RESPIRATORY REGULATION DURING EXERCISE IUNCONDI- TIONED

RESPIRATORYREGULATIONDURINGEXERCISE IN IUNCONDI-TIONED SUBJECTS'

By J. B. HICKAM, W. W. PRYOR,2 E. B. PAGE, AND R. J. ATWELL

(From the Department of Medicine, Duke University School of Medicine, Durham, N. C.)

(Submitted for publication December 13, 1950; accepted, March 12, 1951)

INTRODUCTION

The regulation of breathing during muscularexercise has been the subject of much work andspeculation. Most of the experience has beengained with subjects in good physical training.In general, blood chemical changes in these sub-jects have not been great during exercise, and ithas been difficult to account for the amount ofventilation on the basis of these changes and otherknown respiratory stimulants. The present studywas done on untrained subjects in only fair physi-cal condition. It seemed likely that the coordina-tion of various respiratory drives would be lesssmooth in untrained subjects than in trained per-sons, so that the part played by different stimulimight become more obvious. The conduct of suchstudies has been simplified by the recent develop-ment of techniques for recurrent sampling of ar-terial blood during the course of exercise. Nearlyall past studies have had to rely on indirect meansof estimating arterial blood changes.

METHODS

The subjects were 11 male physicians and medical stu-dents in normal health but not in physical training. Theyranged in age from 22 to 35 years and in surface areafrom 1.86 to 2.12 sq. m. They were not fasting and cameto the experiment directly from their work. Arterialblood samples were taken from the brachial arterythrough an in-lying Cournand needle atfached by flexibleplastic tubing to a series of three way stopcocks. Bloodwas taken into syringes wet with heparin. For pH deter-minations, sodium fluoride was added as a preservative.The syringes were capped and stored in ice water.Analyses were completed within two to four hours. Thesubjects breathed through a Douglas valve into a re-cording spirometer. For determination of oxygen con-sumption the expired air was diverted into a Douglas bag.Gas mixtures other than air were supplied to the inspira-

1 This work was supported by grants from the Life In-surance Medical Research Fund and the Anna H. HanesMemorial Fund.

2 U. S. Public Health Service Postdoctorate ResearchFellow.

tory side of the system from large Douglas bags. Thesemixtures consisted of 15% 02 in N2, tank oxygen (99.5%),and 5% CO2, 21% 0, in N2. Resting measurements weremade with the subject seated. Ventilation was also meas-ured with the subject standing, just before exercise. Ex-ercise was carried out in an airconditioned room at 250 C.on a treadmill at a speed of 5.1 miles per hour on a 3.5%grade. For most of the subjects this exercise rate wasfairly severe. Oxygen consumption, measured in eightsubjects about five minutes after beginning exercise, was2,290±90 cc./min. (range: 1,760 -2,610). Ventilationwas recorded continuously and is expressed as ventilationat room temperature (25 ± 1° C.). 3 Arterial blood sam-ples were taken at points during the breathing of variousgas mixtures, at rest and exercise, when it became evi-dent from the respiratory tracing that the ventilation ratehad been nearly steady for at least 30 seconds. The sub-jects were studied on two separate days, one of whichwas devoted to the effects of the carbon dioxide mixture,and the other to the effects of the various oxygen mixtures.

Respiratory gas analyses were done with the Haldaneapparatus or with a Pauling oxygen meter. Arterialblood oxygen content and per cent oxygen saturationwere determined by a spectrophotometric method (1).The pH of whole blood was measured with a Cam-bridge Model R pH meter equipped with an inclosed glasselectrode. Measurements were made at room tempera-ture, usually 25-26° C., and the result was corrected to 370C. by Rosenthal's factor (2). This temperature was ar-bitrarily chosen for all pH measurements on arterialblood. pH determinations were made in duplicate innearly every case, and duplicate determinations agreed towithin 0.02 pH units or less. The stability of the in-strument was checked with a pH 4.00 acid phthalate buf-fer or a pH 7.01 phosphate buffer between each reading.The carbon dioxide content of whole blood was determinedby the method of Van Slyke and Neill (3), and the plasmacarbon dioxide content was estimated from this value andfrom the pH and hemoglobin by the line chart of VanSlyke and Sendroy (4). From the pH and CO2 contentof arterial plasma at 370 C., the CO2 tension of arterialblood was calculated by the Henderson-Hasselbalch equa-tion. For the calculation of CO2 tension at 37° C., pK'was taken to be 6.11, as found by Dill, Daly and Forbes(5). Where the method was evaluated, as described be-low, by equilibration of blood in a tonometer at 380 C., apK' value of 6.10 was used, as determined for this tem-perature by Hastings, Sendroy, and Van Slyke (6).

The determination of CO2tension by this indirect means

3 Range of barometric pressure: 747-757 mm. Hg.

503

J. B. HICKAM, W. W. PRYOR, E. B. PAGE, AND R. J. ATWELL

depends upon several measurements and calculations, ofwhich pH determination and temperature correction seemmost liable to error. The validity of the procedure wastherefore tested by equilibrating blood samples in tonom-eters with a variety of CO2 tensions at 380 C. Theactual CO, tension or "direct tension" was determined byHaldane analysis of the gas mixture in the tonometer.The blood was then removed from the tonometer, cooledto room temperature, and determination of "indirect CO2tension" carried out as described above, except that pH'swere corrected to 380 C. In this way, 23 comparisonswere made between direct and indirect CO2 tensions.The direct tensions ranged from 19.5 to 58.1 mm. Hg, andthe corrected direct pH's from 7.59 to 7.13. The per centoxygen saturation of these bloods was between 70 and100% in 13 instances, and between 25 and 70% in theremainder. The mean indirect CO, tension exceeded thedirect by 1.62 mm. Hg, with a standard deviation of 2.03,and a standard error of 0.42. This difference, thoughsmall, was significant (P < .01). In addition, a com-parison of indirect arterial CO, tension with alveolarCO2 tension was made in 17 instances. The alveolarairs were collected during rest and exercise with thevariations in technique for these conditions describedby Dill and colleagues (7, 8), and also during hy-perventilation on air and various CO, mixtures. Al-veolar airs do not provide a satisfactorily exact checkon arterial CO2 tensions, but the comparison was thoughtto be required by the possibility that changes in the bloodduring strenuous exercise might seriously invalidate thecorrections made in the indirect measurement of CO,tension. The alveolar tensions ranged from 17.9 to 76.7mm. Hg, and the blood pH's from 7.59 to 7.15. The meanindirect arterial CO, tension exceeded the alveolar ten-sion by 1.6 mm. Hg, with a standard deviation of 3.7.The difference was not significant. These agreementsseem satisfactory. It is not clear why the indirect CO,tension is significantly greater than the direct. It seemsmost probable that the error lies in measurement of bloodpH. The difference could be caused by a mean error ofapproximately .02 units below the true value. The devi-ations between direct and indirect CO2 tensions were notmore marked at extreme pH ranges than at more ordinaryvalues. This is of importance because Rosenthal's temper-ature coefficient for whole blood was established for a pHrange of 7.25 to 7.45, while the temperature coefficient ofserum has been found to vary considerably at extreme pHranges (6.6 and 7.8) (5). The present results indicatethat Rosenthal's coefficient for human whole blood cansatisfactorily be extended over the pH ranges encounteredin this study.

