281

Regulation of Stroke Volume during Submaximal andMaximal Upright Exercise in Normal Man

Michael B. Higginbotham, Kenneth G. Morris, R. Sanders Williams, Philip A. McHale,R. Edward Coleman, and Frederick R. Cobb

From the Department of Medicine, Division of Cardiology, and the Department of Radiology, Duke University Medical Center, and theDurham Veterans Administration Medical Center, Durham, North Carolina

SUMMARY. To characterize the hemodynamic factors that regulate stroke volume during uprightexercise in normal man, 24 asymptomatic male volunteers were evaluated by simultaneous rightheart catheterization, radionuclide angiography, and expired gas analysis during staged uprightbicycle exercise to exhaustion. From rest to peak exercise, oxygen consumption increased from0.33 to 2.55 liters/min (7.7-fold), cardiac index increased from 3.0 to 9.7 liters/min per m2 (3.2-fold), and arteriovenous oxygen difference increased from 5.8 to 14.1 vol% (2.5-fold). The increasein cardiac index resulted from an increase in heart rate from 73 to 167 beats/min (2.5-fold), andan increase in left ventricular stroke volume index from 41 to 58 ml/m2 (1.4-fold). During lowlevels of exercise, there was a linear increase in cardiac index due to an increase in both heartrate and stroke volume index; stroke volume index increased as a result of an increase in leftventricular filling pressure and end-diastolic volume index and, to a much smaller extent, adecrease in end-systolic volume index. During high levels of exercise, further increases in cardiacindex resulted entirely from an increase in heart rate, since stroke volume index increased nofurther. Left ventricular end-diastolic volume index decreased despite a linear increase in pul-monary artery wedge pressure; stroke volume index was maintained by a further decrease in end-systolic volume index. The degree to which stroke volume index increased during exercise inindividuals correlated with the change in end-diastolic volume index (r = 0.66) but not with thechange in end-systolic volume index (r = 0.07). Thus, the mechanism by which left ventricularstroke volume increases during upright exercise in man is dependent upon the changing relation-ship between heart rate, left ventricular filling, and left ventricular contractility. At low levels ofexertion, an increase in left ventricular filling pressure and end-diastolic volume are importantdeterminants of the stroke volume response through the Starling mechanism. At high levels ofexertion, the exercise tachycardia is accompanied by a decrease in end-diastolic volume despite aprogressive increase in filling pressure, so that stroke volume must be maintained by a decreasein end-systolic volume. (Cire Res 58: 281-291, 1986)

MANY attempts have been made to characterize thecentral hemodynamic response to exercise in normalman (Bevegard et al., 1960; Reeves et al., 1961;Astrand et al., 1964; Julius et al., 1967; Ekelund andHolmgren, 1967; Epstein et al., 1967; Ekblom et al.,1968; Thadani and Parker, 1978; Hossack et al.,1980). Despite initial controversy concerning therelative contributions of stroke volume and heartrate to this response (Rushmer, 1959; Chapman etal., 1960), it has now been established that an in-crease in stroke volume is an important cardiovas-cular adjustment to upright exercise (Rowell, 1974;Clausen, 1976; and Astrand and Rodahl, 1977).However, the precise mechanism by which strokevolume increases during upright exercise in manremains to be established.

Measurements of pressure and left ventricular(LV) dimensions during exercise have been used byWildenthal and Mitchell (1969), Vatner et al. (1972),and Horwitz et al. (1972) to clarify the importanceof heart rate, LV contractility, and LV filling (Frank-Starling mechanism) in the regulation of stroke vol-

ume in dogs. However, technical limitations haverestricted similar studies in humans to small num-bers of subjects with cardiac disease, and to lowlevels of submaximal exercise in the supine position(Harrison et al., 1963; Sonnenblick et al., 1965;Braunwald et al., 1967); no study has combinedmeasurements of pulmonary and systemic pressureand LV volume in normal subjects during uprightexercise.

In order to examine the physiological factors thatcontribute to the increase in stroke volume duringupright exercise in normal man, we measured pul-monary and systemic hemodynamics and LV vol-umes simultaneously by right heart catheterizationand radionuclide angiography during submaximaland maximal upright bicycle exercise in 24 healthymale volunteers.

