NORMAL RESPIRATORY AND CIRCULATORYPATHWAYS OF ADAPTATION IN EXERCISE

Robert A. Bruce, … , George B. Brothers, Tulio Velasquez

J Clin Invest. 1949;28(6):1423-1430. https://doi.org/10.1172/JCI102207.

Research Article

Find the latest version:

http://jci.me/102207-pdf

NORMALRESPIRATORYAND CIRCULATORYPATHWAYSOFADAPTATIONIN EXERCISE1

By ROBERTA. BRUCE,2 FRANKW. LOVEJOY, JR.,2 RAYMONDPEARSON,'PAUL N. G. YU,2 GEORGEB. BROTHERS,3 AND TULIO VELASQUEZ4

(From the Chest Laboratory of the Department of Medicine of the University of RochesterSchool of Medicine and Dentistry and the Medical Clinics of Strong Memorial and

Rochester Municipal Hospitals, Rochester, New York)

(Received for publication January 17, 1949)

Rahn and Otis (1) have described the normalchanges in the alveolar gases during work anddemonstrated the quantitative relationships be-tween alveolar oxygen, carbon dioxide, respira-tory quotient and ventilation by means of continu-ous recording of gas analyses. They were able todelineate the pathways of change which are con-trolled by the excretion of carbon dioxide, and ob-serve the effects of hyperpnea, hypoxia, hypo-ventilation and CO2 breathing as well. Pelnar(2), at first independently and later in conjunc-tion with Rahn, observed the changes in expiredair composition in normal subjects as well as inpatients with cardiorespiratory impairment in rela-tion to work. By means of a simple diagram re-lating the changes in 02 and CO2 content of ex-pired air with the respiratory quotient during thestress of exercise, he felt that he could delineaterespiratory function better than by any other meansin a manner that paralleled the intensity of dyspneain patients. Between the extremes of a) normalperformance with a large circular curve and b) theabnormal performance with no change in thesevalues during exercise, Pelnar found a series ofresults which were characteristic of the variedstages of dyspnea. Thus he was able to evaluatefunctional performance largely in terms of changesin the respiratory gases during exercise. In con-trast to this are the older methods reported byKaltreider and McCann (3) who relied chiefly onventilation volumes and ventilation indices to ap-praise pulmonary capacity during work. BothRahn and Otis (1) and Pelnar (2) express the

1 Aided by grants from the Hochstetter Fund, LovejoyFund, Fluid Research Fund, and the Rochester Gas andElectric Company Fund.

2 Bertha Hochstetter Buswell Research Fellow inMedicine.

3 Fellow of the General Education Board.4 Rockefeller Visiting Fellow in Medicine.

view that the level of ventilation is controlled inlarge measure by the CO2 when breathing normalambient air in which the strong stimulus of hy-poxia is lacking. The hypoxic stimulus of breath-ing 15%o oxygen in normals or hypoxemia frompulmonary disease in patients clearly inducesgreater ventilation levels, however.

With these considerations in mind we have at-tempted to repeat these studies in both normalsubjects and patients with disease to ascertain notonly the normal respiratory pathways, but also thecirculatory pathways during exercise and recovery.Thus the inter-relationships between the two canbe considered in a way to demonstrate how theadaptations of each complement and spare theother. The techniques employed consist essen-tially of controlled work-loads by means of walk-ing on a motor-driven treadmill and continuousgas analyses of expired air as well as determina-tions of electrocardiographic changes in rate andpattern, blood pressure and arterial oxygen satu-ration (oximeter) (Figure 1). Observations havenot been limited to the basal state in order to ob-tain an appraisal of performance to a standardizedstress of exercise in relation to ordinary circum-stances in regard to diet, clothing and activity.In a separate report the statistical analyses of vari-al)ility of normal performances under these condi-tions is presented (4). For the purpose of defin-ing the normal pathways in this report, only themean values of all pertinent observations obtainedat minute intervals before, during, and after ex-ercise are graphically considered in relation toeach other.