RESULTS

The data on each subject relating respiratoryvolume to arterial blood measurements at rest andexercise and during the breathing of various gasmixtures are presented in Tables I and II.

Resting arterial blood

The mean values for resting arterial blood arepresented in Table III. Of the 11 subjects, ar-terial blood was obtained from one on only a singleoccasion, and from two subjects on three occasions.In the remaining eight, samples were taken duringtwo separate trials. In determining mean valueseach subject was given equal weight. The arterialoxygen saturation is within the general range ofrecent reports by Comroe and Walker (97.5%)(9), Woodand Geraci (97.9%o) (10), and Drab-kin and Schmidt (98.6%) (11). The pH is some-what lower than that found for arterial blood byDill, Edwards, and Consolazio (12) (7.40) andby Dill and associates (7.39) (13), but higherthan that reported by Barr, Himwich, and Green(7.30-7.36) (14). The CO2 tension is slightlyhigher than that reported by Dill and colleagues(41.0, 40.9) (12, 13), higher than that of Lilien-thal and coworkers, by Riley's direct method(37.8) (15), close to that of Galdston and Wol-lack (42 direct, 44 indirect) (16), and less thanthat of Barr and colleagues (44-45) (14). Asnoted above, the mean CO2 tension is probablyabout 2 mm. Hg too high, and the pH .02 units toolow. In addition, the values are probably affectedby the small dead space of the respiratory appa-ratus, and are almost certainly affected by the factthat the subjects were not in a basal state, but cameto the experiment directly from their daily activi-ties. Under these conditions the variability in pHand CO2 tension among normal subjects was quitestriking.

Arterial blood changes during exercise

The present observations are not directly com-parable to many of those reported in the literaturefor the reason that the exercise period was tooshort in most cases for the subject to reach a steadystate. In addition, the subjects were untrained,and the exercise accordingly varied in its relativeseverity. For some subjects it was severe enoughto preclude reaching a steady state.

The arterial blood oxygen capacity rose in allsubjects, the mean increase being 6%o. This iswithin the general range previously reported (8,14, 17-19). The increase had reached nearly itsfull value after three minutes of exercise. Theper cent oxygen saturation gradually declined dur-

504

RESPIRATORYREGULATIONDURING EXERCISE

TABLE I

Effect of 5%C02 during exercise

Surae sa Inspired Time on | %Arterial Os Arterial Arterial COtSurtace State mixture mixture setlto aturation seum CO2 Arterial pH teso

(sq. m.)R.1.93

R.1.86

M.1.90

K.1.88

D.2.12

L.1.95

C.1.92

G.2.04

S.1.89

T.2.04

Rest

Exercise

Rest

Exercise

Rest

Exercise

Rest

Exercise

Rest

Exercise

Rest

Exercise

Rest

Exercise

Rest

Rest

Exercise

Rest

Exercise

Rest

Exercise

RestExercise

Air5%C02Air5%C02Air

Air5%C02Air5%CO2Air

Air5%C02Air5%C02Air

Air5%C02Air5%C02Air

Air5%C02Air5%C02Air

Air5%C02Air5%C02Air

Air5%C02Air5%C02Air

Air5%C02

Air5%CO2Air5%C02Air

Air5%CO2Air5%C02Air

Air5%C02Air5%C02Air

AirAir5%CO2Air

2:003:251:381:50

1:502:451:100:54

2:003:501:271:00

2:004:002:002:10

1:503:401:002:10

2:003:451:151:50

1:503:301:451:00

2:45

1:453:151:001:10

1:503:271:201:03

1:453:151:021:50

4:102:052:46

(5./mis., 2z C.)6.19.6

51.481.878.6

6.422.671.1

122.084.0

9.711.558.087.067.0

6.517.540.372.751.1

15.916.849.466.065.0

8.315.242.057.0

59.7

6.310.744.381.551.0

7.617.6

9.114.047.058.052.4

7.1

22.237.472.549.9

6.911.830.027.331.0

5.920.534.836.0

95.898.495.995.894.2

97.599.698.898.297.8

99.398.797.197.497.5

98.299.797.899.197.8

99.499.295.196.894.0

98.498.797.097.595.6

98.498.898.398.597.8

97.499.2

98.999.196.997.596.1

97.898.997.397.697.4

96.798.996.897.298.1

97.793.197.395.8

(Vi. %)53.257.142.946.234.5

53.556.240.747.734.5

51.054.443.349.640.3

58.160.148.553.145.1

52.659.551.955.944.6

56.761.351.553.242.8

56.460.653.257.650.0

55.759.9

60.763.152.758.753.0

56.459.252.759.951.4

57.561.858.564.057.8

58.657.860.653.6

7.417.347.267.197.20

7.387.327.297.137.21

7.387.297.207.137.23

7.387.327.297.197.27

7.507.347.297.147.21

7.357.257.207.107.13

7.367.297.317.227.32

7.357.31

7.407.337.287.19

7.27

7.377.327.347.207.27

7.367.297.297.197.29

7.367.237.157.22

(mm. Hi)3847375339

4149386238

3950496542

4452456143

3149487149

4662587356

4556476243

4553

4453506751

4451436749

4657547353

47617558

505


TABLE II

Effect of 15% 02 and tank 02 during exercise

Subject State mInstured Tmixue Ventilation %IArteriall0pHArterionmixture mixture ~~saturation serum CC,Arei2P rra O

H.

R.

M.

K.

D.

L.

C.

B.

G.

S.

T.

RestExercise

*Exercise

RestExercise

RestExercise

RestExercise

RestExercise

RestExercise

RestExercise

RestExercise

RestExercise

RestExercise

RestExercise

* Separate study.