Methods

SubjectsThe study population consisted of 24 asymptomatic

males aged 20-50 years; seven subjects were in their 20's,

by guest on June 6, 2018http://circres.ahajournals.org/

Dow

nloaded from

282

10 were in their 30's, and seven were 40-50 years old. Allhad sedentary occupations; nine subjects performed reg-ular physical activity off the job in the form of jogging,and only three subjects jogged for at least 30 minutesevery day. Subjects varied from 63-103 kg in weight andfrom 173-185 cm in height; body surface area rangedfrom 1.65-2.10 m2. Hemoglobin ranged from 13.2-16.7g/100 ml. No subject demonstrated any abnormality onthe electrocardiogram or on physical examination.

All studies were performed under a research protocolapproved by the Institutional Review Boards of both theDuke University and the Durham Veterans AdministrationMedical Centers.

Study Protocol

Participants initially reported to the cardiac catheteri-zation laboratory at 8:00 AM in the postabsorptive state.Under local anesthesia, a 7-French balloon-tipped ther-modilution Swan-Ganz catheter was introduced into theright pulmonary artery under fluoroscopic control, and an18-gauge 2'/2"plastic cannula was introduced percuta-neously into the left brachial artery. Subjects then weretransferred to the exercise facility, where simultaneoushemodynamic, radionuclide, and gas exchange measure-ments were recorded at rest in the supine and sirringpositions, and then at each work load achieved duringupright bicycle exercise to exhaustion.

Exercise was performed on an isokinetic bicycle ergom-eter (Fitron, Lumex Inc.). Work load was commenced at150 kilopond meters (kpm) per minute (25 watt) and wasincreased by 150 kpm per minute every 3 minutes untillimited by exhaustion. For analysis, maximum exercisewas taken to be the highest completed 3-minute exercisestage. During each stage of exercise, measurements werestrictly timed so that arterial and mixed venous blood weresampled simultaneously with expired gas analysis, duringthe 3rd minute of each exercise stage. Equilibrium radio-nuclide angiograms were acquired during the 2nd and 3rdminute of each exercise stage. Pulmonary artery wedgepressure and thermodilution cardiac output measurementswere recorded during the 2nd minute of each exercisestage. Pulmonary artery wedge pressure measurementswere repeated at the end of the 3rd minute during at leastone of the high work loads in each subject; none of themeasurements made at 3 minutes exceeded those made at2 minutes by more than 2 mm Hg.

Hemodynamic Measurements

All supine pressure measurements were related to areference point at the mid axillary line, i.e., a level closelyapproximating the position of the right atrium. To ensurethat an appropriate reference point was used for uprightmeasurements, a small radioactive source was positionedon the chest wall opposite the center of the right atriumas indicated on the scintigraphic display on the gammacamera; pressure transducers were adjusted to this level.The position was often, but not always, opposite the 4thright intercostal space.

Pulmonary and systemic pressures (systolic, diastolic,and electronically derived mean) were obtained withHewlett-Packard pressure transducers and amplifiers, andwere recorded continuously at a paper speed of 2.5 mm/sec. Values of mean pulmonary artery wedge pressurewere taken at the average point in the respiratory cycle toaverage the effects of forced expiration and inspirationduring exercise.

Circulation Research/Vo/. 58, No. 2, February 1986

Blood samples taken during each exercise stage werechilled in an ice bath immediately after collection. Thepercent O2 saturation of 3-ml arterial and mixed venousblood samples was measured on an Instruments Labora-tories oximeter, calibrated with known dilutions of O2-saturated blood. The O2 content in the mixed venous andarterial blood samples was obtained for calculations ofFick cardiac output. Stroke volume was obtained by divid-ing Fick cardiac output by heart rate. Systemic and pul-monary vascular resistances (in dyn • sec • cm~!) were cal-culated by multiplying mean blood pressure (in mm Hg)by 80, and dividing by cardiac output. Cardiac output alsowas measured by thermodilution on a commercially avail-able dedicated computer (Instruments Laboratories) at restand during each exercise stage. Arterial lactate concentra-tion was estimated by the standard technique of lactateoxidation and NAD+ reduction (Calbiochem-Behringrapid lactate kit).