METHODS

Thirty-five normal adults were used for this investi-gation (20 men, 15 women, ranging in age from 20 to57 years). Each was allegedly healthy and gainfullyemployed, and none had any detectable sequelae of anyprevious cardiac or pulmonary disease. All rested before

1423

1424 R. BRUCE, F. LOVEJOY, JR., R. PEARSON, P. YU, G. BROTHERS, AND T. VELASQUEZ

EQUIPMENT FOR EXERCISE TOLERANCE TESTINGOBSERVER

... L N,-BLOOD

PRESSURECUFF

OXIMETERCONTROL

ELECTRO-CARDIOGRAPH

CONTROLS

TREADMILL:HYDRAULIC

LIFT

PATIENT EAR PIECE- - - FACE MASK

- EXHAUST HOSEA_.......ELECTRODE BELT

GAS METER

_-°./OXIMETER GALV...FLOW METERS..OXYGEN ANALYZERS

CARBON DIOXIDESy -~~n GALVANOMETER

COUNTERS FORHEART RATE,RESPIRATORY

RATE, ANDVENTILATION

tVER PULSE RELAY

FIG. 1

Arrangement of equipment for exercise tolerance testing. One observer remains onthe treadmill platform with the patient, records blood pressure, electrocardiogram, symp-toms and signs. The other observer seated at the continuous analyzer records the heartrate, respiratory rate, ventilation volume, gas concentrations, and oxinmeter at one minuteintervals.

the test period and then were observed during 10 minutesof rest sitting in a chair, 10 minutes level walking on thetreadmill at the rate of 2.6 mph (12 rpm of the conveyorbelt or 70 meters per minute velocity), and again for 10minutes of recovery while sitting in a chair. Each waslightly clothed, and the barometric pressure and tem-perature were the prevailing room air values (747.3 +1.1 mm. Hg, 0.15% coefficient of variability) and 24.9 +1.9° C (7.9% coefficient of variability).

The systemic blood pressure was obtained from ananeroid sphygmomanometer at frequent intervals, evenduring walking. Changes in arterial oxygen saturation,as well as variations in characteristics of the ear (greenfilter), were followed by an improved Millikan oximeter.Three electrodes attached to a rubber belt around thesubject's chest for "ground," "indifferent" (right scapula)and "apical" contacts permitted registration of the chestlead electrocardiogram on a direct-writing instrumentas well as continuous recording of the pulse from theamplified QRS spikes. Expired air gas data were ob-tained from a modified A-13A oxygen mask through a11/2 inch corrugated rubber hose mixing chamber con-ducting to a continuous gas analyzer. Ventilation vol-umes were electrically recorded from a low resistance(differential pressure less than 1 inch of water) wet-test gasometer with accuracy of + 1.65% at 100 literper minute velocities. The respiratory rate was re-corded by activation of a variable time-delay relay coun-ter. Gas analyses were made continuously on aliquot(saturated with water) samples (at flow rates of 100-125 nil per minute regulated by calibrated rotameters),

by means of a thermal-conductivity cell for carbon-dioxide expressed in volumes per cent; and a suitablePauling oxygen analyzer which determined the partialpressure of oxygen. The calibration of each was estab-lished and checked against room air prior to each test.The terminal portion of each expired breath (mid-capacity air)5 was continuously sampled by the alternatepulsations of a very thin rubber balloon incorporated inthe face piece of the mask for analyses of oxygen tensionby a second Pauling analyzer and the derivation ofchanges in the physiological dead space. These valuesincluded the dead space of the mask which was reducedby careful packing with cotton. At the rates of flowstated above, the aliquot gas samples reached the ana-lyzers after a lag of 10-20 seconds for expired air and40-50 seconds for the mid-capacity air. Since the re-sponse of the CO, analyzer was delayed to 99% peakreading in five minutes from zero concentration, acutechanges in CO2 concentration could not be followed withaccuracy. Hence no correction for inspired air volumefrom differences in nitrogen concentration in expired airwere made, and estimates of the expired air respiratoryquotients were only approximated values (averagingabout 4% too high), (Figure 2).

5 In the traditional sense, alveolar air has the composi-tion of that volume of air which is obtained at the endof forced expiration, hence includes in the last portion ofthe reserve air partition of the pulmonary capacity. "Mid-capacity" air corresl)onds in gaseous composition tothat obtained at the end of normal expiration.

