuated. We measured EELV in relation to the passive lung volume change as a global reflection of expiratory muscle activity. It has been previously shown3X7,Rthat abdominal and rib cage expiratory muscles are strongly activated by PEEP or CPAP in anesthetized animals. The difference between EELV during positive pressure breathing and the passive change in lung volume is a global measurement of this activity. Since EELV increased with CPAP and PEEP,
DIAPHRAGMATIC
FUNCTION
DURING
Table 2. The Effect
Airway PESSUR?
CPAP L’ PEEP
139
of Continuous Positive Airway on the Coastal end Crural
0
2
Pressure and Positive Parts of the Diaphragm 6
End-Expiratory
Pressure
10
14
18 _____
Crural
L,,,”
CPAP PEEP
-
2.4 ? 0.6 1.5 c 0.3
4.7 ;t 1.2 3.8 k 0.9
6.7 2 1.8 6.9 -f 1.5
Costal L,,,” CPAP
-
0.6 -t 0.3
1.6 2 0.7
2.2 c 1.0
1.4
6.8 1 2.3
-
0.5 2 0.1
1.3 + 0.5
2.6 2 0.8
3.8 2 1.3
4.9 I 1.7
20 + 1.6 20 f 1.5
21 t 1.6 21 k 1.5
23 + 2.0 23 +- 1.6
24 c 2.5 24 k 2.1
23 2 2.7 24 f 2.6
23 t 3.0 24 + 2.7
7.1 f
1.4
8.0 t 1.5
9.7 t
1.8
IO % 1.9
11.3 k 2.1
12 2 2.2
8.1 f
1.5
8.7 + 1.7
9.5 t 2.0
9.7 + 2.0
10.5 -e 2.0
10.8 + 2.1
PEEP Crural %L,,,t CPAP PEEP Costal %LFBct CPAP PEEP
*LK is the initial resting length of the diaphragm. the resting length before CPAP or PEEP. t%L,,, difference
is the tidal inspiratory between L,,, or %L,,,
diaphragm.
The values
shortening of the diaphragm at the various levels of airway
Both CPAP and PEEP produced
an increase
are expressed
in costal
24,
1
OJ
, 0
2
6 10 Airway Pressure (cm
14
4i
decrease
in resting
13.2 i 3 11.1 k 2.4
length
compared
with
measured during the various levels of CPAP and PEEP. There was no pressure during the two conditions of CPAP or PEEP except for the costal
diaphragmatic resting length (L& decreased in both the costal and crural parts of the diaphragm. The decrease in length was similar during PEEP and CPAP (Table 2). Tidal diaphragmatic shortening in the crural part of the diaphragm did not change during CPAP or PEEP, as shown in Table 2. Tidal costal diaphragmatic shortening increased during CPAP and PEEP compared with control. Crural shortening during resting breathing was greater than costal shortening, as has previously been reported.’ However, shortening did not increase in the crural diaphragm during PEEP or CPAP at any airway pressure. In Fig 2, swings in transdiaphragmatic pressure (APdi) are shown as a function of CPAP
4
as a percentage
10.7 _t 2.6 8.5 + 2.2
18
HZO)
Fig 2. Swings in transdiaphragmatic pressure (APdi) ere shown as a function of increasing airway pressure (Pao) during PEEP (solid circles) or CPAP (open circles). APdi was unchanged during CPAP but increased with PEEP b 6 cm H,O (analysis of variance).
tidal shortening
at Pao t 6 cm H,O (ANOVA).
and PEEP. At an airway pressure of 6 cm H,O and greater, APdi was significantly increased during PEEP compared with CPAP. APdi was unaltered during CPAP; however, during PEEP, APdi increased from 9.9 to 21 cm H,O (Fig 2). The duration of inspiration was different between CPAP and PEEP. lnspiratory duration as measured by airflow was greater than neural T, during CPAP but was less than neural T, during PEEP (Table 1). At a PEEP of 6 cm H,O and higher, T,,,, was greater than T,+ (paired t-test), whereas T,, was greater than TIFMCiat CPAP levels of 14 and 18 cm H,O (paired t-test). Accordingly, with PEEP. a component of active inspiration (associated with diaphragmatic EMG and shortening) was not associated with airflow; on the other hand, during CPAP, there was inspiratory airflow without inspiratory EMG activity. With PEEP, active diaphragmatic contraction was not followed by airflow until later in the breath, whereas with CPAP, diaphragmatic contraction was always associated with airflow and, indeed, inspiratory flow preceded active diaphragmatic contraction. In Fig 3, EMG activation of the diaphragm is shown as a function of increasing airway pressure during PEEP and CPAP. At all levels of airway pressure of 6 cm H,O and above, the degree of diaphragmatic activation in both parts of the diaphragm was distinctly greater during PEEP than CPAP. There was an approximate doubling of the diaphragmatic EMG in both the
140
ROAD,
zooT
’
Fig 3. The area under the integrated costal (triangles) and crural (circles) EMG is shown as a function of Pao during CPAP (open symbols) and PEEP (closed symbols). Electromyogram activity increased in both parts of the diaphragm during PEEP at pressures 2 6 cm H,O, but decreased compared with control at CPAP levels of 14 and 18 cm H,O (analysis of variance).
