Expiratory Muscle Recruitment during Inspiratory Flow-resistive Loading and Exercise1- 3

PETER H. ABBRECHT, KRISHNAN R. RAJAGOPAL, and RICHARD R. KYLE Introduction During strenuous exercise, there is active contraction of the expiratory muscles (1), which decreases end-expiratory lung volume (EELV) and allows the diaphragm to operate at a more advantageous length during inspiration (2).Thus, diaphragmatic contraction is more efficient in meeting the increased ventilatory demands of exercise. This use of expiratory muscles to augment ventilation may delay or prevent the onset of inspiratory muscle fatigue. Abdominal muscle recruitment and decrease in EELV may also occur during inspiratory flowresistive loading (3). Therefore, recruitment of expiratory muscles may be a supportive mechanism that helps the inspiratory muscles meet increased work loads. Little information is available on the effects of combined exercise and inspiratory flow-resistive loading on expiratory muscle recruitment and EELV. There are no previous studies of the time of onset or the magnitude of expiratory muscle recruitment in the setting of combined exercise and inspiratory flow-resistive loading. The combination of exerciseand increased airway resistance may occur during exertion in patients with a variety of clinical conditions, including chronic obstructive pulmonary disease, bronchial asthma, and upper airway obstruction. A better understanding of the pattern of respiratory muscle actions during exercise with flow-resistive loading could lead to interventions having therapeutic benefit. In addition, there are many occupational settings in which workers wear respiratory protective equipment while working at increased metabolic rate. The protective equipment may cause increased resistance to breathing and thus limit the degree of physical effort that the worker can sustain. Accordingly, we examined the respiratory effects of exerciseand various degrees of resistive load to define the normal response of the respiratory muscles to this dual challenge. We studied a group of normal volunteers during prolonged submaximal ex-

SUMMARY Both exerel.. and Inspiratory flow-resistive loading may cau.. recruitment of expiratory muscles. To evaluate the extent of recruitment In combined e.rel.. and fIow.realstlve loading, and to estimate the effect on Inspiratory muscle work, we studied five men, 28 to 39 yr of age, during mild exerel.. with different deg..... of Inspiratory fIow- ..... lstlve loading. Each subJect per· formed four 1·h exerel.. runs at 30% of their maximal oxygen consumption on different dap whl" Inspiring through an external realator of either 1A, 14.5,19.9,or 30.8em H20/&1L. Mouth and eeoph8geal pre..ure, Inspiratory flow rate, and abdominal and rib c.ge motion were recorded continuously. Abdominal exPansion tended to lead and rib cage eXPansion tended to lag the start of Inaplmlon as Judged from the beginning of negative preuure development at the mouth. Th_ time differ· enceslncreased .. resistive load Increased. Plota of abdominal versus rib cage motion also showed Increa.. In ph..e shift, with the abdomen leading the rib cage on Inaplmlon. For all subfecta, the esophageal preaaure at the end of expiration became .... negative .. the resistive load Increased, Indicating that the end..xplratory volume decreased with Increasing resistive load. we conclude that there was Increasing u.. of expiratory muscles .. the resistive load Inc.....ed, and that the Initial eXPansion of the abdomen at high .....Istlve loads represented elastic recoil of structures that had been compreaaed below the volume at FAC by the expiratory musc.... Calculations suggest that, at different resistive load., from 10 to 20% of the work of Inspiration was transferred to the expiratory muscles, and that the power of the Inspiratory muscles approeched a limiting value at high resistive loads. AM REV RESPIR DIS 1111; 144:113-120

ercise while they breathed through different flow-resistive loads. Measurements included mouth and esophageal pressures, inspiratory airflows at the mouth, displacement of the abdomen and rib cage throughout each breath, and FEco2 • From these measurements we determined end-expiratory esophageal pressure and the relationships between the time at which inspiratory airflow begins and the times at which rib cage and abdominal expansion are initiated. In order to better understand the significance of the observed changes in respiratory pattern, we also estimated the contribution of the expiratory muscles to breathing, and the effects of the observed breathing pattern changes on the work of the inspiratory muscles. Methods Informed consent for the study was obtained from five healthy, nonsmoking men 26 to 39 yr of age. Their anthropometric and pulmonary function data are given in table 1. All pulmonary function results were within the normal predicted range. Maximal oxygen uptake (Vo2 max) was determined for each subject in a progressive graded exercisetest using a bicycle ergometer, with the work load incremented by 2S watts/min. To study the ef-

feet of added inspiratory resistance during steady-state exercise,each subject did four experiments on four different days in which he pedaled at 300/0 of his V02 max for I-h periods, with a different inspiratory resistance being added to the breathing circuit during each experiment. The l-h duration for each experiment was arbitrarily chosen to attain steadystate conditions and to permit maximal load adaptation to occur. The subjects breathed through a no. 2700 respiration valve (Hans Rudolph, Kansas City, MO), with a Fleisch pneumotachograph attached to the inlet port. The only resistance in the expiratory limb of the breathing circuit was that of the Hans Rudolph valve, approximately 1.4 em H 20/s/L (Received in originalform October9, 1990 and in revisedform February 11, 1991) 1 From the Departments of Medicine and Physiology, Uniformed Services University of the Health Sciences,Bethesda, Maryland, and the Pulmonary Disease Service, Walter Reed Army Medical Center, Washington, D.C. 1 Supported by Protocol R07648 from the Uniformed Services University of the Health Sciences and by Protocol 170S from the WalterReed Army Medical Center. J Correspondence and requests for reprints should be addressed to Peter H. Abbrecht, Departments of Medicineand Physiology, Uniformed ServicesUniversity of the Health Sciences, Bethesda, Maryland.