PATHWAYSOF ADAPTATION IN EXERCISE

MAGNITUDE OF ERROR IN OXYGEN CONSUMPTIONCALCULATED FROMEXPIRED AIR

- DIRECrLY

. CORRECTEDFOR

M1LJSQ.M/M

300.2001

gooT ,,,,1 _1001

MINUTES 5f 10 15 2O 25 30REST EXERCISE RECOVERY

FIG. 2The differences in oxygen consumption calculated

directly are compared with those obtained by correctionfor changes in nitrogen tension. The magnitude of errorby the short cut calculation is considered almost negligi-ble in relation to the change in oxygen consumptioninitiated by exercise.

The reproducibility and accuracy of gas analyses bythis method was estimated as variations of duplicateanalyses from an ideal value of zero in terms of standarddeviations as follows:

a) Haldane techniqueb) Continuous analyzerc) Between "a" and "b"

Co2 vol. %+ 0.054+ 0.160+ 0.360

02 vol. %+ 0.058+ 0.028+ 0.083 6

Consequently reliance was placed upon the oxygenvalues. Suitable conversion tables permitted rapid de-termination of the volumes per cent of oxygen absorbed(respiratory efficiency) from air from the observed ex-pired air pO2 reading and the dry barometric pressure atthe operating temperature. Similarly a factor was es-tablished for each subject that permitted conversion ofthe ventilation volumes observed into liters per squaremeter of body surface area at 760 mm. Hg and 00 C dry.The oxygen consumption was estimated from the productof the corrected minute ventilation and the respiratoryefficiency and ranged between 96-101% (depending onthe R.Q.) of the true value based upon corrections fornitrogen differences in inspired and expired air. Allprimary observations were recorded and graphed at oneminute intervals (Figure 3). From these data, averagevalues per period, dead space volumes, oxygen debt,ventilation indices, etc. were calculated.

Accessory procedures routinely employed were the de-termination of maximum mask breathing capacity bymeans of voluntary hyperventilation for 30 seconds atthe completion of the above test, changes in oximeter

6 The standard error for oxygen analyses determinedby the Haldane technique was 3.3 times greater than thatobtained by the Pauling oxygen analyzer.

readings with breath-holding, and estimation of lung-to-ear circulation time with the oximeter.

RESULTS

Inasmuch as statistical analyses showed no sig-nificant differences in the data according to thesex of the subjects, all the consecutive one min-ute observations for the mean values of both sexeshave been compiled together. The resting values(allowing for differences in basal state, tempera-

EXERCISE PERFORMANCEORMALMALE)13O.I110 f

TOT6.0 L/IS.N/K

6.0 (STP)4.0

2.0

::O .L./KdL.O,400 NML/SQ.M./I.

300 (STP)300 p

too.100

I10

SATURATI0N

I?7 VOL% OXYGEN.----s-- t~

is t

145 VOL. % CARSON

DIOXIDE4 . t

L;4MINUTES 5

REST

BLOODPRE

HEARTRATE

VENTILATION

VENTILATION

EQUIVALENTS

OXYGEN.0 CONSUMPTION

lIZ

RESPIRATORYRATE

._ ~~OXIMETE

EXPIRED AIR

AIR

EXPIRED AIR

10 15 0to 2 30

EXERCISE RECOVERY

FIG. 3Typical exercise performance graph of a normal male

subject. The values of "11%" and "1.0" adjacent to therecovery curve for oxygen consumption refer to oxygendebt and half-time recovery time, respectively. The valueof minute-by-minute observations of multiple measure-ments is illustrated by the fleeting changes in heart rate,ventilation, and oxygen consumption occurring on theseventh, 19th, and 28th minutes. These variations wereinvoluntary reactions to environmental stimuli initiatedby visitors during the period of observation. The subjectwas unaware of these variations.

1425

..a.

1426 R. BRUCE, F. LOVEJOY, JR., R. PEARSON, P. YU, G. BROTHERS, AND T. VELASQUEZ

RATES OF ADAPTATION

FIG. 4The rates of adaptation to changing states of activity

are compared on the basis of per cent change from thepreceding mean one minute values for ventilation volume,oxygen consumption, volumes per cent of oxygen ab-sorbed from expired air and heart rate for the first fourminutes of exercise and recovery. The 10th minute repre-

sents changes from standing. Note that ventilation andoxygen consumption show the greater changes first, andthat the change in volumes per cent of oxygen absorbed isdelayed, causing the final increment in oxygen consump-

tion during exercise.