costal and crural parts during PEEP, whereas during CPAP, diaphragmatic activation decreased by 30%. DISCUSSION
We have shown that lung volume increases to the same degree during PEEP and CPAP at any given airway pressure in the anesthetized dog. By inference, this suggests that the degree of expiratory muscle activation counteracting the increase in EELV was equal in the two conditions. Similar expiratory muscle activity and EELV during PEEP and CPAP lead to a similar initial length in both parts of the diaphragm at the onset of inspiration (Table 2). Diaphragmatic contraction, however, as assessed by the degree of neural activation and the pressure (APdi) developed, was significantly greater during PEEP than CPAP. Diaphragmatic shortening was similar with PEEP and CPAP because a considerable degree of the shortening was passive (occurred before EMG onset) during CPAP. The expiratory muscles were actively recruited in the two conditions. This expiratory muscle activation might have distorted the chest wall. For example, greater recruitment of the rib cage expiratory muscles during PEEP or CPAP would have placed the diaphragm at a shorter initial length, which could compromise its contraction during the subsequent inspiration. However, diaphragmatic initial resting length, as assessed by sonomicrometry, was similar at equivalent levels of CPAP and PEEP.
LEEVERS,
AND
GRASSINO
Thus, the chest wall configuration was similar in this aspect in the two conditions. Activation of the expiratory muscles during lung inflation is one mechanism whereby tidal volume may be preserved during hyperinflation. Expiratory muscle activity is known to reduce lung volume below the volume expected from the passive pressure volume relationship of the respiratory system. Therefore, if there had been no expiratory muscle activity, EELV would have increased substantially more than in our study (see Table 2). As Agostoni’ predicted, breathing under such circumstances can become an expiratory act so that relaxation of expiratory activity during CPAP can lead to passive recoil of the chest wall and, hence, passive inspiration. Accordingly, active inspiration (EMG activity), as we have shown (see Fig 3) may actually decrease during CPAP and still preserve V,. Indeed, T,,,, decreased, whereas T,v was unchanged, showing that a significant component of inspiration was passive. Therefore, breathing with CPAP produced a component of both diaphragmatic shortening and tidal volume that was passive. When expiratory activity increased, the result was a marked decrease in inspiratory activity (EMG) and preservation of V, with CPAP. When PEEP was applied, the respiratory response was different. Tidal volume was less during PEEP compared with CPAP at Pao levels greater than 6 cm H,O. Compared with CPAP breathing, PEEP breathing was not associated with T,, greater than T,,,,. On the contrary, T,,,, was greater than T,, (Table 1). At the end of expiratory activity, airway pressure is positive during PEEP. Prior to the onset of airflow, airway pressure must become negative. Therefore, at the end of an expiration, there is no component of passive inspiration. Indeed, a significant degree of diaphragmatic EMG, shortening, and APdi must occur to restore the level Pao to 0 before airflow can begin. Therefore, although tidal volume was defended with PEEP, to do so, inspiratory muscle activity, as reflected in the EMG and APdi, had to increase substantially. The costal and crural diaphragms responded differently in terms of length changes (%L& in the two conditions. Although crural EMG activity increased compared with the increase in
DIAPHRAGMATIC
FUNCTION
DURING
141
CPAP V PEEP
costal EMG during PEEP, tidal shortening increased only in the costal diaphragm. Costa1 tidal shortening also increased during CPAP. Increased shortening of the costal diaphragm during CPAP may be explained by our knowledge of passive diaphragm movement. Since CPAP produced a component of passive inspiration, this component may influence costal shortening more than crural, since the costal diaphragm is less dependent (ie, more ventral). Froese and Bryan” have shown that during passive inflation in the supine position, the ventral aspect of the diaphragm is displaced more than the dorsal aspect in humans. However, it is unclear why shortening increased during PEEP; possibly, the relatively smaller costal tidal shortening during control allowed greater recruitment in terms of muscle shortening when the EMG increased. It is clear, nonetheless, that the costal part of the diaphragm contracted less during control and responded differently to both CPAP and PEEP, which provides support for the concept that these two parts of the diaphragm may be separate on a functional basis. As Gherini et al’ have shown, use of the expiratory muscles during CPAP should reduce the inspiratory work of breathing. Our results show that when expiratory muscle activity is comparable between CPAP and PEEP, there is a marked reduction in inspiratory activity with CPAP. Thus, we predict a decrease in the overall oxygen cost of breathing with CPAP compared with PEEP. Indeed, in conditions in which expiratory muscle function is preserved and the inspiratory muscles weakened, CPAP may be used to reduce the inspiratoty work of breathing. However, such an effect will be dependent on expiratory muscle recruitment. Without such expiratory muscle recruitment, FRC increases and diaphragmatic length decreases. It has been previously shown that when diaphragm length decreases, its shortening also decreases” and its ability to produce pressure”.” and volume” is sharply diminished.