AirAir15% 02Air

Air100% 02

AirAir15% 02Air100% 02AirAir15% 02Air100% 02AirAir15% 02100% 02Air

AirAir15% 02Air100% 02AirAir15% 02Air100% 02Air

AirAir15% 02100% 02Air


AirAir15% 02Air100% 02Air


AirAir15%0/ 02100% 02Air

2:500:541:25

8:001:15

3:001:001:101:30

2:401:241:101:10

1:361:151:241:15

3:361:201:151:27

3:361:461:571:461:45

2:361:362:001:54

3:101:362:252:05

3:511:402:181:391:57

2:202:102:402:36

2:361:502:001:20

(I./mix., Z50 C.)8.9

66.694.887.4

8470

7.672.097.072.063.0

12.765.084.073.049.0

8.369.163.064.070.6

13.940.761.253.549.9

6.640.040.048.043.054.0

7.840.656.538.851.0

7.437.253.040.647.0

9.637.646.040.535.040.9

8.836.650.932.429.7

9.931.431.436.336.6

99.497.786.995.4

97.698.086.995.4

100.0

99.298.089.497.3

100.0

99.499.186.7

100.094.5

98.796.278.095.8

100.0

98.397.188.796.0

100.095.0

98.097.990.4

100.097.3

99.796.688.4

100.096.5

99.097.889.397.1

100.096.7

95.395.384.4

100.093.0

98.197.673.6

100.095.3

(vol. %)51.643.537.233.1

55.043.438.136.536.4

52.144.439.440.738.2

54.648.841.840.038.4

51.350.546.543.744.5

57.952.649.849.248.646.9

56.350.947.954.749.4

54.050.849.249.448.8

57.454.053.452.653.153.2

59.355.549.854.156.758.655.254.351.850.0

7.387.287.277.15

7.377.237.217.167.187.447.327.317.257.20

7.327.277.257.197.23

7.427.237.227.177.14

7.337.267.247.217.157.14

7.297.297.267.307.32

7.327.247.267.247.29

7.387.347.367.317.297.31

7.337.207.237.187.20

7.357.287.237.247.20

(mm. Hg)39413641

4346424543

3538354143

4747424640

3653545257

495251546160

5247474943

4752485145

444542474947

50625263634752575356

506

RESPIRATOkY REGULATION DURING EXERCISE

TABLE III

Mean arterial blood values of 11 subjects at rest and during exercise

State Ventilation Blood CO SerumCo0 pH CO tension Os capacity | 0tuatoncontent contentsauto

(I./min., 250 C.) (vi. %) (voI. %) (mm. Hg) (vol. %)Rest 8.7±0.7 46.3±0.7 55.4±0.8 7.37±.01 43.341.5 19.51 40.32 98.3Exercise 3'11"$ 47.1±4.0 41.8±1.2 49.9±1.5 7.27±.01 48.2±1.6 20.52±0.37 97.1Exercise 8'20"1* 57.2±4.6 38.1±1.7 45.4±2.1 7.22±.02 47.6±2.1 20.67±0.33 96.1

*Average time. (Ranges: 1'36" to 4'30" and 5'45" to 11'30".) Indicated range following mean values denotesstandard error.

ing the exercise period, falling by an average of2.2%. A similar fall has been reported by Asmus-sen and Nielsen (20). The arterial oxygen ten-sion has been found to fall during exercise (15).In addition, the pH fall in our subjects would havethe effect of shifting the oxyhemoglobin dissocia-tion curve so that the per cent saturation would beless for a given oxygen tension.

For the group as a whole there was a consider-able fall in pH and in serum CO2content. A largechange had occurred by the time of the first sam-ple, and the decline was continued in the secondsample. Decreases of arterial pH and CO2 con-tent during exercise are a reflection of the accumu-lation of lactic acid (8, 21), and changes of thismagnitude indicate that the rate of exercise wassevere for the group as a whole. With subjectsin better training, exercise rates of this degree ormuch higher may be maintained without the ac-cumulation of large amounts of lactic acid andwithout a decline in pH (8, 22). However, withexercise which is severe for the subject, fall in ar-terial pH and serum CO2 content of this generalmagnitude have been reported (14, 21, 23). Ourexperience is similar to Barr's and Himwich's (23)in the finding that resting pH and CO content maynot be restored for many minutes after exercise.

The mean arterial CO2 tension was slightly in-creased during exercise. Because of the great in-dividual variation, the difference between the rest-ing and exercising levels is not significant for thegroup as a whole. Nielsen (22) reported that theCO2 tension falls during exercise and that the fallis greater with increasing severity of exercise.His observations were made during the steadystate of exercise on two highly trained subjects,and the arterial CO2 tension was calculated fromthe composition of expired air and the respiratorydead space. Data on this subject are relatively few.

The observations of Barr and associates (14),and of Galdston and Wollack (16) showing a fallor no substantial change in arterial CO2 tensionwere made on arterial blood drawn shortly afterexercise was completed. Lilienthal and colleagues(15), measured CO2 tension by Riley's methodin arterial blood drawn during exercise from sixsubjects. The sampling time was comparable tothat of the present study, but, judging from theoxygen consumption, the severity of exercise wassomewhat less. They found an average increaseof approximately 3 mm. Hg in arterial CO2 tensionduring work. Dill and associates have investigatedthe arterial CO2 tension during exercise as meas-ured by "inspiratory" alveolar air. There was con-siderable individual variation, but the average ten-sion, after a steady state was reached, was not farfrom the resting value (7, 8). Four trained sub-jects at graded work rates showed a moderate in-crease in CO2 tension with lighter work rates anda gradual decline toward the resting level as therate was increased (13). Briggs (24) studied sub-jects under a very large work load. In mnany ofthe best performers, there was a high CO2 con-tent in the expired air (up to 8% in two subjects),indicating a very high arterial CO2 tension. Inshort, the available evidence indicates that the ar-terial CO2 tension during exercise may rise, fall,or remain nearly unchanged depending upon theindividual subject, the work rate, and the dura-tion of work. In four cases in the present studythe CO2 tension within one minute after exer-cise was found to be substantially the same as theoriginal resting value.

The course of ventilation during exerciseAt the start of exercise, as found by Krogh and

Lindhard (25), there is an immediate increase inventilation. With our subjects, this increased ven-

507


tilatory rate was maintained with relatively minorfluctuations, particularly in the first four to fivebreaths, for roughly a minute. A rather sharpfurther increase in ventilation then occurred. Thissharp "secondary rise" in ventilation was presentconsistently except in one subject, T., in whomitappeared in only one of three trials. The secondaryrise usually took place within a period of one tothree breaths. The time at which the secondaryrise occurred varied from 30 to 67 seconds afterthe onset of exercise, with an average time of 52seconds. The mean increase at this point was

10.4 + 1.6 1./min. The ventilation thereafter rose

much more gradually and smoothly, reaching a

steady state in some cases. This secondary riseseems not to have been commented upon before,although close observations of breathing duringearly exercise have been made by Krogh and Lind-hard (25) and by Asmussen and Nielsen (26).Their subjects were in good physical condition.Continuous recordings of ventilation throughoutthe course of exercise have been made by Rahnand Otis (27) and by Bruce and coworkers (28),without the finding of a sharp break in ventilation.In these studies the subjects were untrained, butthe work rate was quite mild.