Expired Gas AnalysisExpired gases were analyzed continuously both at rest

and during exercise. The O2 and CO2 contents of theexpired air were recorded continuously with BeckmanOM-14 and LB-2 analyzers, respectively. Minute volumewas recorded continuously with a Pneumoscan spirom-eter. Oxygen consumption (Voj) was calculated at rest andduring the last minute of each exercise stage from standardformulas after correction for temperature and barometricpressure. Peak V02 was used as an objective index ofaerobic work performance or cardiovascular reserve.

To determine whether measurements of V02 during the3rd minute of each exercise stage adequately representedsteady state values, a separate study was performed inwhich five normal subjects performed graded maximumexercise using 6-minute exercise stages. Early (2-3 minute)and late (5-6 minute) measurements of V02 differed byonly 3 ± 2%; the largest difference was 10% in one subjectat peak exercise. The catheterization procedure did notappear to affect individual measurements of peak Vo?.Ten subjects underwent separate exercise tests withoutcatheter placement within several weeks of the invasivestudy. In these subjects, whose peak Vch ranged from 23.0to 50.0 ml/kg per min, no systematic difference was seenbetween the two studies, with mean V02 values of 33.7 ±7.3 and 34.3 ± 7.3 ml/kg per min. Peak Voj differed byonly 1.6 ± 1.8 ml/kg per min.

Radionuclide AngiographyAfter in vivo labeling of red blood cells with 30 mCi

technetium-99m, gated equilibrium radionuclide studieswere acquired, using a Searle LEM mobile y camera witha high-sensitivity 30° slant hole collimator interfaced withan A2 computer (Medical Data Systems). Gating was trig-gered by the R-wave of the electrocardiogram. All studieswere acquired in the left anterior oblique projection thatallowed optimum separation of left and right ventricles(approximately 40°); data acquisition time ranged from1.5—2 minutes for each study. During the exercise studies,particular care was taken to minimize movement of thesubject, to avoid firm gripping of the camera, and tomaintain a constant work load throughout each radio-nuclide acquisition.

Using standard computer algorithms, we defined LVborders by a semiautomated edge-detection method; back-ground was selected automatically by reference to theend-systolic frame, and ejection fraction (EF) was com-puted from the end-diastolic (ED) and end-systolic (ES)

by guest on June 6, 2018http://circres.ahajournals.org/

Dow

nloaded from

Higginbotham et al. /Regulation of Stroke Volume during Exercise 283

counts, thus: EF = (ED counts — background) — (EScounts — background) + (ED counts — background).

To minimize the influence of background and tissueattenuation on measurements of LV volume in this study,LV end-diastolic and end-systolic volumes were derivedentirely from radionuclide ejection fraction and Fick strokevolume. Ejection fraction is affected little by the technicalfactors mentioned above, and is widely accepted as beingthe most accurate and reproducible measurement in quan-titative radionuclide angiography (Gould, 1982). The de-cision to measure LV volumes in this manner was basedon the fact that reported methods for radionuclide volumeestimation have shown a standard error of approximately30 ml when compared with contrast angiography (Dehmeret al., 1980; Links et al., 1982). This error constitutesapproximately 25% of the normal end-diastolic volumeand 50% of the normal end-systolic volume; in contrast,standard errors for estimations of ejection fraction havebeen of the order of 0.05, or 10% of the measurement(Dehmer et al., 1980). Although independent validationof measurements of LV volume during exercise is notpossible, it seems highly likely that a technique employingFick stroke volume and radionuclide ejection fractionwould involve smaller errors than measurement of vol-umes by radionuclide techniques alone.

Measurement of ejection fraction by radionuclide an-giography has been compared with contrast angiographicejection fraction measurements in 27 males in our labo-ratory; linear regression analysis between radionuclideand contrast values demonstrated a correlation coefficientof 0.93 and a standard error of 0.047. Ejection fractionhas been shown also to be a highly reproducible measure-ment both at rest and during exercise. In previous studiesin our laboratory (Morris et al., 1983), ejection fractionmeasured on two occasions varied by approximately 0.03both at rest and during exercise.