ture and pressure) are inl accord Nwith those re-

ported by others.The rates of changes in the observations of

ventilation, oxygen consumption, respiratory effi-ciency (volumes per cent of oxygen absorbed)and the heart rate are shown in Figure 4. Thesechanges are expressed as percentage changes inthe several factors from the immediately precedingminute during the periods of adaptation to andfrom a steady state of exercise. The values for the10th minute of observation represent the changesinduced by standing up and are of smaller magni-tude than those for walking. As a result of exer-

cise both ventilation and heart rate promptly rise,but the greatest change is confined to the firstminute. The oxygen consumption initially risesin proportion to the ventilatory response, butthereafter it continues to rise during the next min-ute because of the greater respiratory efficiency.By the 13th minute of observation all factors haveapproached a new and relatively stable range

during the steady state of exercise. In recovery

the reverse changes promptly occur, but in this

instance there are two apparent differences: a)there is no lag in decline of respiratory efficiency,and b) the heart rate is restored toward the rest-ing level more rapidly than the other factors.

There are changes in the physiological deadspace along with the increments in the intid-ca-pacity ventilation associated with exercise (Fig-ure 5). The latter shows an almost linear riseduring the first three minutes of exercise as wellas a linear decline in three minutes during re-covery from exercise. During the last seven min-utes of exercise there is a just perceptible increasein the mid-capacity ventilation. The slight dis-crepancies between total ventilation and mid-capacity ventilation appear to be related to fluctua-tions in the physiological dead space. On theaverage, throughout the three periods of observa-tion the dead space represents 21% of the tidalvolume. Both the relative and absolute values forthe dead space are lower than those reported byothers (about 26 to 30% [ 5, 6] ), and recent im-provements in the face mask sampling device yieldsomewhat higher dead space values indicating atechnical error as partly responsible for these dis-crepancies. Although the dead space shows a

MEAN INTRAPULMONARYVENTILATION14Il MID-CAPACITYIo VENTILATION

150 DEAD SPACE

MINUTES 5 10 15 to 25 30REST EXERCISE RECOVERY

.,24XS2002.S

U 100 .oo o o

4 200 460 600 500 1000TIDAL AIR (ML..STPS)

FIG. 5Relationships of mean intrapulmonary ventilation,

physiological dead space, and tidal air volume in ambientvalues, B. T. P. S., in normal subjects.

%CHANGEFROMPRECEDING MINUTE

60 --- VENTILATION

.50 .-- OXYGENCONSUMPTION

40 -- VOL.% 02 ABSORBED

30.DO - HEART RATE

.20I10

-201-401-4

213 20 22 24 MINUTES

PATHWAYSOF ADAPTATION IN EXERCISE

MEAN RESPIRATORY PATHWAYS

511R5 R

"ST

.5s -5.0 -45 -4.0 .35VOL%OXrff £ASORK1 FROM AIR

'. '1

0.0 * LOREIPIATORY QUOTINT

LSQ/5.NIM. VNTILATION EOIALITVCNTILATION 0o z t.

7.0'I LZM

T4.Ol. .

REST

0oo 00 500 400 500

OY0M COBSUPTI"AS

&941 T1i

VOL.% AMDIOXIDEEXCRETED

courtf #

ERoS PATHWAYSOVERYMATceY

FIG. 6The mean respiratory pathways in normal subjects illustrate the various

inter-relations between ventilation, oxygen consumption, ventilation equiva-lents for oxygen, respiratory quotient, and changes in activity from restto exercise and back to rest again. The arrows indicate the direction of theloops, and the numbers adjacent to the loops indicate the specific minutes ofobservation during the testing procedure. The iso-R.Q. lines are indicatedby the background grid; note that the ultimate exercise R.Q. is slightlylower than the resting value. (Inasmuch as the oxygen consumption anddead space change with activity, iso-ventilation lines cannot be applied tothis graph). Note that the ventilation equivalent for oxygen is lower dur-ing exercise than at rest, indicating increased pulmonary capillary flow andmore effective oxygenation of the pulmonary arterial blood. See text forfurther discussion.

disproportionately large value early in exercisefollowed by a transient decline, the average deadspace value during exercise is 62%o larger thanthe resting values.