Lung inflation produced by CPAP can therefore place the diaphragm and the other inspiratory muscles at a disadvantage. In humans, this effect of increase in lung volume and decrease in diaphragm length is believed to be offset by an augmentation of inspiratory EMG activity. This would imply that to maintain tidal volume in the face of an increase in lung volume. an augmentation of inspiratory muscle activity is required. It is clear, therefore, that for CPAP to reduce the inspiratory work of breathing, augmentation of expiratory muscle activity is a prerequisite. That is, for CPAP to produce an inspiratory assist, expiratory muscle activation must drive the respiratory system below the relaxed position so that during the subsequent inspiration the positive inspiratoty pressure can, with chest wall recoil, produce an assist to inspiration. Without expiratory muscle recruitment CPAP produces no such assistance to inspiration. Anesthetized animals recruit their expiratoty muscles vigorously in response to this load. Human subjects can also recruit then expiratory muscles during positive pressure breathing, however this effect is greater in the upright than in the supine position and occurs at higher levels of CPAP.14 Therefore, CPAP may not be relied on to uniformly recruit the expiratory muscles in humans. The inspiratory activation of the diaphragm needed to preserve tidal volume was clearly less during CPAP. However, this reduction was dependant on expiratory muscle recruitment. Although PEEP is an expiratory threshold load, it clearly places a load on the inspiratory muscles as well as the expiratory muscles. Not only is the diaphragm shortened by the increase in FRC, but also a considerable part of diaphragmatic contraction is not accompanied by airflow and is hence ineffective. ACKNOWLEDGMENT The authors thank typing and preparation
Bernice Robillard for her assistance of the manuscript.
in
REFERENCES I. Gherini S, Peters RM, Virgilio RW: Mechanical on the lungs and work of breathing with positive expiratory pressure and continuous positive airway sure. Clin Invest 76:251-2X, 1979
work endpres-
2. Schlobohm RM, Falltrick RT, Quan SF, et al: Lung volumes, mechanics and oxygenation during spontaneous positive-pressure ventilation. Anesthesiology SS:416-422. 1981
142
3. Road JD, Leevers AM: Inspiratory and expiratoty muscle function during continuous positive airway pressure in dogs. J Appl Physiol68:1092-1100, 1990 4. Newman S, Road JD, Bellemare F, et al: Respiratory muscle length measured by sonomicrometry. J Appl Physiol 56:753-763,1984 5. De Troyer A, Sampson M, Sigrist S, et al: The diaphragm: Two muscles. Science 213:237-238,198l 6. Layon J, Banner MJ, Jaeger MJ, et al: Continuous positive airway pressure and expiratory positive airway pressure increase function residual capacity equivalently. Chest 89517-521, 1986 7. Bishop B: Reflex control of abdominal muscles during positive pressure breathing. J Appl Physiol 19:224-234,1964 8. Bishop B, Bachofen H: Vagal control of ventilation and respiratory muscles during elevated pressures in the cat. J Appl Physiol32:103-112, 1972 9. Agostoni E: Diaphragm activity and thoracoabdomi-
ROAD,
LEEVERS,
AND
GRASSINO
nal mechanics during positive pressure breathing. J Appl Physiol17:215-220,1962 10. Froese AB, Bryan AC: Effects of anesthesia and paralysis on diaphragmatic mechanics in man. Anesthesiology 41:242-255,1974 11. Road JD, Leevers AM: Effect of lung inflation on diaphragmatic shortening. J Appl Physiol 65:2383-2389, 1988 12. Road J, Newman S, Derenne JP, et al: In vivo length-force relationship of canine diaphragm. J Appl Physiol60:63-70,1986 13. Kim M, Druz WS, Danon J, et al: Mechanics of the canine diaphragm. J Appl Physiol41:369-382,1976 14. Urbscheit N, Bishop B, Bachofen H: Immediate effects of continuous positive pressure breathing on abdominal expiratory activity, minute ventilation, and end-tidal PCO, of conscious man. Phys Ther 53:258-265,1973
and A. Grassino
Continuous positive airway pressure (CPAP) and positive end-expiratory pressure (PEEP) both increase lung volume and hence may compromise diaphragm function. However, the effects of these two positive airway pressure modalities on inspiratory work of breathing are conflicting. In this study, we compared the effect of CPAP versus PEEP on diaphragm function in spontaneously breathing anesthetized dogs. Eight sodium pentobarbital-anesthetized dogs were randomly exposed to various levels of CPAP and PEEP. Measurements of diaphragmatic shortening, transdiaphragmatic pressure swings, and diaphragmatic electromyogram (EMG) were made. The change in lung volume and diaphragm length was similar at equivalent airway pressures during PEEP or CPAP. Therefore, expiratory muscle recruitment in the two conditions was equivalent.
However, tidal diaphragmatic EMG and transdiaphragmatic pressure swings increased markedly during PEEP compared with CPAP. At a PEEP of 18 cm H,O, crural and costal EMG activities were 185% r 16% and 163% ? 8% of control, respectively, whereas during CPAP the EMG activity was 66% 2 11% of control for both the costal and the crural diaphragms (&SE). During PEEP, the duration of neural inspiration (T,,,) was greater than the duration of inspiration as measured by airflow (TJ. On the other hand, during CPAP, T ,EMo was less than T,. We conclude that although expiratory muscle recruitment is comparable and tidal volume greater during CPAP, the inspiratory activation of the diaphragm decreases with CPAP but increases markedly with PEEP. Copyright o 1991 by W. B. Saunders Company
B
PEEP and CPAP and the work of breathing in these two studies, however, have been complicated by the different changes in EELV produced by the two modalities. We therefore wished to define the effect of these two positive airway pressure modalities on the function of the main inspiratory muscle, the diaphragm. The work of breathing as estimated by pressure and volume ignores the work done by the respiratory muscles on distortion of the chest wall. Direct measurements of respiratory muscle contraction can therefore provide information not available from analysis of pressure and volume change. Therefore, using the techniques of sonomicrometry4 and electromyography (EMG) to directly measure muscle contraction and activation, we compared the effects of PEEP and CPAP on the diaphragm in anesthetized dogs.