It seemed possible that the secondary rise mightreflect the beginning of arterial blood changes re-

sulting from exercise. This possibility was exam-

ined in two further subjects who showed a well-marked secondary rise on moderate exercise (cor-responding to an 02 consumption of about 1,400cc./min.). The results are presented in Table IV.It is apparent that there was a rapid fall in pH anda rise in arterial CO2tension during early exercise.The total extent of these changes, as measured bythe effect on ventilation of breathing CO2 at rest,could account roughly for the magnitude of thesecondary rise. However, the ventilation rise issharp, while the blood chemical changes are more

gradual. This difference may reflect a lag be-tween arterial blood changes and the ventilatoryresponse. It seems possible only to conclude thatchemical changes may occur quite early in the ar-

terial blood of untrained subjects even during mod-erate exercise, and that these changes may be re-

sponsible for the secondary rise in ventilation.On stopping exercise, there was usually a sharp

decrease in the ventilation rate, as noted by As-mussen and Nielsen (26). The magnitude of thisdecrease varied greatly. The mean fall was 8.8 -4

2.6 1./min. Because of individual variation, there

TABLE IV

Ventilation and arterial blood changes during early exercise and while breathing COrenriched air at rest

Subject Sampling time after Comment Ventilation Serum COs pH at 370 C. CO tensionstarting exercise

(I./min., 250 C.) (vol. %) (mm. Hg)R. M. Rest 8.8 53.6 7.37 41

32"-41" 12.7* 55.5 7.34 45Secondary vent. rise to19.8 I./min. begins at 63sec.

71"-82" 19.8 55.8 7.28 523'23"-3'33" 36.5 50.9 7.21 55

Rest 28 min. after exercise 8.1 49.0 7.32 42Rest Breathing 3-4% CO2for 15.4 53.7 7.29 49

4 mmn.E. S. Rest 8.5 56.0 7.38 43

16"-26" 55.5 7.34 4620.7*

4511-5511 56.6 7.32 49Secondary vent. rise to32.5 1./min. begins at 52sec.

2'49"-3'04" 43.6 53.1 7.27 51Rest 22 min. after exercise 9.7 51.5 7.34 43Rest Breathing 4-5% CO2for 25.2 56.5 7.28 53

2i min.

* Average after first four to five respirations (about 15 seconds) until secondary rise in ventilation. Treadmillexercise at 4 mph and a 4%grade, subjects in poor training. A sharp "secondary rise" in ventilation is associated withan increase in CO2 tension. Subjects show normal response to breathing CO2at rest.

508


was no significant difference between the magni-tude of the immediate increase at the beginning ofexercise and the decrease at the end.

The effect of breathing 5%o CO2

In comparing the effect on ventilation of breath-ing 5% CO2 at rest and during exercise, the diffi-culty arose that a number of our subjects were un-

able to continue exercise with the gas mixture longenough to reach a steady state. Accordingly, themixture was given until it was evident from therespiratory tracing that the ventilation was no

longer changing rapidly, and arterial samples were

then taken. For comparison, the same procedurewas adopted in the resting state. The average timeon CO2 at rest was two minutes and during exer-

cise one and one-half minutes. The average timeneeded to reach a steady state on 5%CO2 at restprobably exceeds five minutes and in individualcases may be much more (22, 29, 30). However,the breathing time used in the present study covers

the period of rapid change in ventilation. In meas-

uring the effect of CO2 inhalation during exercise,it was necessary to select control levels for ventila-tion and arterial CO2 tension. This was compli-cated by the fact that the subjects were not in a

steady state, and ventilation was consequently in-creasing with time, apart from the effect of CO2.As shown in Table I, during the same exercise pe-

riod control measurements were made while thesubject was breathing air, both before and afterthe administration of CO2. The control value forventilation while the subject was breathing CO2was estimated by assuming that the ventilation timecurve would have followed a straight line betweenthese two points if the subject had continued on

air. The mean values for the two points on air,before and after giving C02, were 46.6 1./min. atthree and one-half minutes of exercise, and 58.31./min. after seven and one-half minutes of total

exercising time.4 The mean estimated control ven-

tilation during CO2 breathing was 53.3 1./min.,while the mean actual ventilation was 73.0 1./min.The estimated control ventilation is a little too

'The four minutes between these sampling times in-cluded one and one-half minutes each on CO2 and air, withapproximately 30 additional seconds after each periodwhich were occupied in completing the blood samplingand observing the ventilation rate to insure that it wasnot changing rapidly.

TABLE V

Mean effect on 10 subjects of breathing 5% CO2at restand during exercise

StatVeni CO, Ventilation change/COsState Ventilation tension tension change

(1.mmn.,250 C.) (mm. Hg) (I./min./mm. Hg)

Rest 8.3±-0.9 41.8 ±1.5Rest, 5%CO 15.5±1.4 52.6±1.4 .80±.24Exercise 53.3 44.2* 46.9 42.1*Exercise, 5% COs 73.0 ±7.5 65.5 2.0 1.05±4.20

* Expected values, calculated from measurements duringexercise before and after the administration of CO2. Seetext.

low because the ventilation-time curve is not astraight line but is convex upward. The apparenteffect of CO2 is thereby magnified, but the effectis not great. In the second control period, meas-urements were made at an average of one and one-half minutes after changing from CO2 to air. Itmight have been expected that measurements dur-ing this period would still be much influenced bythe preceding CO2 inhalation. However, the meanventilation at this point (58.3 ± 4.4 1./min.) wasnot significantly different from that of the lastcontrol period in the oxygen experiments (56.0 +5.4 1./min.), which are described in the next sec-tion. In most of the latter, the preceding gas mix-ture was tank oxygen, which depresses ventilationinstead of elevating it. It is concluded that ventila-tion in the second control period was not signifi-cantly influenced by the preceding gas mixture ineither case. The mean CO2 tension was approxi-mately 47 mm. Hg in each control period. Therewas very little individual change between periods,and a simple average was taken for the CO2tensionto be expected if the subject had continued on air.It appears that there is a rapid recovery from theeffect of inhaling CO2during exercise.

The mean results for 10 subjects are presentedin Table V. The mean ventilation response to anincrease in arterial CO2 tension during exerciseexceeds that at rest, but the difference is not sta-tistically significant (0.3 < P < 0.4). It was sug-gested by Nielsen (22) that the respiratory centerbecomes more sensitive to CO2 during exercise,and that this change is an important factor in theregulation of breathing during exercise. Underthe conditions of the present experiment, there isno evidence that exercise has sensitized the re-spiratory center to CO2in the sense that the ventil-atory response to a given increment in CO2 ten-

509


sion becomes significantly greater. The sameconclusion was reached by Grodins (31) from anexamination of Nielsen's data. The magnitude ofthe response to 5% CO2 at rest is somewhat lowerthan usual, probably because of the short exposuretime. The mean pH fall as a result of breathingCO2 was slightly greater during exercise than atrest. As a corollary to the first conclusion, thereappears to be no evidence that exercise producesany marked respiratory sensitivity to a fall in ar-terial pH.