113

114

AB8RECH~ ~~L,

TABLE 1 ANTHROPOMETRIC AND PULMONARY FUNCTION DATA FOR EACH SUBJECT Subjects

Age. yr Height, cm Body weight, kg BSA, m2 FEV" L FVC, L FEV1/FVC, % \t02max, mUmin

28 173 80.9 1.95 4.02 4.83 83 3,211

2

3

4

5

26 166 59.1 1.66 3.34 3.61 93 3,500

37 170 87.3 1.99 3.19 4.11 78 3,940

39 188 95.5 2.22 4.44 5.68 78 2,940

28 183 72.2 1.94 4.60 5.45 84 3,030

Definition of abbreviations: eSA • body surface area;

at an expiratory flow of 1 Lis. The measured resistance of the inspiratory limb of the circuit at a flow of 1 Lis was 1.4 em H 20/s/L (Rl). Additional resistors could be added to the inspiratory limb, bringing the total inspiratory resistance to 14.5(R2), 19.9(R3), or 30.6 (R4) em H 20/s/L at a flow of 1 Lis. All external resistors showed linear pressure-flow relationships over the ranges of flows encountered in the studies. On each of the four l-h exercise runs, the subject breathed through a different one of the four inspiratory resistances, with the order of testing determined in random order. Exercise and inspiratory flowresistive loading were started simultaneously. The pneumotachograph flows were integrated to obtain inspired air volumes. Inspired tidal volume was multiplied by respiratory frequency to obtain minute ventilation. Mouth pressure was recorded continuously through a port in the mouthpiece using an MP45 pressure transducer (Validyne Engineering, Northridge, CAl. Esophageal pressure was measured with a similar transducer using a balloon 10 em long attached to a nasogastric tube. End-expiratory esophageal pressure was read at the point where mouth pressure just began to fall below zero. There was no effect of time (P > 0.2) on the endexpiratory esophageal pressures recorded during the steady-state period (last 40 min) of the runs, indicating that the esophageal balloon readings were stable over that period of time. The fraction of CO 2 in respiratory gas obtained from a sampling port in the Hans Rudolph valvewas measured using an LB2 medical gas analyzer (Beckman Instruments, Fullerton, CAl. Abdominal and rib cage motion were monitored using a Respitrace'" respiratory inductive plethysmograph (Ambulatory Monitoring, Ardsley, NY). The abdominal band was placed just below the rib cage, and the thoracic band was placed over the lower middle chest. Maximal negative mouth pressures that could be developed by the subjects were measured before and immediately after each run. All measured variables were recorded on a model 7758strip chart recorder and a model 3968 FM tape recorder (Hewlett-Packard, Cupertino, CAl. End-tidal CO 2 , tidal volume,

v0

2 -

oxygen uptake.

respiratory rate, minute ventilation rate, respiratory duty cycle, and the starting times for outward motion of the rib cage, abdomen, and inspiratory airflow at the mouth during each run wereread from the strip chart recording. The times by which rib cage or abdominal motion led or lagged the beginning of inspiratory airflow at the mouth were calculated by subtracting the time when outward rib cage or abdominal motion started from the time at which inspiratory flow started. There was no effect of time (P > 0.1) on the lead and lag times calculated during the steady state (last 40 min) of the runs, indicating that the Respitrace readings were stable over that period of time. To produce rib cage-abdominal phase loops and estimates of the work of breathing, the analog data for Respitrace displacements, mouth and esophageal pressures, and inspiratory flows were digitized at 200 samples/sl channel using a DAS-500 data system (Keithley Instruments, Cleveland, OH). Because the ventilatory variables for each subject became nearly constant within 15 to 20 min of starting each run, and because of the large amount of data involved in processing each breath, only the last 50 to 70 breaths of each run were analyzed. This corresponded to a 2- to 3-min time period. The beginning of each breath was identified as the point at which mouth pressure started to become negative and at which inspiratory flow began. The beginning of changes in mouth pressure and flow coincided at all loads. In order to obtain waveforms for averaging, the digitized data were interpolated to obtain 512 evenly spaced points per breath for each variable. The Respitrace data for each breath were then normalized by dividing the excursion at each of the 512 points by the maximal excursion during that breath. The data for all breaths wereaveraged for each point. The normalized data were then used to generate loops of rib cage versus abdominal displacement. The start ofinspiratory rib cage and abdominal motion were located on the loops as the points at which direction of displacement changed from inward to outward. To determine the amount of lead or lag for the abdomen and rib cage, the time at which inspiratory motion started on the abdominal-rib cage loop was compared with

AND KYLE

the time in the breath at which mouth pressure started to become negative. Toallow comparisons at different breath durations, time leads or lags were converted to degrees of phase difference using the relationship cp = 360 t/Ttot, where cp is phase shift, t = lead (- ) or lag ( +), and not = total respiratory cycle duration. To determine if motion artifact caused by ergometer pedaling affected the calculated phase lags, fast fourier transforms were done (usingthe Igor program by WaveMetrics,Lake Oswego, OR) for the data from individual runs and for the averaged data for all subjects at a given load. Although there was a significant component of power at a frequency of lOO/min (twice the pedaling frequency) in the individual runs, there was no significant contribution at that frequencyin the averaged data used to generate the curves shown in figure 1. Thus, pedaling artifact did not affect the calculated phase relationships. Changes in EELVat different resistiveloads were estimated for each subject using measured lung compliance and changes in endexpiratory esophageal pressure (pesEE). Dynamic pulmonary compliance was estimated for each breath using the relationship: L\P/V