OTH POSITIVE end-expiratory pressure (PEEP) and continuous positive airway pressure (CPAP) have been used to increase functional residual capacity (FRC) and improve gas exchange. Both positive airway pressure modalities may increase the work of breathing; however, the increase has been shown to be less with CPAP compared with PEEP.’ The main mechanism leading to a relative reduction in work of breathing during CPAP was attributed to the reduction in inspiratory work at the expense of an increase in expiratory work. Increased expiratory work prevents the expected increase in end-expiratory lung volume (EELV) and, with CPAP but not with PEEP, allows an inspiratory assist. Schlobohm et al,’ however, reported that oxygen consumption was not different with either modality. The inspiratory assist produced by CPAP3 should have reduced the inspiratory work of breathing and hence reduced inspiratory muscle work and oxygen consumption. Comparisons between From the Department of Medicine, University of British Columbia, Vancouver, BC; and Notre Dame Hospital and The Meakins-Christie Laboratories, Montreal, Quebec. Received December 19, 1990; May 14,199l. Dr Road was a scholar of the British Columbia Health Care Research Foundation. Address reprint requests to J.D. Road, MD, Department of Medicine, University Hospital (UBC), 221 I Wesbrook Mall, Vancouver, BC V6T 1 W5. Copyright 0 1991 by W.B. Saunders Company 0883-9441191/0603-0008$05.00/O 136
METHODS To compare PEEP versus CPAP, eight pentobarbital sodium (25 to 30 mgikg intravenously) anesthetized dogs (mean weight, 24 f 3 kg [*SD]) were studied. The dogs were intubated with a no. 9 endotracheal tube and then placed in the supine position. Supplemental doses of pentobarbital sodium were given throughout the experiment to suppress the cornea1 reflex. Pairs of piezoelectric transducers (length transducers) were then inserted into the crural and costal parts of the diaphragm via a midline laparotomy, as previously described.? Bipolar stainless steel hook electrodes were placed in the crural and costal diaphragms close to the transducers to measure the diaphragmatic EMG. The abdominal inciJournalof
Critical
Care, Vol6,
No 3 (September),
1991: pp 136-142
DIAPHRAGMATIC
FUNCTION
DURING
137
CPAP L’ PEEP
sion was then tightly sutured to prevent air leaks and free air was aspirated from the abdominal cavity through a catheter. Two 5-mL balloons connected to PE 200 tubing were positioned for the measurement of transdiaphragmatic pressure. One balloon was positioned in the midesophagus and tilled with 0.5 mL of air, and the abdominal balloon was placed between the cranial surface of the liver and the central tendon of the diaphragm. The abdominal balloon was filled with 1 mL of air. The catheters were connected to a differential pressure transducer (Validyne MP 45 2 100 cm H,O, Engineering Corp, Northridge, CA). The output from this transducer summed gastric and pleural pressure and gave transdiaphragmatic pressure. Airflow was measured by a pneumotachograph (Fleisch #l; Whittaker Medical Manufacturing, Richmond, VA) attached to the endotracheal tube. Volume was obtained by integrating the flow signal obtained from the pneumotachograph. A twoway valve (#2600; Hans-Rudolph Inc, Kansas City, MO) was placed distal to the pneumotachograph. This valve separated inspiratory and expiratory airflow. Airway pressure (Pao) was recorded from a port in the middle chamber of the two-way valve. The raw diaphragmatic EMG was amplified and filtered (Grass model P511; band width, 40 to 3,000 Hz with a 60-Hz notch filter; Grass Instrument Co, Quincy, MA) and then fully rectified and integrated (resistance capacitance filter; time constant, 100 milliseconds). Electromyographic activity was determined by measuring the surface area under the integrated EMG signal. Diaphragmatic length measurements were made by connecting the piezoelectric transducers to a sonomicrometer (Sonomicrometer 120; Triton Technology, San Diego, CA). We have previously described the use of this technique to measure diaphragmatic length.4 The technique has been used extensively for measurement of cardiac dimensions and it can provide accurate measurements of diaphragmatic and other skeletal muscle length change. In brief, one transducer emits an ultrasonic burst that is propagated through the diaphragm to the receiving transducer. Measurement of the transit time is made and, given that the conduction velocity in the muscle is a constant, a calculation of the intervening distance is made. This measurement is made at a frequency of 1,537 Hz. The resultant analog signal can be used to determine the velocity and amount of diaphragmatic shortening. Diaphragmatic length change is expressed as a percentage change from the length at functional residual capacity (%L&. Diaphragmatic length and EMG activity were recorded from both the costal and crural parts of the diaphragm because each part has been shown to have separate innervation and function.’ Continuous positive airway pressure was produced by increasing airflow on the inspired side of the two-way valve and regulated by varying the resistance to aitiow (by partially clamping the expiratoty line) on the expiratory side of the two-way valve. Airway pressure during this maneuver was recorded from the port in the middle chamber of the two-way valve. Airway pressure did not fluctuate by more than 1 cm during inspiration and 2 cm during expiration using this technique. Positive end-expiratory pressure was applied by placing a tube connected to the expiratory side of the two-way valve under varying levels of water. The level of PEEP was again recorded from the port in the two-way
valve. Airway pressure during expiration oscillated in early expiration, but was always within 1 cm of the desired level prior to inhalation, when a regular breathing pattern was established. The duration of inspiration (T,,,) was recorded from points of zero flow on the flow signal. Neural T, was recorded from the crural EMG and termed T,,,,. Neural T, was the time between the initial upward deflection of the integrated EMG to the peak of the integrated EMG.