Left ventricular ejection time and peak ejection ratewere derived from the volume curve of the radionuclidestudy using a method identical to that described by Ma-gorien et al. (1983).

AnalysisGroup data for each variable are expressed as mean ±

SD. The significance of changes for the group betweensupine and upright rest, and between upright rest andpeak exercise, was assessed by two-tailed paired Mests.

The relationship between progressive changes in var-ious cardiovascular parameters and changes in Vch wereexpressed as both individual and group responses. Foranalysis of the group response, a polynomial regressionanalysis was performed on the data for each patient,beginning with the first power of Vo2. An F-test was usedto determine whether adding the next highest power ofVo? resulted in a significant reduction in the regressionmean square. The regression equation that fit all the datapoints was determined by averaging the individual regres-sion coefficients across all patients. Figures were con-structed by plotting all data points; the regression line wasconstructed from the average regression coefficients.

ResultsNo serious complications resulted from these

studies. Three cases of superficial thrombophlebitisresulted from the insertion of the Swan-Ganz cath-eter, but each case resolved with conservative ther-apy.

Table 1 illustrates mean values (±SD) and rangesfor hemodynamic variables measured at rest in thesupine and sitting positions, and during maximalupright bicycle exercise.

Hemodynamics at Rest; Effects of PostureIn the supine position, before sitting on the bicy-

cle, all subjects were allowed to relax, and measure-ments were made when the heart rate was stable.Heart rate averaged 64 beats/min and exceeded 80beats/min in only three of the 24 subjects. All sub-jects were normotensive, and values for all hemo-dynamic variables were within the accepted normalrange (Table 1).

When the subjects assumed a sitting position onthe bicycle, heart rate increased, diastolic blood pres-sure increased, pulmonary artery pressures de-creased, and systemic and pulmonary vascular re-sistances increased. Left ventricular end-diastolic,end-systolic, and stroke volume indexes decreased,so that cardiac index decreased despite the smallincrease in heart rate. Arteriovenous O2 differenceincreased to compensate for the decrease in cardiacindex. Despite the decrease in LV filling as reflectedby lower pulmonary pressures and LV volumes, nosystematic differences were seen in ejection fractionsbetween the supine (0.59) and upright (0.61) posi-tions. Peak ejection rate and LV ejection time bothdecreased significantly from the supine to the up-right position, consistent with reflex-mediated in-creases in contractility and heart rate.

Exercise Performance and End-PointsGroup values for peak work load and V02 are

shown in Table 1. These data confirm the widerange of exercise performance to be expected froma heterogeneous population of normal subjects. Fig-ure 1 confirms a linear increase in V02 for eachsubject with increasing work load; the slightly moreabrupt increase from rest to the initial stage ofexercise resulted from additional work to overcomethe inertia of the legs. As in other studies of un-trained individuals (Mitchell et al., 1958), 'true'maximum V02 could not be confirmed in this study,since V02 did not plateau at peak work loads. How-ever, arterial lactate levels were consistent with theachievement of maximum or near-maximum levelsof exertion: serum lactate exceeded 65 mg/dl in eachsubject, and was more than 90 mg/dl in 14 of the24 subjects (Astrand, 1960).

Hemodynamic Responses to Maximal UprightBicycle Exercise

The magnitude of changes in hemodynamic pa-rameters from rest to peak exercise are shown inTable 1.

Maximal exercise hemodynamic data varied con-siderably between individual subjects. Peak heartrate varied from 136-190 beats/min; this variabilitywas due primarily to age differences in the popula-

by guest on June 6, 2018http://circres.ahajournals.org/

Dow

nloaded from

284 Circulation Research/Vo/. 58, No. 2, February 1986

TABLE 1

Hemodynamic Data at Rest and during Peak Exercise (Mean ± SD, and Range)

W (kpm/m)

Voj (liters/min)

HR (beats/min)

SAPs (mm Hg)

SAPd (mm Hg)

TSVR (dyne-sec-cnrr5)

PAPs (mm Hg)

PAPd (mm Hg)