= L\V/(V

co + R

(1)

where L\P = change in trans pulmonary pressure from that at EELV, L\V = volume change from EELV, V = instantaneous volumetric flow rate, R = pulmonary resistance, and CL = pulmonary compliance. The digitized pressure and flow data were used to form a linear regression of L\P/V on L\V/V. The r 2 for the regressions averaged 0.999, indicating that the model of equation 1 fits the data well. The reciprocal of the slope

1.0

1.0

R2

R1

0.5

I

0.5

/ II

I

o

0 III IV

o

0.5

1.0

0

0.5

1.0

o

0.5

1.0

1.0 R3

0.5

0.5

o

o

0.5

1.0

Abdominal Excursion Fig. 1. Normalized rib cage versus abdominal excursion loops obtained for the four different inspiratory resistances. Each loop is the averageacross five subjectsfor all breathsduring the last5 min of the run using the algorithm described by Abbrecht and colleagues (25).Thefourquadrantsareidentifiedby Romannumerals I to l'l

EXPIRATORY MUSCLE RECRUITMENT DURING INSPIRATORY RESISTIVE LOADING AND EXERCISE

of the regressionline for each breath wastaken as the average compliance of the lungs for that breath. For each subject, the values of PesEEmeasured prior to the start of exercise in the four runs were averaged to estimate esophageal pressure at FRC. The change in EELV from FRC for the breaths analyzed at the end of each run was estimated from the relationship: AEELV

=

CL (PesFRc - PesEE) (2)

where Pesrac = end-expiratory esophageal pressure before exercise and PesEE = endexpiratory esophageal pressure during exercise with resistive loading. To obtain an approximation of inspiratory muscle work in our subjects, we used the pressure-volume relationships shown in figure 2A. The curve BCE, representing the total resistive pressure drop across lung tissue, airway, and external resistor at each volume, was drawn using the esophageal pressurevolume data obtained for the subject during inspiration. The additional pressure required to overcome thoracic viscous resistance was estimated by multiplying a thoracic resistance value estimated to be 0.92 em H 20/s/L by averaging values from the literature (4-7) by inspiratory flow rate, and the resulting values added to curve BCE to obtain curve BDE, which represents the total pressure difference across the thorax-lung-external resistor system during inspiration, except for pressures caused by elastic recoil and deformation of the chest wall. The line AGF representingelastic recoil of the thorax was drawn to intersect the FRC volume line at a point where the pressure difference was equal to the esophageal pressure measured at the end of resting expiration. Compliance of the chest wall (CT)was estimated to be 0.22 L/cm H 2 0 (8-10). Deformation of the chest wall during resistive loaded breathing requires work that is in addition to that estimated from a relaxed pressure-volume curve (11, 12).Agostoni and coworkers (13) presented a method for estimating the pressure difference required for chest wall deformation from data on changes in chest circumference with volume during relaxation and during breathing maneuvers, and the relationship between pleural pressure

Fig. 2. Different components of pressure drop aeross the lung-thorax system estimated for Subject 5 at load R3 when (A) the subject is starting inspiration at an encklxpiratory lung volume below FRC, and (B) when subject is starting inspiration at the FRC and develop. ing the same inspiratory flow-lung volume pattern as in panel A. See text for explanation.

....

oa:

TABLE 2 AVERAGE VALUES OF RESPIRATORY VARIABLES FOR ALL SUBJECTS AT EACH RESISTIVE LOAD* External Resistance (em H2O/slL) 1.4 14.5 19.9 30.6

End-tidal CO2 (%) 5.63 5.51 5.66 5.68

± ± ± ±

0.11 0.16 0.15 0.08

Tidal Volume (L) 1.36 1.56 1.50 1.72

± ± ± ±

0.06 0.17 0.11 0.14

that showed significant change (p

< 0.05

and chest circumference during relaxation. Agostoni and Mognoni (11) present data on changes in chest dimensions for a subject rebreathing through a resistance of 17 em H 20/s/L (similar to the value of R3 in the present experiments) who showed pleural pressurechanges and EELVchanges verysimilar in magnitude to those found in our subjects during runs using the R3 resistor. Thus, weestimated the pressure difference required to deform the chest wall at different volumes during inspiration using the data from Agostoni and Mognoni (11) and the method of Agostoni and coworkers (13).The pressure differencesrequired for deformation at a given volume were added to the chest wall elastic pressure values to obtain the curve AHF, which is the net expanding pressure exerted by the chest wall at a given volume during inspiration. The work of deforming the chest wall during inspiration is given by the area AHFGA. The total inspiratory work is area BDEB minus area ABEFHA. To obtain an idea of how much of the work of inspiration was transferred to the expiratory muscles, we calculated the inspiratory muscle work that would have been required to produce the same inspiratory resistivepressure drops and tidal volume with the breath starting at FRC (figure 2B). In figure 2B the line JK represents the pressure caused byelastic recoil of the lung. It was drawn to intersect the FRC volume line at the esophageal pressure difference measured in the subject at the end of resting expiration, with a slope CL determined for that subject from equa-

B

1.0

E

,g

0.5

ca z:

o

•E ~

~

0.0

29.1 24.2 26.4 20.0

± ± ± ±

Inspired Ventilation (Umin)

1.99 2.74 2.82 1.22

38.8 35.5 37.4 33.5

± ± ± ±

1.55 0.99 1.41 2.24

Duty Cyelet (%) 47.6 62.4 63.9 67.5

± ± ± ±

1.13 2.48 1.00 2.43

by ANOVA) with inspiratory resistance.