Protocol Control parameters of the dependent variables were made during spontaneous breathing. The dependent variables included air flow, tidal volume, APdi, costal and crural diaphragmatic EMG, and costal and crural diaphragmatic length change. These animals were randomly assigned to receive either PEEP or CPAP first, then varying levels of positive airway pressure were applied in a random order. The levels of aitway pressure used were 2,6. 10, 14, and 1X cm H,O. Each level of CPAP or PEEP was applied for a period of 4 to 6 minutes to allow stabilization of breathing. Between each series of measurements, a rest period of 5 minutes allowed all signals to return to baseline. The dependent variables were recorded at each level of PEEP or CPAP. The change in end-expiratory lung volume. EELV produced by PEEP, or CPAP was recorded by measuring the amount of air expired at the end of each CPAP or PEEP application. The change in diaphragmatic initial length of the costal and crural diaphragms was also recorded at that time. At the completion of the experiment, the passive relationship between lung volume and Pao was measured. After the administration of pancuronium bromide (0.05 mgikg intravenously), two inflations to total lung capacity (TLC) were performed. Lung volume and Pao were then recorded during a slow inflation to TLC to obtain the passive values. A mean of six breaths was analyzed for each parameter, and the mean values were compared with the control values at the start of the experiment by analysis of variance with repeated measurements. Comparisons between CPAP and PEEP at any given airway pressure were made by Student’s paired c-test. Significance was accepted at a level off’ .C .05. Values reported are mean ? standard error unless otherwise
stated.
RESULTS
In Table 1, a comparison of the breathing patterns between CPAP and PEEP is shown. During both PEEP and CPAP, the expired volume over 1 minute (V,) decreased compared with control at Pao > 10 cm H,O. The decrease in V, was mainly due to a reduction in breathing frequency. In both conditions, respiratory frequency, as expressed by the total duration of inspiration and expiration (TTOT)from the flow tracing, decreased significantly compared with control. This decrease was similar during CPAP and PEEP except at airway pressures of 14 and 18 cm H,O, at which breathing frequency de-
138
ROAD,
Table
1. The Effect
of Continuous
Positive
on Respiratory Airway PreSStIre
Ahway
Timing
Pressure
and Functional
and Positive Residual
End-Expiratory
AND
GRASSINO
Pressure
Capacity 10
6
LEEVERS,
0
2
6.4 2 0.7 6.1 t 0.4
8.2 -c 1 7.6 r 0.8
11.1 + 1.1 9.7 2 1.4
12.7 + 1.3 10.3 f 1.7
14
18
T TOT* CPAP (set) PEEP
13.1 + 2.2 8.3 + 1.6
13.4 k 2.6 8.6 k 1.7
Lt CPAP (set) PEEP T, EM* CPAP (set) PEEP AEELVS
1.6 + 0.1
1.64 + 0.1
1.72 + 0.16
1.55 2 0.1
1.48 + 0.1
1.49 2 0.11
1.64 + 0.17 1.41 2 0.13
1.6 + 0.17 1.3 + 0.13
1.43 -c 0.14 1.19 + 0.12
1.6 2 0.1 1.55 + 0.08
1.5 * 0.1 1.62 + 0.1
1.5 2 0.15 1.80 + 0.1
1.5 k 0.15 1.81 + 0.1
1.3 + 0.14 1.75 2 0.17
1.2 5 0.13 1.7 r 0.18
-
77 2 23
138 2 24
148 + 23
212 2 31
319 + 58
-
42 r 6
104 + 16
54+
206 ‘- 34
153 r 16 409 2 63
181 + 27 623 2 71
223 + 28 850 + 96
2.1 -c 1.0 2.3 2 1.0
352
CPAP (ml) PEEP ALung Vol(l
0
IO
V, CPAP (L/min) PEEP
3.9 + 2.1 3.9 2 1.4
*T,,, (set) is the total duration flow signal.
3.3 t 1.7 3.2 2 1.2
of the respiratory
cycle
2.6 L 1.4 2.4 +- 1.0 and increased
compared
with control
2 k 0.6
2 -t 0.6 2.8 -c 2.3
at Pao of 26 cm H,O measured
from
the
tT,, (set) is the duration of inspiration measured between points of zero airflow. of inspiration measured from the crural diaphragmatic EMG (see Methods for further ST, EMOor neural T, (set) is the duration description). T, IMG was greater during PEEP than control at Pao 2 6 cm H,O and less than control at CPAP 14 and 18 cm H,O. T,, was greater
than T, EMGat all levels
of PEEP > 2 cm H,O (paired
f-test).