TPVR(dyneseccm- ' )

PAWPm (mm Hg)

CI (liters/min per m2)

SVI (ml/m2)

LVEDVI (ml/m2)

LVES1 (ml/m2)

LVEF

PER (EDV/sec)

LVET(sec)

Acb (vol%)

CV05 (vol%)

A-Vo, (vol%)

Supine

0.31 ± 0.06(0.20-0.45)

64 ± 12(48-90)

130 ± 14(110-160)71 ± 6(60-85)

1148 ±290(648-1900)20 ± 4(15-31)9 ± 2(6-14)

65 ±31(13-140)9 ± 3(4-12)

3.6 ± 0.8(2.5-5.4)57 ± 14(38-77)97 ± 25(68-147)41 ± 6(27-54)

0.59 ± 0.06(0.50-0.75)

-2.85 ± 0.49(-3.49H-2.10)

0.36 ± 0.04(0.30-0.47)20.6 ± 1.4(18.4-23.3)16.2 ± 1.3(14.1-19.1)4.4 ±0.7(3.8-5.0)

Rest

Upright

0.33 ± 0.06(0.25-0.49)

73 ± 12(51-92)

136 ± 12(115-160)74 ± 6(70-95)

1446 ± 260(976-1837)15 ± 4(7-22)7 ± 3(2-11)

103 ± 43(27-191)3 ± 2(0-8)

3.0 ± 0.6(1.9-4.1)41 ± 9(24-58)68 ±18(32-96)28 ± 5

(8-50)0.61 ± 0.09(0.45-0.78)

-3.49 ± 0.88(-6.20H-2.61)

0.32 ± 0.04(0.26-0.39)20.8 ± 1.4(18.6-23.5)15.0 ±1.3(12.6-16.9)5.8 ± 0.9(3.5-7.2)

P(Sup vs. Upr)

0.31±0

0.007

0.153

0.001

<0.001

<0.001

0.001

<0.001

<0.001

0.005

<0.001

O.001

<0.001

0.641

<0.001

<0.001

0.702

<0.001

0.001

Peak exercise

963 ± 146(750-1200)

2.55 ± 0.31(2.02-3.80)167 ± 16(136-190)220 ± 24(160-270)94 ± 8(85-115)

585 ± 101(384-758)34 ± 6(26-45)17 ± 6(7-35)

67 ±22(36-112)10 ± 3(5-15)

9.7 ± 1.5(7.5-12.4)58 ± 8(47-76)77 ±13(46-106)19 ±10(4-44)

0.76 ±0.10(0.53-0.90)

-6.76 ± 1.25(-11.10H-5.22)

0.19 ±0.02(0.13-0.23)21.2 ± 1.7(18.7-23.4)7.1 ±0.8(4.5-9.4)

14.1 ± 1.8(10.4-17.5)

P

(Upr Restvs. Ex)

<0.001

<0.001

<0.001

<0.001

<0.001

<0.001

<0.001

O.001

O.001

O.001

O.001

O.001

0.011

O.001

O.001

O.001

O.001

0.824

O.001

O.001

Abbreviations: W = work load; V02 = oxygen consumption; HR = heart rate; SAPs, SAPd and SAPm = systemic blood pressure(systolic, diastolic, and mean); TSVR = total systemic vascular resistance; PAPs, PAPd and PAWPm m pulmonary arterial pressure(systolic, diastolic and mean); TPVR = total pulmonary vascular resistance; CI = cardiac index, SVI = stroke volume index; LVEF = leftventricular ejection fraction; PER = peak ejection rate; LVET = left ventricular ejection time, LVEDVI = left ventricular end-diastolicvolume index; LVESVI = left ventricular end-systolic volume index; A02 = systemic arterial oxygen content; CVOj = pulmonary arterialoxygen content; A-V02 = arteriovenous oxygen difference.