The mean respiratory pathways of adaptationfrom a steady state of rest to that of exercise andthe converse during recovery are shown in Figure6. The oxygen consumption at first increases pro-portionately with the increments in ventilation,but a little later it rises more rapidly because ofgreater respiratory efficiency (volumes per centof oxygen absorbed, or reciprocally, diminishingventilation equivalents for oxygen). The im-proved respiratory efficiency is a function of moreeffective alveolar ventilation and circulation aswell as alterations in blood flow and composition(see below). Concurrently the physiological deadspace enlarges. After the 13th minute of observa-tion (third minute of exercise), the estimated

oxygen consumption gradually declines 5%o asthe respiratory efficiency slowly recedes (Figure6) with a perceptible rise in mid-capacity venti-lation (Figure 5). On the average there isa 22%o increase in respiratory efficiency at thislevel of work. The estimated R.Q. of expired airshows wide fluctuations during the adaptive phasesbecause of the differences in rates of change inthe several factors representing ventilation andcirculation. The direction of change in the R.Q.is always a clockwise loop showing a low valueearly in exercise as the highest values for respira-tory efficiency are achieved by the 13th minute ofobservation (Figure 6). During steady states ofperformance, either rest or activity, the expired airR.Q. may be considered representative of tissuemetabolism throughout the body. With this levelof work, the R.Q. does not rise above the restingvalue as it does with more severe levels of work

1427

wwv.w ..W-

1428 R. BRUCE, F. LOVEJOY, JR., R. PEARSON, P. YU, G. BROTHERS, AND T. VELASQUEZ

MEAN CIRCULATORY PATHWAYS

FIG. 7The mean circulatory pathways in normal subjects il-

lustrate the various inter-relations of heart rate, oxygen

consumption, pulse equivalents for oxygen, amount ofoxygen carried per heart beat and electrocardiographicK constant during changes in activity from rest to exer-

cise, and back to rest again. Note that the heart raterises rapidly during the first minute, and maintains a

plateau thereafter during exercise. The decline in pulseequivalents indicates a rise in stroke volume. The ar-

rows indicate the direction of the loops, and the numbersadjacent to the loops indicate the specific minutes ofobservation during the testing procedure.

(7). With the onset of recovery the R.Q. of ex-

pired air transiently exceeds 1.0 as retained CO2continues to be excreted in excess of the rate ofoxygen absorption reduced with diminished tis-sue demands for oxygen.7 Simultaneously thelevel of ventilation is maintained above the tissuedemands by the stimulus of retained C02, andconcurrently most of the aerobic oxygen debt isrepaid.

The mean circulatory pathways, based upon

consecutive observations of minute heart rate,oxygen consumption, oxygen transport per heartbeat, and systolic time are shown in Figure 7.The systolic time is derived from the (chest lead)

7 Pelnar utilizes the abnormally high R.Q. values dur-ing steady states of either rest or moderate activity torecognize the presence of hyperventilation, beyond themetabolic needs of the tissues, caused by neuropsychogenicfactors. The usage of the term "respiratory quotient"in this sense may be unfortunate, especially when con-

sideration is given to the limitations in accurately mea-

suring the concentration of carbon dioxide by the presentauthors. Possibly the term "respiratory ratio" or ratioof carbon dioxide excreted to oxygen absorbed would bemore appropriate.

electrocardiographic changes in the K constant,or observed QT interval divided by the squareroot of the cycle length to correct for changes dueto rate. The oxygen transport is derived fromthe following equations:

(a) Cardiac Output= Heart rate X stroke volume

Oxygen consumptionA-V 02 difference

and by substitution:

(b) Oxygen TransportOxygen consumption

Heart rate= Stroke volume X A-V 02 difference.

The oxygen transport also represents the averagerate of oxygen transfer from either alveolar airto mixed venous blood, or arterial blood to periph-eral tissues, per unit of physiological time of oneheart beat. Statistical analyses of rates of changein large series of observations on normal subjectsshow that there is a 7% increase in cardiac outputfor every 10%o increment in oxygen consumption,and that the stroke volume shows somewhatgreater increases than the A-V oxygen differencewith increasing cardiac output (8). Similarlyanalyses of the present series of mean values showsa linear relationship of a high order of significancebetween oxygen consumption and oxygen trans-port (r = 0.938). With the initiation of exer-cise the heart rate promptly accelerates during thefirst minute of exercise and thereafter remains ona plateau (Figure 4). Since the oxygen consump-tion continues to rise, and hence the cardiac out-put, further increases in cardiac output must beachieved by larger stroke volumes. The rates ofchange in oxygen transport show the greatest in-crease during the second minute of exercise whenthe A-V oxygen difference probably is also widen-ing due to a reduction of mixed venous oxygencontent caused by blood returning from the exer-cising muscles (7). Concurrently the respira-tory efficiency is attaining its highest and the R.Q.its lowest values. During these adaptive changesin circulation (Figure 3) there are correspondingchanges in the systolic time manifested by theelectrocardiographic K constant. Initially withexercise there is an appreciable rise during thefirst minute followed by gradual fluctuations dur-