§AEELV (m) is the change the two conditions.
in lung volume
produced
by the various
/JALung Vol is the change
in lung volume
produced
by passive
creased more during CPAP than PEEP (paired t-test). However, minute ventilation was similar in the two conditions at all airway pressures because tidal volume (V,) increased significantly during CPAP compared with PEEP (Fig 1). Tidal volume was reduced compared with control at a PEEP of 18 cm H,O, but was similar to control at all other levels of PEEP. Tidal volume did not change with respect to control during CPAP.
3001 0
2
6
Airway Pressure
10
14
16
(cm H20)
Fig 1. Tidal volume is shown as a function of increasing airway pressure (Pao) during PEEP (solid circles) or CPAP (open circles). Tidal volume was greater with CPAP compared with PEEP at all levels of Pao 1 6 cm H,O (paired t-test).
levels
inflation
of CPAP or PEEP. There after muscle
was no difference
in AEELV between
relaxant.
T,, as measured by airflow (T,,), was unchanged at the various levels of Pao during PEEP and CPAP (Table 1). Although T,+ was unchanged during CPAP, neural T,,,, progressively decreased from 1.6 to 1.2 seconds with increasing CPAP levels. Interestingly, during PEEP T,+ was less than T,,,, and during CPAP TIYwas greater than T,,,, (Table 1). As has been previously reported,6 the increase in EELV at each airway pressure during PEEP or CPAP was similar in this preparation. The increases in EELV were relatively small (Table 1); indeed, compared with the passive values at the same airway pressure, the lung volume change was markedly attenuated. We measured EELV in relation to the passive lung volume change as a global reflection of expiratory muscle activity. It has been previously shown3X7,Rthat abdominal and rib cage expiratory muscles are strongly activated by PEEP or CPAP in anesthetized animals. The difference between EELV during positive pressure breathing and the passive change in lung volume is a global measurement of this activity. Since EELV increased with CPAP and PEEP,
DIAPHRAGMATIC
FUNCTION
DURING
Table 2. The Effect
Airway PESSUR?
CPAP L’ PEEP
139
of Continuous Positive Airway on the Coastal end Crural
0
2
Pressure and Positive Parts of the Diaphragm 6
End-Expiratory
Pressure
10
14
18 _____
Crural
L,,,”
CPAP PEEP
-
2.4 ? 0.6 1.5 c 0.3
4.7 ;t 1.2 3.8 k 0.9
6.7 2 1.8 6.9 -f 1.5
Costal L,,,” CPAP
-
0.6 -t 0.3
1.6 2 0.7
2.2 c 1.0
1.4
6.8 1 2.3
-
0.5 2 0.1
1.3 + 0.5
2.6 2 0.8
3.8 2 1.3
4.9 I 1.7
20 + 1.6 20 f 1.5
21 t 1.6 21 k 1.5
23 + 2.0 23 +- 1.6
24 c 2.5 24 k 2.1
23 2 2.7 24 f 2.6
23 t 3.0 24 + 2.7
7.1 f
1.4
8.0 t 1.5
9.7 t
1.8
IO % 1.9
11.3 k 2.1
12 2 2.2
8.1 f
1.5
8.7 + 1.7
9.5 t 2.0
9.7 + 2.0
10.5 -e 2.0
10.8 + 2.1
PEEP Crural %L,,,t CPAP PEEP Costal %LFBct CPAP PEEP
*LK is the initial resting length of the diaphragm. the resting length before CPAP or PEEP. t%L,,, difference
is the tidal inspiratory between L,,, or %L,,,
diaphragm.
The values
shortening of the diaphragm at the various levels of airway
Both CPAP and PEEP produced
an increase
are expressed
in costal
24,
1
OJ
, 0
2
6 10 Airway Pressure (cm
14
4i
decrease
in resting
13.2 i 3 11.1 k 2.4
length
compared
with
measured during the various levels of CPAP and PEEP. There was no pressure during the two conditions of CPAP or PEEP except for the costal
diaphragmatic resting length (L& decreased in both the costal and crural parts of the diaphragm. The decrease in length was similar during PEEP and CPAP (Table 2). Tidal diaphragmatic shortening in the crural part of the diaphragm did not change during CPAP or PEEP, as shown in Table 2. Tidal costal diaphragmatic shortening increased during CPAP and PEEP compared with control. Crural shortening during resting breathing was greater than costal shortening, as has previously been reported.’ However, shortening did not increase in the crural diaphragm during PEEP or CPAP at any airway pressure. In Fig 2, swings in transdiaphragmatic pressure (APdi) are shown as a function of CPAP
4
as a percentage
10.7 _t 2.6 8.5 + 2.2
18
HZO)
Fig 2. Swings in transdiaphragmatic pressure (APdi) ere shown as a function of increasing airway pressure (Pao) during PEEP (solid circles) or CPAP (open circles). APdi was unchanged during CPAP but increased with PEEP b 6 cm H,O (analysis of variance).
tidal shortening
at Pao t 6 cm H,O (ANOVA).