&i.

8. C

Respiratory Ratet (1/min)

* All values are means ± SE.

t Variables

A

1.5

115

+---=......'-t----~~ FRC

A

-D.5+--~-+--~-~·20 0 20 40

Pre..ure Difference, em H20

-0.5 +----+--~-~·20 0 20 40

Pressure Difference, em H20

tion 1. Inspiration was assumed to start at FRC. The resistive pressure drops across lung and external resistor measured at each volume increment above end-expiratory volume were then added to the lung elastic recoil to obtain curve BJCKE, and the estimated resistive loss across the chest wall was added to obtain curve BJDKE. The work of deforming the chest wall during inspiration is given by the area AHFA. The work of the inspiratory muscles is then area BJDKEB minus area HBEFH. The work obtained from figure 2B was subtracted from that in figure 2A to obtain the difference in inspiratory muscle work caused by change in EELV. For each subject and each load, esophageal tension-time indices (TTes) were computed using the relationship: TTes

=

(Pes/PIPes)(1i/not) (f/PIPes) Jon Pesdt)

= (3)

where Pes = mean esophageal pressure during inspiration, PIPes = esophageal pressure during maximal static inspiration at residual volume, n = duration of inspiration, not = duration of breath, t = time from start of inspiration, and f = respiratory rate. For each run, an average Tfes was computed using the same breaths as those used for calculating the rib cage-abdominal phase loops and the work of inspiration.

Results

Values of end-tidal Pco., tidal volume, respiratory frequency, and minute ventilation for the four resistive loads are shown in table 2. Each value represents the mean of all analyzed breaths for all subjects at the specified resistive load. Repeated measures ANOVA showed no change with time for any of the variables listed in table 2. However,increasing load caused a significant decrease in frequency and an increase in duty cycle. Although tidal volume values were greater at higher loads, the trend was not statistically significant (p = 0.06). The average rib cage versus abdominal displacement loops for the group of five subjects are shown for each of the four resistive loads in figure 1. The start of each loop (at point 0,0) is the point

118

ABBRECHT, AAJAGOPAL, AND KYLE

in the respiratory cyclethat represents the beginning of inspiratory airflow measured at the mouth. At the lowest load (RI) the loop showed little hysteresis. At higher loads there was greater hysteresis, with the changes occurring predominantlyon the inspiratory limb. For load RI the loop was confined almost entirely to the first quadrant, with only a very mi-

nor excursion into the fourth quadrant. As load increased, there were greater excursions into both the second and the fourth quadrants. The amount of hysteresis also increased with increasing resistive load. Typical phase relationships between the abdominal and rib cage displacement waveforms are shown for one breath in

AS

RC

G»

m

co 0

.

.a '0 i% co

AS

...

RC

~

'0

c

~

co

0

c G» E

Fig. 3. Typical rib cage (Re) and abdominal(AB)excursion waveforms obtainedfor one subjectat the four differentinspiratory resistances, R1 to R4.Inspiratoryairflowat the mouthstartedat timezero. Thestartingtimesforabdominal and rib cageexpansion are indicated by Ta and Tr, respectively.

0

'0

.a

..

AS

~

Re

-e

'0

~

co

0

0..c

.6

0

R3

f

~

U >C

W

AS

R4

RC

T~ -1

o

Tr

2

4

3

Time from Start of Inspiratory Flow, sec A 200

I

= :c en

•E

Fig. 4. Relationships amongthebeginning of rib cage expansion, abdominal expansion, andinspiratory airflowatthe mouth for different inspiratory resistances. All valuesare averages ± SE for five subjects for breaths recorded duringthe last few minutesof eachrun. A. Timedifferencebetween start of airflow and start of rib cage or abdominal expansion. Abdominal mean value for R1 differssignificantly frommeanvalues for R2, R3,and R4 (Duncan's multiple. rangetest). Ribcagemeanvaluefor R1 is significantly differentfromthatfor R4. B.Phase difference between startof airflow and start of rib cage or abdominal expansion. Abdominal mean value for R1 differssignificanUy frommeanvalues for R3andR4.Ribcagemeanvaluefor R4is significantly differentfromthat for R1 and R2.

100

o

i=

-100 30

•~

i ",

= :c en

Ias

.c

a.