OXYSEN TRARSPORTPER HEART SEAT

4.0 ML.1S0.N./N.

3.5 I- 3.0 - -

.5 REST

.AO .4t .44 .4A .4A 100 too 500 400 5ooMECTCANSIOSRAPINIO KR OXYGEN CONSUMPTION MLS./ZMfN.

X- 110 ,,X R lEXERCISE E

a,- so', REST ~~HEART RATE il 4'0 It5

PULSE EQUIVALENTS

PATHWAYSOF ADAPTATION IN EXERCISE

ing the remainder of exercise; on the average theexercise systolic time is 7%o greater than that ob-served during the resting period. With recoverythe K value falls during the 21st and 22nd minuteof observation to 9% below the exercise range andthen gradually resumes the resting value. In noneof the normal subjects studied at this level ofwork were there any significant alterations of theQRScomplexes, ST segments, or T waves, otherthan the effects of a more rapid rate; i.e., nochanges in the electrocardiogram that could beassociated with possible or probable insufficiencyof the coronary circulation.

DISCUSSION

The continuous recording of the heart rate elec-trocardiographically, ventilation volume, and re-spiratory efficiency of expired and mid-capacityair, as well as the carbon dioxide concentrationduring changes in the metabolic demand initiatedby exercise greatly supplements the existing in-formation about either ventilatory performance oralterations in gaseous composition. By thesemeans the pathways of respiratory and circula-tory adaptation to the stress of exercise can be ob-served in terms of degrees, as well as rates, ofchange in a manner that demonstrates the com-plementary sparing action of the several variables.Appreciable uniformity of results can be approxi-mated by control of the work-load in proportion tothe body size by means of walking on a motor-driven treadmill. These studies have not beenperformed under basal conditions in order to ob-serve the responses in both normal subjects andpatients with cardio-respiratory diseases underordinary circumstances of daily activity. Ananalysis of the range of variability and reproduci-bility of results in normals and alterations pre-sented by patients with disease will be the sub-ject of later reports. The resting values observedby these methods are in reasonable accord withthose observed by others employing the Tissot-Haldane methods (9). The pathways of responsesfor expired air are similar to the findings of Rahnand Otis (1) for normal alveolar pathways in ex-ercise as well as those of Pelnar (2) for thechanges in expired air composition initiated byexercise.

These studies demonstrated the followingchanges which regularly occur with exercise:

ventilation and heart rate promptly rise to ahigher plateau, but the heart rate levels off by thesecond minute of exercise whereas the mid-ca-pacity ventilation continues to increase in a linearfashion for three minutes. The R.Q. of expiredair initially declines with exercise as the respira-tory efficiency (denominator, or volumes per centof oxygen absorbed) rises and overshoots thesteady state value by the third minute of exercise(Figure 6). This transient change is an expres-sion of greater alveolar ventilation and circulation(together with enlarged physiological dead space)as well as greater diffusion of oxygen into morerapidly circulating blood which has a lower con-tent of oxygen as a result of increased musculardemand in exercise (7). The arterial oxygensaturation is maintained; hence the A-V oxygendifference widens; and since the oxygen consump-tion continues to rise for three minutes (and heartrate does not), there is a progressive increase instroke volume. The striking increase in oxygentransport per heart beat during the second minuteof exercise is an expression of the changes in boththe A-V oxygen difference and the stroke volumesince it is the product of these two factors. As theapparent circulatory lag in adaptation to exerciseis corrected, the expired air R.Q. gradually risesduring the further attainment of the steady stateof exercise and the respiratory efficiency slowlydeclines; ventilation just perceptibly rises; andoxygen consumption diminishes by about 6%.These changes are considered to reflect peripheralcirculatory adjustments as vasodilatation occurs todissipate heat and slight arteriolization of venousblood from small peripheral shunts permits themixed venous oxygen content to rise (and narrowthe A-V oxygen difference) ( 10, 11 ). With thesechanges there is no significant change in diastolicblood pressure although the systolic pressure, andhence, the pulse pressure increase. The alterationsin the systolic time expressed by the electrocardio-graphic K constant are considered expressions ofchanges in blood flow as well as contractile forceand ventricular emptying associated with increasedstroke volume. Someindividuals exhibit slight re-ductions in both heart rate and ventilation aboutthe seventh minute of exercise, indicating furtheradjustments in performance; possibly these changesaccount for the phenomenon of the "second wind"