and PEEP. At an airway pressure of 6 cm H,O and greater, APdi was significantly increased during PEEP compared with CPAP. APdi was unaltered during CPAP; however, during PEEP, APdi increased from 9.9 to 21 cm H,O (Fig 2). The duration of inspiration was different between CPAP and PEEP. lnspiratory duration as measured by airflow was greater than neural T, during CPAP but was less than neural T, during PEEP (Table 1). At a PEEP of 6 cm H,O and higher, T,,,, was greater than T,+ (paired t-test), whereas T,, was greater than TIFMCiat CPAP levels of 14 and 18 cm H,O (paired t-test). Accordingly, with PEEP. a component of active inspiration (associated with diaphragmatic EMG and shortening) was not associated with airflow; on the other hand, during CPAP, there was inspiratory airflow without inspiratory EMG activity. With PEEP, active diaphragmatic contraction was not followed by airflow until later in the breath, whereas with CPAP, diaphragmatic contraction was always associated with airflow and, indeed, inspiratory flow preceded active diaphragmatic contraction. In Fig 3, EMG activation of the diaphragm is shown as a function of increasing airway pressure during PEEP and CPAP. At all levels of airway pressure of 6 cm H,O and above, the degree of diaphragmatic activation in both parts of the diaphragm was distinctly greater during PEEP than CPAP. There was an approximate doubling of the diaphragmatic EMG in both the
140
ROAD,
zooT
’
Fig 3. The area under the integrated costal (triangles) and crural (circles) EMG is shown as a function of Pao during CPAP (open symbols) and PEEP (closed symbols). Electromyogram activity increased in both parts of the diaphragm during PEEP at pressures 2 6 cm H,O, but decreased compared with control at CPAP levels of 14 and 18 cm H,O (analysis of variance).
costal and crural parts during PEEP, whereas during CPAP, diaphragmatic activation decreased by 30%. DISCUSSION
We have shown that lung volume increases to the same degree during PEEP and CPAP at any given airway pressure in the anesthetized dog. By inference, this suggests that the degree of expiratory muscle activation counteracting the increase in EELV was equal in the two conditions. Similar expiratory muscle activity and EELV during PEEP and CPAP lead to a similar initial length in both parts of the diaphragm at the onset of inspiration (Table 2). Diaphragmatic contraction, however, as assessed by the degree of neural activation and the pressure (APdi) developed, was significantly greater during PEEP than CPAP. Diaphragmatic shortening was similar with PEEP and CPAP because a considerable degree of the shortening was passive (occurred before EMG onset) during CPAP. The expiratory muscles were actively recruited in the two conditions. This expiratory muscle activation might have distorted the chest wall. For example, greater recruitment of the rib cage expiratory muscles during PEEP or CPAP would have placed the diaphragm at a shorter initial length, which could compromise its contraction during the subsequent inspiration. However, diaphragmatic initial resting length, as assessed by sonomicrometry, was similar at equivalent levels of CPAP and PEEP.
LEEVERS,
AND
GRASSINO
Thus, the chest wall configuration was similar in this aspect in the two conditions. Activation of the expiratory muscles during lung inflation is one mechanism whereby tidal volume may be preserved during hyperinflation. Expiratory muscle activity is known to reduce lung volume below the volume expected from the passive pressure volume relationship of the respiratory system. Therefore, if there had been no expiratory muscle activity, EELV would have increased substantially more than in our study (see Table 2). As Agostoni’ predicted, breathing under such circumstances can become an expiratory act so that relaxation of expiratory activity during CPAP can lead to passive recoil of the chest wall and, hence, passive inspiration. Accordingly, active inspiration (EMG activity), as we have shown (see Fig 3) may actually decrease during CPAP and still preserve V,. Indeed, T,,,, decreased, whereas T,v was unchanged, showing that a significant component of inspiration was passive. Therefore, breathing with CPAP produced a component of both diaphragmatic shortening and tidal volume that was passive. When expiratory activity increased, the result was a marked decrease in inspiratory activity (EMG) and preservation of V, with CPAP. When PEEP was applied, the respiratory response was different. Tidal volume was less during PEEP compared with CPAP at Pao levels greater than 6 cm H,O. Compared with CPAP breathing, PEEP breathing was not associated with T,, greater than T,,,,. On the contrary, T,,,, was greater than T,, (Table 1). At the end of expiratory activity, airway pressure is positive during PEEP. Prior to the onset of airflow, airway pressure must become negative. Therefore, at the end of an expiration, there is no component of passive inspiration. Indeed, a significant degree of diaphragmatic EMG, shortening, and APdi must occur to restore the level Pao to 0 before airflow can begin. Therefore, although tidal volume was defended with PEEP, to do so, inspiratory muscle activity, as reflected in the EMG and APdi, had to increase substantially. The costal and crural diaphragms responded differently in terms of length changes (%L& in the two conditions. Although crural EMG activity increased compared with the increase in
DIAPHRAGMATIC
FUNCTION
DURING
141
CPAP V PEEP
costal EMG during PEEP, tidal shortening increased only in the costal diaphragm. Costa1 tidal shortening also increased during CPAP. Increased shortening of the costal diaphragm during CPAP may be explained by our knowledge of passive diaphragm movement. Since CPAP produced a component of passive inspiration, this component may influence costal shortening more than crural, since the costal diaphragm is less dependent (ie, more ventral). Froese and Bryan” have shown that during passive inflation in the supine position, the ventral aspect of the diaphragm is displaced more than the dorsal aspect in humans. However, it is unclear why shortening increased during PEEP; possibly, the relatively smaller costal tidal shortening during control allowed greater recruitment in terms of muscle shortening when the EMG increased. It is clear, nonetheless, that the costal part of the diaphragm contracted less during control and responded differently to both CPAP and PEEP, which provides support for the concept that these two parts of the diaphragm may be separate on a functional basis. As Gherini et al’ have shown, use of the expiratory muscles during CPAP should reduce the inspiratory work of breathing. Our results show that when expiratory muscle activity is comparable between CPAP and PEEP, there is a marked reduction in inspiratory activity with CPAP. Thus, we predict a decrease in the overall oxygen cost of breathing with CPAP compared with PEEP. Indeed, in conditions in which expiratory muscle function is preserved and the inspiratory muscles weakened, CPAP may be used to reduce the inspiratoty work of breathing. However, such an effect will be dependent on expiratory muscle recruitment. Without such expiratory muscle recruitment, FRC increases and diaphragmatic length decreases. It has been previously shown that when diaphragm length decreases, its shortening also decreases” and its ability to produce pressure”.” and volume” is sharply diminished.