B

20 10 0 -10 -20

0

10

20

30

External Resistance, em H20 sec I L

figure 3. Time zero indicates the time when mouth pressure started to fall below zero. With increasing resistive load, the outward (inspiratory) motion of the abdomen occurred at greater times before the start of inspiratory airflow at the mouth. In contrast, the inspiratory motion of the rib cage lagged more behind the start of inspiratory airflow as inspiratory resistance increased. The times at which rib cage and abdomen changed direction for each subject were determined from their averaged rib cage-abdominal loops for each load. The times were averaged across subjects for each load. The resulting average time differences between start of abdominal and rib cage expansion and beginning of inspiratory airflow for the group of five subjects are shown in figure 4A. Abdominal expansion led the initiation ofinspiratory airflow, with abdominal lead increasing significantly as load increased (p < 0.025 by ANOVA). Rib cage expansion lagged behind the beginning of airflow with increasing lag as load increased (p < 0.05 by ANOYA). Similar relationships were exhibited by the phase relationships between airflow and rib cage and abdominal motion, as shown in figure 4B. The average inspired volumes at the times that the rib cage motion changed direction during inspiration are 0.2, 2.8, 7.0, and 31.0 ml for RI, R2, R3, and R4, respectively. The lead and lag times calculated from the abdominal-rib cage loops did not differ (p > 0.1 by ANOYA) from lead and lag values obtained for the last 40 min of the runs using times read from the strip chart data. This fmding indicates that the abdominal-rib cage motion loops, which were calculated only for the last few minutes because of data processing limitations, were representative of the whole steady-state portions of the runs. End-expiratory esophageal pressures were measured for the same breaths as used for the timing and flow data that are presented in figures I, 3, and 4. Average values of PesEE for the five subjects are shown for the different resistiveloads in figure 5. PesEE became less negative as resistive load increased (p < 0.005 by ANOYA). Estimated changes in EELV from FRC for each resistive load, calculated from equation 2 using the dynamic compliance values for each subject, are shown in figure 6. End-expiratory lung volume (as indicated by the lower value on each bar) decreased significantly with increasing resistance (P < 0.005 by ANOVA), whereas the fraction of tidal volume contributed by excursion below FRC

117

EXPIRATORY MUSCLE RECRUITMENT DURING INSPIRATORY RESISTIVE LOADING AND EXERCISE

0N

2

2

:I:

3

0.8

2

0.4

E Co)

ar ~

U) U)

0

-J

0

!

a:

0-

E

CD

0)

as

.c

,g

-2

>C

w

iii

CD

0 U) W

."

i

U)

a.

e ~'0.

ii

LL

'iii

fI)

T

i

•.

,ii i i

CD

~

0)

D.

c as .c

0

~I

CD

11::::I e

FRC

0

:2

E ::::I '0

S

>

l!

-a

C

w -6 0

1

0.2

fI)

10

20

.E

30

-1 ,

External Resistance, em H20 sec I L

o

Fig. 5. Esophageal pressures at the end of expiration for eachresistance (closedcircles). All valuesareaverages ± SEforfivesubjectsfor breathsrecorded during the last part of each run. Also shown (open circle) is the average valueobtained for the five subjectsduring the control period before start of exercise or loaded breathing. ThemeanvalueforR4wassignificantly different fromthe meanvaluesfor R1, R2, and R3(Duncan's multiple-range test).

increased significantly with resistance. There was no significant difference with load in the tidal excursion above FRC. The values estimated for work required to overcome viscous resistances and tissue elastic recoil during inspiration at the different resistive loads are given in table 3. The work required to deform the chest wall during inspiration at load R3 calculated from the data of Agostoni and Mognoni (11), as described above, was 2.2070 of the total inspiratory work. Because a higher external resistor would simultaneously increase the tendency toward deformation (12)and resistivework, we used the 2.2070 figure for the loads Rl, R2, and R4. Also shown in table 3 are inspiratory muscle work calculated for the actual ex-

' 20

10

30

Fig. 6. Tidal volumesfor the group of five subjectsfor the different inspiratoryresistances. Average tidal volumeis givenbythe totalheightof each bar. Theportion of each bar belowthe FRCline represents the change in end-expiratory lung volume estimated from the change in end-expiratory esophageal pressure. Error barsshowSEfor changein EELV. The meanEELV value for R4 was significantlyless than the meanvalues for R1, R2, and R3 (Duncan's multiple-range test).

periments, and estimates of what the inspiratory muscle work would be if the subjects started inspiring from FRC instead of the reduced EELV actually used. The work required to overcome pulmonary (including external resistor) flow resistance varied from 51070 (Rl) to 96070 (R4) of estimated inspiratory muscle work. The predicted decrease in inspiratory muscle pressure-volume work caused by decreases in EELV ranged from 10.4 to 20.7070. The estimated power expended by the inspiratory muscles at each resistive load is shown in figure 7. Power increased with resistance (p < 0.001 by ANOVA). The mean power at load Rl was significantly less (Duncan's multiple-range test) than the mean power at loads R2, R3, and R4.

O+------r----------,.---or-..--L 0 o 10 20 30 External Resistance, em H20 sec I L

Fig. 7. Estimated inspiratorymusclepowerfor the four differentinspiratoryresistive loads. Uppercurve(closed circles) showsestimated powerrequirements if subjects had started inspiration from FRC. Middlecurve (open circles) showspowercalculated for actualbreathswhen subjects started inspiration ata volume belowFRC. l0wer curve (opensquares) showstension-time index. All pointsare meanvalues for all subjects. Brackets show 1 SEe

The mean powers at R2 and R3 were also less than that at R4. There was no significant difference between powers at loads R3 and R4. Also shown in figure 7 are the average esophageal tension-time indices for the five subjects at each resistive load. The average increased monotonically with load from a value of 0.045 at Rl to 0.187 at R4. Discussion During normal resting respiration, both rib cage and abdominal excursions occur only within the first quadrant. In our