tion and was independent of the degree of exerciseeffort as judged from the serum lactate levels (r =0.27). Oxygen consumption increased from 0.33 to2.55 liters/min (7.7-fold), cardiac index from 3.0 to9.7 liters/min per m2 (3.2-fold), and arteriovenousO2 difference from 5.8 to 14.1 vol% (2.5-fold). Theincrease in cardiac index resulted from an increasein heart rate from 73 to 167 beats/min (2.5-fold)and of stroke volume index from 41 to 58 ml/m(1.4-fold). It is noteworthy that values for strokevolume index, end-diastolic volume index, andmean pulmonary artery wedge pressure measured

in the sitting position at peak exercise did not exceedvalues recorded at rest in the supine position. Al-though, on the average, stroke volume index in-creased by 17 ml/m2 (40%) from rest to peak exer-cise, individual responses varied, ranging from 1ml/m2 (2%) to 29 ml/m2 (70%). The change in strokevolume index from rest to peak exercise correlatedwith the change in end-systolic volume index (r =0.66, P = 0.001) but not with the change in end-systolic volume index (r = 0.07, P = 0.735). Thus,individual differences in the stroke volume responseto exercise were explained largely by differences in

by guest on June 6, 2018http://circres.ahajournals.org/

Dow

nloaded from

Higginbotham et al./Regulation of Stroke Volume during Exercise 285

4000

g 3000

2000

ouN

o1000

300 600 900 1200

Work Rate (kpm/min)

FIGURE 1. Oxygen consumption during staged maximal upright bi-cycle exercise in 24 normal males, 20-50 years old. Data for individualsubjects are plotted as a function of work rate.

the extent to which end-diastolic volume increased,rather than the extent to which end-systolic volumedecreased.

Changes in vascular resistance during upright ex-ercise differed considerably for the systemic andpulmonary circulations. Total systemic vascular re-sistance decreased dramatically from 1446 to 585dyne X sec X cm"5; the value at peak exercise wasmuch lower than the supine value at rest. In con-trast, there was a small decline in calculated pul-monary vascular resistance from 103 to 67 dyne Xsec X cm"5, with the exercise value equaling the restsupine value.

Pattern of Hemodynamic Responses toProgressive Upright Exercise

To examine cardiovascular mechanisms duringprogressive exercise independent of individual dif-ferences in maximal aerobic capacity, each variablewas considered as a function of V02 normalized toits peak value.

Heart Rate

Heart rate increased progressively with increasinglevels of work in each subject and increased linearlyin relation to V02.

I?<f> E

0-<

a>as?; E

300

175

50300

175

50300

175

en50

_ 2000m"Eo

1000a>

~ 00 25 50 75 100

Relative 02 Consumption (% of max)FIGURE 2. Systemic arterial pressure (SAP) and vascular resistance(SVR) during staged maximal upright bicycle exercise in 24 normalmales, 20-50 years old. Each variable is plotted against normalizedO i consumption and is represented as a computer-generated regres-sion curve with 95% confidence limits.

50.0

37.5

• E 25.0

W 12.5

037.5

3 25.0

£ 12.5

037.5

o— 25.0

3 12.5

0_ 375

o 250

Q. J 125c

25 50 75 100

Relative 08 Consumption (%of max)

FIGURE 3. Pulmonary artery pressure (PAP) and vascular resistance(PVR) during staged maximal upright bicycle exercise in 24 normalmales, 20-50 years old. Data are presented as in Figure 2.

by guest on June 6, 2018http://circres.ahajournals.org/

Dow

nloaded from

286 Circulation Research/Vo/. 58, No. 2, February 1986

37.5 r

\ 27.8

I 18.2

8.6

-1.025 50 75 100 0 25 50

Relative 02 Consumption (% of max)

75 100

FIGURE 4. Mean pulmonary artery wedge pres-sure (PAWP) during staged maximal uprightbicycle exercise in 24 normal males, 20-50 yearsold. The left panel illustrates individual data.The group response (right panel) is representedas in Figures 2 and 3.

Systemic Arterial Pressure

Systolic arterial pressure increased linearly withincreasing V02 (Fig. 2). In contrast, diastolic bloodpressure did not increase until subjects had exercisedto approximately 50% of peak Vo2, resulting in asecond order polynomial curve fit. Mean arterialpressure increased linearly due to the predominanteffect of changes in systolic pressure. Systemic vas-cular resistance declined exponentially.