Lung inflation produced by CPAP can therefore place the diaphragm and the other inspiratory muscles at a disadvantage. In humans, this effect of increase in lung volume and decrease in diaphragm length is believed to be offset by an augmentation of inspiratory EMG activity. This would imply that to maintain tidal volume in the face of an increase in lung volume. an augmentation of inspiratory muscle activity is required. It is clear, therefore, that for CPAP to reduce the inspiratory work of breathing, augmentation of expiratory muscle activity is a prerequisite. That is, for CPAP to produce an inspiratory assist, expiratory muscle activation must drive the respiratory system below the relaxed position so that during the subsequent inspiration the positive inspiratoty pressure can, with chest wall recoil, produce an assist to inspiration. Without expiratory muscle recruitment CPAP produces no such assistance to inspiration. Anesthetized animals recruit their expiratoty muscles vigorously in response to this load. Human subjects can also recruit then expiratory muscles during positive pressure breathing, however this effect is greater in the upright than in the supine position and occurs at higher levels of CPAP.14 Therefore, CPAP may not be relied on to uniformly recruit the expiratory muscles in humans. The inspiratory activation of the diaphragm needed to preserve tidal volume was clearly less during CPAP. However, this reduction was dependant on expiratory muscle recruitment. Although PEEP is an expiratory threshold load, it clearly places a load on the inspiratory muscles as well as the expiratory muscles. Not only is the diaphragm shortened by the increase in FRC, but also a considerable part of diaphragmatic contraction is not accompanied by airflow and is hence ineffective. ACKNOWLEDGMENT The authors thank typing and preparation
Bernice Robillard for her assistance of the manuscript.
in
REFERENCES I. Gherini S, Peters RM, Virgilio RW: Mechanical on the lungs and work of breathing with positive expiratory pressure and continuous positive airway sure. Clin Invest 76:251-2X, 1979
work endpres-
2. Schlobohm RM, Falltrick RT, Quan SF, et al: Lung volumes, mechanics and oxygenation during spontaneous positive-pressure ventilation. Anesthesiology SS:416-422. 1981
142
3. Road JD, Leevers AM: Inspiratory and expiratoty muscle function during continuous positive airway pressure in dogs. J Appl Physiol68:1092-1100, 1990 4. Newman S, Road JD, Bellemare F, et al: Respiratory muscle length measured by sonomicrometry. J Appl Physiol 56:753-763,1984 5. De Troyer A, Sampson M, Sigrist S, et al: The diaphragm: Two muscles. Science 213:237-238,198l 6. Layon J, Banner MJ, Jaeger MJ, et al: Continuous positive airway pressure and expiratory positive airway pressure increase function residual capacity equivalently. Chest 89517-521, 1986 7. Bishop B: Reflex control of abdominal muscles during positive pressure breathing. J Appl Physiol 19:224-234,1964 8. Bishop B, Bachofen H: Vagal control of ventilation and respiratory muscles during elevated pressures in the cat. J Appl Physiol32:103-112, 1972 9. Agostoni E: Diaphragm activity and thoracoabdomi-
ROAD,
LEEVERS,
AND
GRASSINO
nal mechanics during positive pressure breathing. J Appl Physiol17:215-220,1962 10. Froese AB, Bryan AC: Effects of anesthesia and paralysis on diaphragmatic mechanics in man. Anesthesiology 41:242-255,1974 11. Road JD, Leevers AM: Effect of lung inflation on diaphragmatic shortening. J Appl Physiol 65:2383-2389, 1988 12. Road J, Newman S, Derenne JP, et al: In vivo length-force relationship of canine diaphragm. J Appl Physiol60:63-70,1986 13. Kim M, Druz WS, Danon J, et al: Mechanics of the canine diaphragm. J Appl Physiol41:369-382,1976 14. Urbscheit N, Bishop B, Bachofen H: Immediate effects of continuous positive pressure breathing on abdominal expiratory activity, minute ventilation, and end-tidal PCO, of conscious man. Phys Ther 53:258-265,1973