TABLE 3

Work (cm H,O/L)

1.4 14.5 19.9 30.6

7.7 27.2 36.5 53.9

Thoracic Resistance

Elastic Recont

Deformation:l:

Inspiratory Muscle (actual EEL V)§

2.14 1.37 1.43 1.38

4.75 4.58 4.03 -0.19

0.32 0.73 0.92 1.21

14.9 33.7 42.9 56.3

Inspiratory Muscle (if at FRC)'

Reduction Caused by Decreased EELV

Reduction in Inspiratory Work

18.8 37.6 48.2 68.7

3.9 3.9 5.3 12.4

20.7 10.4 11.0 18.1

Definition of abbreviation: EELV • end-expiratory lung volume. * Includes viscous resistance of tissues, airway resistance, and external resistor.

t Includes elastic

recoil of the lungs and thorax. as 2.2% of the total nondeformed work for R1, R2, and R4. § Inspiratory muscle work estimated during actual breaths starting at EELV. , Inspiratory muscle work assuming subjects generated same tlm.flow profile but started inspiration at FRC.

:I: Estimated

~

External Resistance, em H20 sec I L

COMPONENTS OF ESTIMATED WORK AND THE ESTIMATED INSPIRATORY MUSCLE WORK FOR DIFFERENT EXTERNAL RESISTORS

Inspiratory External Flow Resistance (cm H,O/slL) Resistance *

~ Iii: E.

CD

(%)

118

study, with increasing resistive load there was progressive extension of the abdominal-rib cage excursions into the second and fourth quadrants (figure 2). Movement of the abdominal-rib cage excursion loops into the second quadrant indicates that the abdomen is starting to expand before inspiratory airflow has started at the mouth. In addition, the rib cage is still decreasing its volume, so that the outward abdominal motion represents abdominal paradox. Because the flow at the mouth is outward, some volume loss from the rib cage must be transferred to the abdomen. Movement of the loops into the fourth quadrant indicates that the rib cage is continuing to decrease its volume even after inspiratory airflow has begun and while the abdomen is expanding. During that time there is rib cage paradox, and volume is still being transferred from the rib cage to the abdomen. The paradoxical motions of rib cage and abdomen at higher resistive loads are also evident on examination of waveforms for single breaths, as shown in figure 3. In interpreting the inductance plethysmographic data, one must consider that increasing resistive loads, especially during exercise, may have produced changes in chest configuration (14, 15)and affected the calibration of the Respitrace. Sackner and colleagues (16) compared tidal volumes measured by spirometry with those estimated with the respiratory inductive plethysmograph in a group of subjects exercising at 800 kpm/min on a bicycleergometer. Although they found substantial differences in the tidal volumes measured by the two methods during exercise, the values for variables related to the timing of respiration (frequency, duration of breath, and fractional inspiratory time) obtained by the two methods agreed within 3 OJo. Because high levels of exercise change chest wall configuration, the data of Sackner and colleagues (16) provide some confirmation that the effects of chest wall configurational changes on timing should be small. It should be noted that the loops in figure 1 show relative motions of the rib cage and abdomen, and thus were not used to examine absolute changes in volume. Thus, the effects of configurational changes on measurements of absolute volume are not of concern here. A likelyexplanation for the timing patterns observed for abdominal and rib cage motion is the recruitment of expiratory muscles during expiration, with

ABBAECHT, AAJAGOPAL, AND KYLE

the degree of recruitment increasing with higher resistive loads. The finding of increasingly less negative end-expiratory esophageal pressures as resistive load increased is consistent with recruitment of abdominal muscles during expiration, producing an EELV below the subject's normal FRC. Use of abdominal muscles with resultant decrease in EELV during exercise has been reported in several studies (17-19). Recently, Henke and coworkers (1)measured EELV using He dilution during upright exercise on a bicycle ergometer. They found that EELV became smaller, and end-expiratory pressures became more positive as cyclework increased. Inspiratory flow-resistiveloading has also been reported to cause recruitment of abdominal muscles (3). A recent study using inspiratory elastic loading in humans showed preferential recruitment of the transversus abdominis during expiration (20). On the basis of the studies cited above, the combination of exercise and respiratory resistive load would be expected to also result in recruitment of expiratory muscles. The extension of the abdominal-rib cage displacement loops into the second and fourth quadrants with higher resistive loads can be explained by differences in the activation periods for various expiratory muscles. In human subjects during expiration below FRC, De Troyer and coworkers (21) have shown that there is electrical activation of both abdominal and the triangularis sterni (tranversus thoracis) muscles. The abdominal muscle activity ceases before the beginning of abdominal expansion, whereas the triangularis sterni activity continues for a substantial time after start of abdominal expansion. Contraction of the triangularis sterni and the internal intercostals causes an increase in intrapleural pressure (12). With the diaphragm relaxed near the end of expiration, the increased pleural pressure willbe transmitted to the abdominal compartment. Thus, continuing contraction of the triangularis sterni and internal intercostals near the end of expiration might contribute to the early outward motion of the abdomen while the rib cage is still deflating. Furthermore, the persistent activity of the triangularis sterni and expiratory rib cage muscles during early inspiration, in the absence of abdominal contraction, will cause the rib cage volume to diminish even after the start of inspiration, tending to cause excursion of the loop into the fourth quadrant. Therefore, in early