Pulmonary Arterial Pressure

As shown in Figure 3, the pattern of changes inpulmonary arterial pressure during progressive up-right exercise was different from that seen in thesystemic circulation. Systolic pulmonary arterialpressure increased more rapidly during initial stagesof exercise than during the later stages; this responsewas best described by a curvilinear function. Dia-stolic pulmonary artery pressure increased in aslightly curvilinear fashion. A small curvilinear de-cline in pulmonary vascular resistance was seenduring progressive exercise.

Pulmonary Artery Wedge Pressure

Pulmonary artery wedge pressure increased line-arly during progressive upright exercise (Fig. 4).Curves for pulmonary artery diastolic pressure andmean pulmonary artery wedge pressure had similarslopes, but the wedge pressure was approximately3 mm Hg less than the diastolic pressure at lowlevels of exercise and 5 mm Hg less at higher levels.

Cardiac Output

Figure 5 illustrates a linear relationship betweencardiac index and V02 during progressive uprightexercise. The linear relationship between these twoparameters is comparable to previous studies (Row-ell, 1974; Clausen, 1976; Astrand and Rodahl, 1977)and provides validation of the Fick cardiac outputmeasurements. To validate Fick measurements ofcardiac output further, cardiac output was measuredalso by the thermodilution technique immediatelybefore the Fick measurements, during each stage ofexercise. Measurements of cardiac index by the twoindependent techniques were highly correlated (r =0.95) with a standard error of 0.96 liter/min per m2.Values were distributed evenly about the line ofidentity except for outputs greater than 9 liters/minper m2, where thermodilution values exceeded Fick;overestimation of cardiac output at high flow ratesis a common observation with indicator dilutiontechniques (Reeves et al., 1961; Hanson and Ta-bakin, 1964).

Left Ventricular Volumes

Progressive changes in LV stroke volume index,end-diastolic volume index, and end-systolic vol-ume index are illustrated in Figure 6, with eachparameter shown on the same scale.

Stroke volume index increased initially, reached aplateau at approximately 50% of peak VO2, andtended to decrease slightly at peak exercise.

End-diastolic volume index increased uniformly

15.0

"e 11.2

E1 7.5

0 3.7

FIGURE 5. Cardiac index (CI) during stagedmaximal upright bicycle exercise in 24 normalmales, 20-50 years old. Individual and groupdata are represented as in Figure 4.

1000 2000 3000 4000 1000 2000 3000 4000

O2 Consumption (ml/min)

by guest on June 6, 2018http://circres.ahajournals.org/

Dow

nloaded from

Higginbotham et al./Regulation of Stroke Volume during Exercise

160

287

FIGURE 6. Stroke volume index (SVI), end-dia-stolk volume index (EDVI), and end-systolicvolume index (ESVI) during staged maximalupright bicycle exercise in 24 normal males,20-50 years old. Data are represented as INFigures 4 and 5.

0 25 50 75 100 25 50 75

Relative 02 Consumption (% of max)

100

during the initial exercise levels, reached its peakvalue at approximately 40% of maximum Vo2, andthen remained unchanged or decreased as peakexercise was approached. End-diastolic volume in-dex tended to decrease in 17 of the 24 subjectsduring the last exercise stage, compared with a slightincrease in the remaining seven subjects.

End-systolic volume index decreased linearly,though by a small extent, during progressive uprightexercise.

Systolic Left Ventricular Performance

Systolic LV performance, as measured by the ejec-tion phase indices, ejection fraction, peak ejectionrate, and LV ejection time, increased progressivelyduring upright bicycle exercise (Fig. 7). Ejection frac-tion increased more rapidly during the early stagesof exercise, while changes in peak ejection rate andejection time were linear throughout exercise.