The effect of high and low oxygen concentrationsIn 1920 Briggs (24) observed that breathing

oxygen-enriched air during severe exercise de-creased the ventilation rate. The effect wasgreater the more severe the exercise, and for agiven exercise rate the effect varied inversely withphysical fitness. The finding has been confirmedabundantly (20, 22, 27, 32, 33). Conversely, it isknown that ventilation for a given work rate isgreater at high altitudes (13, 34, 35) or whilebreathing low-oxygen mixtures (15, 27, 36, 37)than it is at normal oxygen tensions. In the pres-ent study a trial was made of the effect on ventila-tion of changing from air to 15%o 02 in N2 and totank oxygen.

The first control measurements were made afterabout three minutes of exercise. The subject wasthen switched to 15% oxygen, and thereafter toair or tank oxygen. In all cases there were two orthree control periods on air, as indicated in TableII. The time on each mixture was roughly one andone-half minutes before measurements were made.The subjects did not know the order in whichmixtures were given and the changes were madewithout their knowledge. The method of deter-

TABLE VI

Mean effect on 11 subjects of breathing 15% 02 andtank 02 during exercise

%reilVentilationState Ventilation os saturation change

(I./mix., 250 C.) (I./min.)Exercise, air 52.1:1=4.8* 96.9*Exercise, 15% 02 61.6±6.5 85.7 9.5±3.1tExercise, air 55.8±5.3* 96.1*Exercise, tank 02 47.4±3.9 100.0 8.442.3t

* Expected values, calculated from measurements duringcontrol periods. See text.

t P <.02.tp <.01.

mining "expected" ventilation and per cent oxy-gen saturation from the control periods was thesame as described in the preceding section. Themean results for 11 subjects are presented in TableVI. Fifteen %o 0° caused a significant increase inventilation (P < .05) and tank oxygen a signifi-cant decrease (P < .01), with reference to therate on air. On switching gas mixtures, the changein ventilation rate was extremely rapid. This wasparticularly true in passing to oxygen, when amarked slowing was often evident within 10 sec-onds, or four to five breaths. With the ventilationchanges there were corresponding changes in ar-terial CO2tension and pH. That is, with the lowerventilation under tank oxygen the CO2 rose andthe pH fell, while reverse changes occurred with15% 02. These secondary changes must havelimited the extent of the oxygen effect, since theywould force ventilation in the opposite direction.

It is of some interest to note that the meanchange of CO2 tension was only about 3 mm. Hgon changing from air to either gas mixture, whichwas less than might have been anticipated from thechange in ventilation. The short exposure timemay not have permitted the development of a fullchange in CO2 tension. In addition, it is possiblethat the different gas mixtures may have influencedthe rate of bicarbonate breakdown and conse-quently of CO2 excretion. From the rate of totalblood CO2 loss, it is apparent that lactic acid wasbeing formed rapidly. The rate of production oflactic acid during work can be markedly slowed byhigh oxygen mixtures or accelerated by low oxy-gen (20, 36, 38, 39).

The effect on ventilation of changing arterialoxygen tension is far greater during exercise ofthis severity than at rest. The effect on ventila-tion of inhaling high and low oxygen concentra-tions at rest has been studied by Dripps and Com-roe (40). They found a small, immediate drop of3.1%o in ventilation on passing from air to oxygen.On 12% oxygen, which produced a mean fall of17%o in arterial saturation, approximately corre-sponding to that of the present study, there was anaverage ventilation increase of 0.5 1./min.

The effect of changes in arterial pHThere has been some disagreement as to the role

played by arterial pH in the regulation of breath-

510


TABLE VII

Arterial blood pH, ventilation, and CO2 tensio inthe post-exercise period

Subject State Ventilation Arterial blood C2ObloH tension

(I./mins.,250 C.) (mm. Hg)

R. M.* Rest 8.8 7.37 41post-exercise 8.1 7.32 42

28 min.

E. S.* Rest 8.5 7.38 43post-exercise 9.7 7.34 43

22 min.

S. Rest 7.1 7.37 44post-exercise 8.4 7.33 45

6 min.

P. S. Rest 7.8 7.35 45post-exercise 10.7 7.29 47

18 min.

* Data repeated from Table IV.

ing during exercise. Barr (14, 41), Dill, Talbottand Edwards (8), and Nielsen (22) found littlecorrelation between pH and ventilation. Fromstudies on diabetic acidosis Kety and associates(42) have suggested that there is a threshold forrespiratory response at about pH 7.2. Betweenthis point and pH 7.0, ventilation increased as thehydrogen ion concentration rose. These investi-gators pointed out that the arterial CO2 tensionof their patients was simultaneously decreased.As emphasized by Gray (43) and Grodins (31)such a change could mask a stimulant effect whichmight otherwise appear at higher pH levels.

Some data on this point are available from thepresent study and are presented in Table VII. Inthe post-exercise period, ventilation and arterialCO2 tension in these subjects have returned sub-stantially to their former values, while the pH isstill .04 to .06 units below the control level. ThepH changes are not large, but it seems possible toconclude that small pH changes, produced in thisway, do not evoke much respiratory response.No such data are available, from this study, forlarge pH changes. As pointed out above, the find-ing that exercise does not generally increase re-spiratory sensitivity to CO2also implies that it doesnot cause a marked increase in sensitivity to a fallin arterial blood pH, since the pH was decreasedconsiderably both at rest and exercise by CO2 in-halation.

Individual variationIndividual variation is conveniently examined

by reference to Figure 1. There is a striking dif-ference among subjects in response to variousstimuli.

The most consistent response is the immediateincrease in ventilation on beginning exercise (Fig-ure 1B). Even so, this shows up to 100%o varia-tion among subjects, and its magnitude is notclosely related to the final ventilation volume. InFigure 1C is presented the abrupt fall in ventila-

H R M K D L C B G S T AV

FIG. 1. VARIATION IN PERFORMANCEOF INDIVIDUAL SUB-JECTS DURING EXERCISE

Columns at the extreme right represent the mean foreach category. The values of categories A, B, C, E, andF are the averages of observations during two separateexercise periods, except for subject B, who had only oneexercise period. A: ventilation during arterial bloodsampling in the latter part of the exercise period. B:abrupt increase in ventilation on beginning exercise. C:abrupt decrease in ventilation on stopping exercise. D:increase in ventilation (1./min.) per mm. Hg rise in ar-terial CO, tension resulting from administration of5%b CO, during exercise. E: difference between arter-ial CO, tensions at exercise and rest. F: fall in arterialblood pH during exercise. G: increase in ventilation re-sulting from administration of 15% 02 during exercise.H: decrease in ventilation resulting from administrationof tank oxygen during exercise.