inspiration the increase in volume of the abdominal compartment is due to both inhalation of air at the mouth and transfer of volume from the rib cage to the abdomen. It should be noted that the total volume inspired before the rib cage starts expanding is very small (a maximum of 31 ml for R4), so that the total volume of the rib cage and the abdominal compartment is nearly constant during that time. Also, if the abdominal muscles begin to relax in late expiration, the caudal displacement of the abdominal contents and the action of gravity will cause the abdominal wall to move outward before inspiratory airflow begins at the mouth, thus adding to the observed excursion of the displacement loop into the second quadrant. Calculated values of EELV decreased with increasing resistive load. However, the decrement in EELV between R3 and R4 was much greater than between lower loads. A large change in breathing pattern between loads R3 and R4 could have resulted from the marked increase in perceived respiratory effort reported by all subjects for load R4. However, because changes in EELV were calculated from changes in endexpiratory esophageal pressure using equation 2, errors in estimating either CL or esophageal pressure may have contributed to the large decrease in EELV calculated for resistor R4. It is possible that, at the lower lung volumes resulting from expiratory muscle action with high resistive loads, a portion of the tidal volume occurred in a region where compliance was not constant. This would result in overestimating the amount of decrease in EELV. In performing these calculations we used values for lung dynamic compliance estimated over the same volume excursion as the breaths for which end-expiratory esophageal pressures were measured. The excellent fit (r 2 = 0.999) to the data of equation 1 indicates that compliance was essentially constant over the volume range studied, Another possible source of error in.estimating EELV by this method might be changes in configuration of the thorax or position changes of the esophageal balloon, both of which could alter the relationship between esophageal and pleural pressures. In summary, although the values for EELV based on esophageal pressure changes can be considered only to be semiquantitative in nature, they do appear to substantiate a decrease in EELV

EXPIRATORY MUSCLE RECRUITMENT DURING INSPIRATORY RESISTIVE LOADING AND EXERCISE

with increasing resistive load, consistent with increasing abdominal muscle recruitment with additional load. Possible causes of the change in breathing and respiratory muscle patterns observed in our study include adaptive responses to resistive loading or respiratory muscle fatigue. Esophageal tension time indices in our subjects at load R4 ranged from 0.146to 0.234, with a mean of 0.187. Bellemare and Grassino (22) reported that subjects could breathe for greater than 45 min at a diaphragmatic tension-time index (TTdi) of 0.15,but for only 25 to 35 min at a TTdi of 0.24. Although tension-time indices based on esophageal pressure are not directly comparable with TTdi values,it appears likely that our subjects breathing against load R4 were on the threshold of developing respiratory muscle fatigue. However, actual occurrence of significant fatigue is unlikely in view of the following findings: (1) maximal negative mouth pressure developed by the subjects immediately after the exercise runs was not significantly different from the preexercise values, (2) there were no significant changes in ventilatory parameters with time over the last portions of the runs, and (3) there was no effect of resistive load on the end-tidal Pco, values. Thus, it is likely that the change in respiratory pattern with load represented an adaptive mechanism to minimize the energy required by the respiratory muscles, or to increasethe overalleffectivenessof the respiratory muscular apparatus. For example, behavioral factors might result in augmented motor output in response to the sensation of increased effort, or the response might represent a metabolic adaptation to limit inspiratory muscle power requirements. Calculations of the net inspiratory power, results of which are presented in figure 7, suggest that power approaches a limiting value at higher work loads. It must be noted that calculation of net inspiratory work required use of severalassumptions. A value for the viscous resistance of the thorax was obtained from the literature. Because we did not measure thoracic compliance, weused the average of several extensive studies reported in the literature (8-10). Weconsidered thoracic compliance to be constant over the range of tidal volumes used by our subjects. Although examination of thoracic pressure-volume curves reported in other studies (8-10) indicates that the assumption of constant thoracic compli-

ance is reasonable in our subjects when above FRC, there is a question as to whether thoracic elastic recoil is underestimated at our subjects' largest excursions below FRC. The effects of such an underestimation would be partially offset by the fact that esophageal pressure tends to underestimate the degree of negativity of the pleural space. During inspiratory resistive loading, work is done to distort the rib cage in addition to the work that accompanies changes in volume (11, 12). To estimate the magnitude of this work, we used data on changes in rib cage dimensions and rib cage pressure-volume relationships presented in the literature (11) for a subject breathing against a resistive load nearly equal to our R3. That subject showed changes in pleural pressure and EEL~ and also showed a tidal volumevery similar to that of our subjects at load R3. Thus, it is reasonable to assume that the work of rib cage deformation in our subjects would also be similar to that calculated from the data of Agostoni and Mognoni (11). The calculations indicate that at load R3 the work of distorting the chest wall would be about 2.2070 of the total inspiratory work. Incidentally, Agostoni and colleagues (23) report that the deformation work during the expiratory phase for the same subject when rebreathing through the same resistance was 1.5070 of the work of the expiratory muscles. That the work of deformation appears to be a relatively small fraction of the total work of breathing in the case where there are large resistive loads may be due to the following: (1) the large amount of flow-resistive pressure drop (cf. figure 1)and (2) the net deformation work represents the difference between the positive work done on some parts of the system and the negative work done by other parts of the system (24). Thus, the net work required for deformation may be relatively small. In estimating the total work of inspiration for loads Rl, R2, and R4, it was assumed that the work of distorting the chest wall would also be equal to 2.2070 of total inspiratory work. Even if this percentage estimate was in error by a factor of2 or 3, it would not have a significant effect on the conclusions regarding total inspiratory work. The estimates of net inspiratory muscle power (figure 7) indicate that power tends to approach an asymptotic value as external inspiratory resistance increases. This observation suggests that, during inspiratory flow-resistive loading,