1429

1430 R. BRUCE, F. LOVEJOY, JR., R. PEARSON, P. YU, G. BROTHERS, AND T. VELASQUEZ

subjectively observed at more severe levels ofwork.

Recovery from exercise begins at once, and innormal subjects at this level of work it is largelycompleted within three minutes. The first vari-able to approach the resting level is the heart ratewhich corroborates the usefulness of the recoverypulse as a measure of physical fitness (12). Dur-ing this time the expired air R.Q. rises above 1.0as retained carbon dioxide continues to be ex-creted, delaying somewhat the restoration of ven-tilation to the resting level although favoring therepayment of most of the aerobic oxygen debt.

SUMMARY

1. The mean respiratory and circulatory path-ways of adaptation to the stress of exercise havebeen studied in 35 normal adults. The work-loadwas regulated by treadmill walking, and multipleobservations were made continuously at one min-ute intervals before, during, and after the exercise.

2. The previously reported changes in gas com-position were confirmed.

3. The additional values of simultaneous con-sideration of circulatory changes are discussed.

4. The adaptive changes in respiration and cir-culation are shown to complement and spare eachother by the differences in time and rate of change.

ACKNOWLEDGMENT

The authors wish to acknowledge the helpful criticismsof Doctors W. S. McCann and N. L. Kaltreider of theDepartment of Medicine of the University of Rochesterin the editing of this paper.

BIBLIOGRAPHY

1. Rahn, H., and Otis, A. B., Continuous analysis ofalveolar gas composition during work, hyperpnea,hypercapnia, and hypoxia. J. Appl. Physiol., 1949,1, 717.

2. Pelnar, P., A new method of examining the efficiencyof the respiratory and circulatory system. Nadi-ladem Ceshe Akademie Vid A UmEni, V Praze,1948.

3. Kaltreider, N. L., and McCann, W. S., Respiratoryresponse during exercise in pulmonary fibrosis andemphysema. J. Clin. Invest., 1937, 16, 23.

4. Bruce, R. A., Pearson, R., Lovejoy, F. W., Yu, P. N.G., and Brothers, G. B., Variability of respiratory andcirculatory performance during standardized exer-cise. J. Clin. Invest., 1949, 28, 1431.

5. Hurtado, A., Fray, W. W., Kaltreider, N. L., andBrooks, W. D. W., Studies of total pulmonarycapacity and its subdivisions. V. Normal values infemale subjects. J. Clin. Invest., 1934, 13, 169.

6. Fowler, W. S., Lung function studies. II. Respira-tory dead space. Am. J. Physiol., 1948, 154, 405.

7. Robinson, S., Experimental studies of physical fitnessin relation to age. Arbeitsphysiol., 1938, 10, 251.

8. Pearson, R., and Bruce, R. A., Unpublished studies,1949.

9. Baldwin, E. de F., Cournand, A., and Richards,D. W., Pulmonary insufficiency. I. Physiologicalclassification, clinical methods of analysis, stand-ard values in normal subjects. Medicine, 1948, 27,243.

10. Dill, D. B., The economy of muscular exercise.Physiol. Rev., 1936, 16, 263.

11. Grant, R. T., and Bland, E. F., Observations onarterio-venous anastomoses in the human skin andin the bird's foot with special reference to thereaction to cold. Heart, 1931, 15, 385.

12. Johnson, R. E., Brouha, L., and Darling, R. C., Testof physical fitness for strenuous exertion. Rev.Canad. de biol., 1942, 1, 491.