511


tion on stopping work. Here there is even agreater variation. In some subjects there is littleor no immediate fall, while others show a fallgreater than the original increase. The initial in-crease, because of its rapidity, has been consideredneurogenic from the time of its description byKrogh and Lindhard, and the same etiologic reas-oning is evidently applicable to the abrupt fall.Since the discovery by Harrison, Calhoun, andHarrison (44) of respiratory reflexes from pas-sive limb motion, considerable attention has beendevoted to this phase of respiratory control, mostnotably by Comroe and associates (45, 46), whohave recently concluded that passive stretch re-flexes can play only a minor part in respiratoryregulation. Whether peripheral or central in ori-gin, the present effect is of considerable magnitude.It is quite possible that the actual magnitude of this"neurogenic factor" during exercise is not fullymeasured by the abrupt changes in ventilation onstarting and stopping exercise. A preliminary ex-amination of this possibility has been made in sub-ject H., who customarily shows an immediateincrease in ventilation of about 12 l./min. on start-ing exercise and no immediate decrease on stop-ping, once a large ventilation volume has beenreached. This subject was caused to exercise re-peatedly for short periods with only brief recoveryintervals. The immediate increase on passing fromrest to exercise amounted in the first trial to 181./min., in the second, to 28 l./min., and in thethird to 32 l./min., the last representing an im-mediate jump from 18 to 50 l./min. The possibil-ity exists that the neurogenic factor may increaseduring exercise and become much larger thanindicated by the average values for starting andstopping effects.

There is a very large range in sensitivity to ar-terial CO2 tension, as shown by Figure 1D, repre-senting ventilation increase per mm. of arterialCO2 tension when 5%b CO2 is administered duringexercise. Wide variation among subjects in re-sponse to carbon dioxide at rest has also been re-marked (29, 30). Figure 1E presents the dif-ference between arterial CO2 tension during thelatter part of exercise and the tension at rest.There is some correspondence between CO2 sen-sitivity (Figure 1D) and the behavior of the ar-terial CO2 tension during exercise. Where sen-sitivity is high, the CO2 tension tends to fall or to

rise only slightly during exercise. The large in-crease shown by D. is probably owing to anxietyand hyperventilation during the resting period,with a resultant subnormal CO2 tension at rest.An exception is provided by S., who was sensitiveto CO2, but still showed a considerable increase inCO2 tension during exercise. Evidently, the im-portance of CO2 as a respiratory regulator duringexercise is subject to considerable individual vari-ation.

The reaction to 15%b oxygen and tank oxygenis presented in Figures 1G and 1H. Once more,there was a considerable range of response. Inseveral cases there was much difference in sensi-tivity to the two mixtures. The response to tankoxygen was a little more consistent than that to15%oxygen.

DISCUSSION

In 1944 Comroe (47) reviewed the evidencedealing with respiratory regulation during exer-cise and concluded that a number of factors mustact together. Subsequent work, including thepresent, has confirmed the conclusion. Effortshave been made to quantitate the role played byvarious of these factors. Most of this work hasbeen done with well-trained subjects. It is gen-erally believed that neurogenic stimuli play somepart, although there is disagreement as to whetherreflexes originating from the limbs themselves areimportant or not (46, 48, 49). It has been held,on the one hand, that neurogenic stimuli are ofprimary importance in light to moderate exercise(50), and, on the other, that they are of little im-portance. Carbon dioxide has been invoked as amajor factor on the ground that exercise increasesthe responsiveness of the respiratory center to thisstimulus (22), and the significance of carbon di-oxide as a major factor has been denied on theground that the alveolar carbon dioxide tension re-mains unchanged or falls during exercise (47).pH change has not been considered of first-lineimportance by most observers (8, 22, 41), but itspossible importance has recently been asserted(31). Arterial oxygen tension has been dismissedas a regulatory factor because changes in arterialoxygen during exercise are slight, and change inarterial oxygen tension over a considerable rangeyields little respiratory response at rest (20). Arise in body temperature during exercise may stim-

512


ulate respiration, but this is slow to develop andprobably would not be very effective in less than 10minutes (51). Because it has often seemed diffi-cult to account for the magnitude of ventilationduring exercise, it has been recurrently postulatedthat an unidentified substance which acts as a re-

spiratory stimulant is produced by working mus-

cles.In the present study, using untrained subjects,

it is apparent that a number of stimuli were opera-

tive. Neurogenic factors, of whatever origin,played a significant part in initiating and maintain-ing ventilation. This stimulus may have beenmuch larger than can be estimated from initial in-creases and terminal decreases in ventilation.

Arterial carbon dioxide tension appears to havebeen effective in increasing ventilation during ex-

ercise in at least some subjects. Since the respira-tory center did not become highly sensitive to car-

bon dioxide during exercise, the effect as cal-culated after several minutes of exercise is notparticularly large except perhaps in subject S., whowas sensitive to carbon dioxide and had a sub-stantial increase in arterial tension during exercise.In those subjects who have a rapid, early decreasein blood CO2 content, presumably owing to lacticacid accumulation, a rising arterial CO2 tensionearly in exercise-may assist in increasing ventila-tion. As pointed out above, this may account forthe "secondary rise" in ventilation early in exer-

cise. In trained subjects this secondary rise hasnot been described, and trained subjects show little,if any, early increase in alveolar CO2 tension (8,22) corresponding to that found in the arterialblood of some of our untrained subjects.

The effect of changing the arterial oxygen ten-sion is of considerable interest. The ventilation re-

sponse is very rapid, and its magnitude is many

times greater than at rest. This effect appears inexercise which is severe for the subject, and theeffect becomes more pronounced as the severity in-creases (20, 24). The response is too immediateto make it likely to result from changes in bloodconcentration of some substance produced byanaerobically working muscles. It has been sug-

gested by Asmussen and Nielsen (20) that an

increase in oxygen tension could act by decreasingchemoreceptor sensitivity to such a substance. Ifso, one would need to postulate that the substancecontinues to be produced on oxygen at the same

rate as on air, although lactic acid production isdecreased by oxygen. This is because the resultsof these investigators show an immediate returnof ventilation to the same steady rate as beforewhen the subject is switched back to air. It seemsa little simpler to assume that the chemoreceptorsystem becomes more sensitive to the arterial oxy-gen tension itself, or that the respiratory centerbecomes more sensitive to the "tone" which isknown to exist at normal oxygen tensions in theresting state (40). In the latter case, the oxygeneffect could be regarded as a variety of neurogenicstimulus which becomes more active the more se-vere the exercise. There is some evidence in ani-mals for a respiratory drive from increased chemo-receptor tone, or increased responsiveness to theexisting tone, during electrically induced work(49, 52). The development of this effect duringexercise can make a significant contribution to thetotal ventilation. Normally, it can not providefor a very close regulation of ventilation becausethe effect does not vary greatly with a small changein oxygen tension. It is probably important inthe close regulation of breathing during exerciseat high altitudes and in patients with pulmonarydisease who have a decreased arterial oxygen ten-sion. For reasons outlined by Bohr (53) andBarcroft (54) and put to experimental trial byLilienthal, Houston, and others (15, 35, 55), thearterial oxygen tension can fall markedly duringexercise when it is already low at rest.

The large falls in arterial pH which occurredduring exercise in some of our subjects would beexpected to have some stimulant effect on respira-tion. Our experience, however, is similar to thatof others in finding no close correlation betweenventilation and pH in a given subject. As pointedout before, we find no evidence that the respiratoryresponse to a fall in pH becomes very great duringexercise.