119

the respiratory system tends to limit inspiratory power to a fixed maximum by transferring any additional power requirements to the abdominal muscles. Use of abdominal muscles during inspiratory resistive loading helps prevent inspiratory muscle fatigue by reducing the power requirements for the inspiratory muscles and by placing the diaphragm at a better mechanical advantage at the start of inspiration. The estimates presented in table 3 indicate that the amount of unloading resulting from expiratory muscle recruitment is on the order of 10 to 20070. Although the expiratory muscles thus appear to make a relatively small contribution in terms of reducing power demands, the difference may be enough to prevent occurrence of inspiratory muscle fatigue at the resistive loads that we studied. Acknowledgment The writers thank William Slivka for his excellent technical assistance.

References 1. Henke KO, Sharratt M, Pegelow D, Dempsey JA. Regulation of end-expiratory lung volume during exercise. J Appl Physiol 1988; 64:135-46. 2. Martin JO, De Troyer A. The behavior of the abdominal muscles during inspiratory mechanical loading. Respir Physiol 1982; 50:63-73. 3. Martin J, Aubier M, Engel LA. Effects of inspiratory loading on respiratory muscle activity during expiration. Am Rev Respir Dis 1982; 125:352-8. 4. Bhansali PV, Irvin CO, Dempsey JA, Bush R, Webster JO. Human pulmonary resistance: effect of frequency and gas physical properties. J Appl Physiol 1979; 47:161-8. 5. Cha EJ, Park KS, LeeCW, Kim KY, Min BO. Estimation of respiratory impedance and source pressure using a Thevenin equivalent circuit model. J Biomech 1983; 16:635-41. 6. Nagels J, Landser FJ, Van der Linden L, Clement J, Van de Woestijne KP. Mechanical properties of lungs and chest wall during spontaneous breathing. J Appl Physiol 1980; 49:408-16. 7. Sobol BJ. Tests of ventilatory function not requiring maximal subject effort. II. The measurement of total respiratory impedance. Am Rev Respir Dis 1968; 97:868-79. 8. Cherniack RM, Hodson A. Compliance of the chest wall in chronic bronchitis and emphysema. J Appl Physiol 1963; 18:707-11. 9. Heaf PJD, Prime FJ. The compliance of the thorax in normal human subjects. Clin Sci 1965; 15:319-27. 10. Naimark A, Cherniack RM. Compliance of the respiratory system and its components in health and obesity. J Appl Physiol 1960; 15:377-82. 11. Agostoni E, Mognoni P. Deformation of the chest wall during breathing efforts. J Appl Physiol 1966; 21:1827-32. 12. Sampson MO, De 'Iroyer A. Role of intercostal muscles in the rib cage distortions produced by inspiratory loads. J Appl Physiol 1982; 52:517-23. 13. Agostoni E, Mognoni P, Torri 0, Agostoni AF. Static features of the passive rib cage and

120 abdomen-diaphragm. J Appl Physiol 1965; 20: 1187-93. 14. Grimby G, Bunn J, Mead J. Relative contributions of the rib cage and abdomen to ventilation during exercise. J Appl Physiol 1968; 24:159-66. 15. Goldman MD, Grimby G, Mead J. Mechanical work of breathing from rib cage and abdominal V-P partitioning. J Appl Physiol 1976; 41: 752-63. 16. Sackner JD, Nixon AJ, Davis B, Atkins N, Sackner MA. Noninvasive measurement of ventilation during exercise using a respiratory inductive plethysmograph. Am Rev Respir Dis 1980; 122: 867-71. 17. Kiers A, Van der Mark TW, Woldring MG,

A8BRECHT, RAJAGOPAL, AND KYLE

Peset R. Determination of the functional residual capacity during exercise. Ergonomics 1980; 23:955-9. 18. Lind F, Hesser CM. Breathing pattern and lung volumes during exercise. Acta Physiol Scand 1984; 120:123-9. 19. Sackner JD, Nixon AJ, Davis B, AtkinsN, Sackner MA. Effects of breathing through external dead space on ventilation at rest and during exercise. II. Am Rev Respir Dis 1980; 122:933-40. 20. De 'Iroyer A, Estenne M, Ninane V, van Gansbeke 0, Gorini M. Transversus abdominis muscle function in humans. J Appl Physioll990; 68:1010-6. 21. De Troyer A, Ninane V,Gilmartin J J, Lemerre C, Estenne M. Triangularis sterni muscle use in su-

pine humans. J Appl Physiol 1987; 62:919-25. 22. Bellemare F, Grassino A. Effect of pressure and timing of contraction on human diaphragm fatigue. J Appl Physiol 1982; 53:1190-5. 23. Agostoni E, Mognoni P, Torri G, Miser G. Forces deforming the rib cage. Respir Physioll966; 2:105-17. 24. Kono K, Mead J. Static volume-pressure of the rib cage and abdomen. J Appl Physiol 1968; 24:544-8. 25. Abbrecht P, Thio 1M, Kyle RR. A simple algorithm for averaging multiple cycles of waveforms having varying cycle periods. Comput Appl Biosci (In Press).