Among the factors known to be operative in thepresent study, then, are a "neurogenic factor"which may increase during exercise, a changingarterial CO2 tension, the development of what isprovisionally regarded as an increased chemore-ceptor tone to oxygen, and a considerable fall inpH. Recently, Gray (43) and Grodins (31) haveapproached the problem of multiple respiratorystimuli by assigning quantitative values to thestimulant effects of CO2 tension, hydrogen ion con-

513


centration, and oxygen tension. If these valuesare accepted and the quantities measured, it is pos-sible to draw conclusions as to the intensity andmode of action of additional stimuli which are re-quired to account for the ventilation observed ina given situation. One of the most striking find-ings of the present study is the tremendous vari-ability among untrained subjects in the responseto different stimuli, as well as in the extent to whichvarious stimuli develop during exercise. Thissuggests that any respiratory formula be appliedonly with reservations to a single subject or evento small groups of subjects, particularly when theyare not in good physical training.

This great variability among normal subjectsin poor physical condition is in contrast with themuch more uniform performance of subjects ingood training. It is probable that still greater de-viations will occur in the course of disease. Pres-ent concepts of the relative importance of variousrespiratory stimuli in muscular exertion have beengained largely through the study of well-condi-tioned subjects. It does not appear likely that theseconcepts can be carried over intact to persons withacute or chronic disease.

SUMMARY

1. Observations have been made on ventilationand arterial blood changes in 11 untrained sub-jects during short periods of moderately severetreadmill exercise. Particular attention was paidto the ventilation changes at the start and con-clusion of work, to the course of ventilation inearly exercise, to the effect on ventilation of breath-ing gas mixtures having, respectively, an increasedCO2 content, a low oxygen, and a high oxygencontent, and to associated changes in arterial CO2tension, pH and per cent oxygen saturation.

2. With untrained subjects there may be largechanges during exercise in arterial CO2 content,CO2 tension, and pH, and these may begin to de-velop very early in exercise.

3. Exercise did not increase significantly thesensitivity of the respiratory center to arterial CO2tension, as measured by increase in ventilation permm. Hg increase in CO2 tension.

4. An attempt is made to estimate in semi-quan-titative terms the role played by various respira-tory stimuli in regulating ventilation during exer-

cise. These stimuli are a "neurogenic" effect; anoxygen effect, which is provisionally regarded asan increase in sensitivity of the chemoreceptorsystem to oxygen tension; fall in blood pH; andarterial CO2 tension. There is great variationamong untrained subjects in the relative impor-tance of these factors.

REFERENCES

1. Hickam, J. B., and Frayser, R., Spectrophotometricdetermination of blood oxygen. J. Biol. Chem.,1949, 180, 457.

2. Rosenthal, T. B., The effect of temperature on thepH of blood and plasma in vitro. J. Biol. Chem.,1948, 173, 25.

3. Peters, J. P., and Van Slyke, D. D., QuantitativeClinical Chemistry. Vol. II, Methods. The Wil-liams & Wilkins Co., Baltimore, 1932.

4. Van Slyke, D. D., and Sendroy, J., Jr., Studies ofgas and electrolyte equilibria in blood. XV. Linecharts for graphic calculations by the Henderson-Hasselbalch equation, and for calculating plasmacarbon dioxide content from whole blood content.J. Biol. Chem., 1928, 79, 781.

5. Dill, D. B., Daly, C., and Forbes, W. H., The pK' ofserum and red cells. J. Biol. Chem., 1937, 117, 569.

6. Hastings, A. B., Sendroy, J., Jr., and Van Slyke, D.D., Studies of gas and electrolyte equilibria in blood.XII. The value of pK' in the Henderson-Hassel-balch equation for blood serum. J. Biol. Chem.,1928, 79, 183.

7. Dill, D. B., Lawrence, J. S., Hurxthal, L. M., andBock, A. V., The carbon dioxide equilibrium inalveolar air and arterial blood. III. Exercisingsubjects. J. Biol. Chem., 1927, 74, 313.

8. Dill, D. B., Talbott, J. H., and Edwards, H. T.,Studies in muscular activity. VI. Response of sev-eral individuals to a fixed task. J. Physiol., 1930,69, 267.

9. Comroe, J. H., Jr., and Walker, P., Normal humanarterial oxygen saturation determined by equilibra-tion with 100 per cent oxygen in vivo and by theoximeter. Am. J. Physiol., 1948, 152, 365.

10. Wood, E. H., and Geraci, J. E., Photoelectric de-termination of arterial oxygen saturation in man.J. Lab. & Clin. Med., 1949, 34, 387.

11. Drabkin, D. L., and Schmidt, C. F., Spectrophoto-metric studies. XII. Observation of circulatingblood in vivo, and the direct determination of thesaturation of hemoglobin in arterial blood. J. Biol.Chem., 1945, 157, 69.

12. Dill, D. B., Edwards, H. T., and Consolazio, W. V.,Blood as a physicochemical system. XI. Man atrest. J. Biol. Chem., 1937, 118, 635.

13. Dill, D. B., Edwards, H. T., Folling, A., Oberg, S. A.,Pappenheimer, A. M., Jr., and Talbott, J. H.,Adaptations of the organism to changes in oxygenpressure. J. Physiol., 1931, 71, 47.

514

RESPIRATORYREGULATION DURING EXERCISE

14. Barr, D. P., Himwich, H. E., and Green, R. P., Stud-ies in the physiology of muscular exercise. I.Changes in acid-base equilibrium following shortperiods of vigorous muscular exercise. J. Biol.Chem., 1923, 55, 495.

15. Lilienthal, J. L., Jr., Riley, R. L., Proemmel, D. D.,and Franke, R. E., An experimental analysjs inman of the oxygen pressure gradient from al-veolar air to arterial blood during rest and exer-cise at sea level and at altitude. Am. J. Physiol.,1946, 147, 199.

16. Galdston, M., and Wollack, A. C., Oxygen and car-bon dioxide tensions of alveolar air and arterialblood in healthy young adults at rest and after ex-ercise. Am. J. Physiol., 1947, 151, 276.

17. Bock, A. V., Dill, D. B., Hurxthal, L. M., Lawrence,J. S., Coolidge, T. C., Dailey, M. E., and Hender-son, L. J., Blood as a physicochemical system. V.The composition and respiratory exchanges of nor-mal human blood during work. J. Biol. Chem.,1927, 73, 749.

18. Lange, H. F., The normal plasma protein values andtheir relative variations. Acta med. Scandinav.,1946, Suppl. 176, 1.

19. Nylin, G., The effect of heavy muscular work on thevolume of circulating red corpuscles in man. Am.J. Physiol., 1947, 149, 180.

20. Asmussen, E., and Nielsen, M., Studies on the regu-lation of respiration in heavy work. Acta physiol.Scandinav., 1946, 12, 171.

21. Terrell, E. S., and Robinson, S., The acid-base equi-librium of the blood in exercise. Am. J. Physiol.,1942, 137, 742.

22. Nielsen, M., Untersuchungen iiber die atemregulationbeim menschen. Skandinav. Arch. f. Physiol., 1936,74, Suppl. 10, 83.

23. Barr, D. P., and Himwich, H. E., Studies in thephysiology of muscular exercise. III. Developmentand duration of changes in acid-base equilibrium.J. Biol. Chem., 1923, 55, 539.

24. Briggs, H., Physical exertion, fitness and breathing.J. Physiol., 1920, 54, 292.

25. Krogh, A., and Lindhard, J., The regulation of res-piration and circulation during the initial stages ofmuscular work. J. Physiol., 1913, 47, 112.

26. Asmussen, E., and Nielsen, M., Studies on the initialchanges in respiration at the transition from restto work and from work to rest. Acta physiol.Scandinav., 1948, 16, 270.

27. Rahn, H., and Otis, A. B., Continuous analysis of al-veolar gas composition during work, hyperpnea,hypercapnia, and anoxia. J. Appl. Physiol., 1949,1, 717.

28. Bruce, R. A., Lovejoy, F. W., Jr., Pearson, R., Yu, P.N. G., Brothers, G. B., and Velasquez, T., Normalrespiratory and circulatory pathways of adaptationin exercise. J. Clin. Invest., 1949, 28, 1423.

29. Heller, E., Killiches, W., and Drinker, C. K., Theevaluation of 5 and 7 per cent carbon dioxide mix-

tures as respiratory stimulants. J. Indust. Hyg.,1929, 11, 293.

30. Dripps, R. D., and Comroe, J. H., Jr., The respiratoryand circulatory response of normal man to inhala-tion of 7.6 and 10.4 per cent CO2 with a compari-son of the maximal ventilation produced by severemuscular exercise, inhalation of CO2 and maximalvoluntary hyperventilation. Am. J. Physiol., 1947,149, 43.

31. Grodins, F. S., Analysis of factors concerned in regu-lation of breathing in exercise. Physiol. Rev., 1950,30, 220.

32. Nielsen, M., and Hansen, O., Maximale k6rperlicheArbeit bei Atmung 0,-reicher Luft. Skandinav.Arch. f. Physiol., 1937, 76, 37.

33. Hewlett, A. W., Barnett, G. D., and Lewis, J. K., Theeffect of breathing oxygen-enriched air during ex-ercise upon pulmonary ventilation and upon the lac-tic acid content of blood and urine. J. Clin. Invest.,1926, 3, 317.

34. Christensen, E. H., Sauerstoffaufnahme und respira-torische Funktionen in grossen Hohen. Skandinav.Arch. f. Physiol., 1937, 76, 88.

35. Houston, C. S., and Riley, R. L., Respiratory and cir-culatory changes during acclimatization to high al-titudes. Am. J. Physiol., 1947, 149, 565.

36. Lundin, G., and Strom, G., The concentration of bloodlactic acid in man during muscular work in rela-tion to the partial pressure of oxygen of the inspiredair. Acta physiol. Scandinav., 1947, 13, 253.

37. Str6m, G., The influence of anoxia on lactate utiliza-tion in man after prolonged muscular work. Acta.physiol. Scandinav., 1949, 17, 440.

38. Asmussen, E., Dobeln, W., and Nielsen, M., Bloodlactate and oxygen debt after exhaustive work atdifferent oxygen tensions. Acta. physiol. Scandinav.1948, 15, 57.

39. Tepperman, J., and Tepperman, H. M., On the bloodlactic acid response to measured exercise in hy-poxic human subjects. J. Clin. Invest., 1948, 27,176.

40. Dripps, R. D., and Comroe, J. H., Jr., The effect ofthe inhalation of high and low oxygen concentra-tions on respiration, pulse rate, ballistocardiogramand arterial oxygen saturation (oximeter) of nor-mal individuals. Am. J. Physiol., 1947, 149, 277.

41. Barr, D. P., Studies in the physiology of muscularexercise. IV. Blood reaction and breathing. J.Biol. Chem., 1923, 56, 171.

42. Kety, S. S., Polis, B. D., Nadler, C. S., and Schmidt,C. F., The blood flow and oxygen consumption ofthe human brain in diabetic acidosis and coma.J. Clin. Invest., 1948, 27, 500.

43. Gray, J. S., The multiple factor theory of the controlof respiratory ventilation. Science, 1946, 103, 739.

44. Harrison, W. G., Jr., Calhoun, J. A., and Harrison,T. R., Afferent impulses as a cause of increasedventilation during muscular exercise. Am. J.PlWsiol., 1932, 100, 68.

515


45. Comroe, J. H., Jr., and Schmidt, C. F., Reflexes fromthe limbs as a factor in the hyperpnea of muscularexercise. Am. J. Physiol., 1943, 138, 536.

46. Bahnson, E. R., Horvath, S. M., and Comroe, J. H.,Jr., Effects of active and "passive" limb movementsupon respiration and 02 consumption in man. J.

Appl. Physiol., 1949, 2, 169.47. Comroe, J. H., Jr., The hyperpnea of muscular exer-

cise. Physiol. Rev., 1944, 24, 319.48. Asmussen, E., Nielsen, M., and Wieth-Pedersen, G.,

Cortical or reflex control of respiration during mus-

cular work? Acta physiol. Scandinav., 1943, 6,168.

49. von Euler, U. S., and Liljestrand, G., The regulationof respiration during muscular work. Acta physiol.Scandinav., 1946, 12, 268.

50. Asmussen, E., Christensen, E. H., and Nielsen, M.,Humoral or nervous control of respiration duringmuscular work? Acta physiol. Scandinav., 1943,6, 160.

51. Nielsen, M., Die regulation der Korpertemperatur beiMuskelarbeit. Skandinav. Arch. f. Physiol., 1938,79, 193.

52. Heymans, C., Jacob, J., and Liljestrand, G., Regula-tion of respiration during muscular work, as stud-ied on the perfused isolated head. Acta physiol.Scandinav., 1947, 14, 86.

53. Bohr, C., Uber die spezifische Tatigkeit der Lungenbei der respiratorischen Gasaufnahme und ihr Ver-halten zu der durch die Alveolarwand stattfinden-den Gasdiffusion. Skandinav. Arch. f. Physiol.,1909, 22, 221.

54. Barcroft, J., The Respiratory Function of the Blood.Part I. Lessons from High Altitudes. CambridgeUniversity Press, London, 1925.

55. Lilienthal, J. L., Jr., Riley, R. L., and Proemmel, D.D., Arterial oxyhemoglobin saturations at criticalpressure-altitudes breathing various mixtures ofoxygen and nitrogen: with a note on the effect ofexercise. Am. J. Physiol., 1946, 145, 427.

516

Documents

RESPIRATORY REGULATION DURING EXERCISE IUNCONDI- TIONED