Effect of AgeTo determine whether the variability in age

among subjects could have influenced the cardio-vascular mechanisms described in this study, mul-tiple linear regression analyses were performed to

determine the relationship between age andcardiovascular variables at rest and during peakexercise. No significant age relationship was seenfor any hemodynamic measurement at rest. Signifi-cant relationships (r = 0.41) were seen between ageand some variables at peak exercise: V02 ml/m2 (r =—0.43), heart rate (r = —0.63), systemic vascularresistance (r = 0.59), pulmonary vascular resistance(r = 0.42) cardiac index (r = —0.54), and LV ejectiontime (r = 0.42). However, no correlation was seenbetween age and stroke volume index, end-diastolicvolume index, end-systolic volume index, ejectionfraction, peak ejection rate, or pulmonary arterywedge pressure during exercise. Thus, whereas, inour study, age did affect peak heart rate, cardiacindex, and O2 consumption, it had no significanteffect on the LV pressure, volume, and contractilityresponses to exercise, and as such was unlikely tohave affected the major findings of this study.

It may be of interest to note that the magnitudeof decline in end-diastolic volume, from its highestvalue during submaximal exercise to that at peakexercise, was related inversely to age (r = —0.47; P— 0.03). Since this measure of the apparent decreasein cardiac compliance during peak exercise wasgreater in younger subjects, our findings do not

by guest on June 6, 2018http://circres.ahajournals.org/

Dow

nloaded from

by guest on June 6, 2018http://circres.ahajournals.org/

Dow

nloaded from

by guest on June 6, 2018http://circres.ahajournals.org/

Dow

nloaded from

290

tricular end-diastolic diameter and stroke volume,in contrast to the uncontrolled exercise state.

The quantitative changes in central and peripheralcardiovascular variables seen in the present studywere consistent with other studies of normal sub-jects, as summarized in well-referenced reviews(Rowell, 1974; Clausen, 1976; Astrand and Rodahl,1977). Exercise performance as assessed by peak O2consumption was somewhat less in our study (2.55±3.1 liters/min or 30.4 ± 6.2 ml/kg per min) thanhas been described in many studies of healthy malevolunteers (Mitchell et al., 1958; Saltin and Astrand,1967); however, many of these studies have in-volved well-trained individuals who would be ex-pected to have a higher aerobic capacity than ouruntrained population. In recent studies of sedentaryindividuals, Martin et al. (1983) and DeBusk et al.(1983) reported peak levels of O2 consumption sim-ilar to those seen in our subjects. A near-maximumlevel of exertion was confirmed in our study by thehigh plasma lactate levels obtained.

Our findings confirm previous observations of alinear increase in heart rate, an increase in bloodpressure, and a decrease in systemic vascular resist-ance during exercise (Bevegard et al., 1960; Astrandet al., 1964; Robinson et al., 1966; Epstein et al.,1967; Julius et al., 1967; Ekblom et al., 1968; Thadaniand Parker, 1978). An increase in stroke volumeduring the first half of exercise, followed by a pla-teau, is also consistent with these studies. The smalldecline in stroke volume at peak exercise in ourstudy has not been a consistent finding, but hasbeen noted by other investigators (Keul et al., 1981).Maximum values for stroke volume index and car-diac index in our study were very similar to valuesobtained by Hossack and Bruce (1980), who ob-served that indirect estimates of cardiac index in-creased from 2.6 to 9.6 liters/min per m2 (comparedwith our values of 3.0 and 9.7 liters/min per m2)and stroke volume index from 37 to 56 ml/m2 (com-pared with our values of 41 and 58 ml/m2 duringtreadmill exercise to a maximum O2 consumption of38 ml/kg per min. Our findings of a small increasein end-diastolic volume index during upright bicycleexercise is consistent with previous noninvasivestudies (Rerych et al., 1978; Crawford et al., 1979;Sorensen et al., 1980); Higginbotham et al., 1983).Maximum values of arteriovenous O difference inthe present study 14.1 ±1.8 vol% are slightly lowerthan previously reported (Astrand et al., 1964), per-haps reflecting differences in physical conditioningor technical differences between previously derivedvalues and the direct measurements used in thepresent study.

Few previous studies have measured pulmonaryartery and pulmonary artery wedge pressures duringupright exercise in humans (Bevegard et al., 1960;Granath et al., 1964; Julius et al., 1967; Messin etal., 1970; Thadani and Parker, 1978), and evenfewer of these have examined asymptomatic normalvolunteers (Granath et al., 1964; Messin et al., 1970).

Circulation Research/Vo/. 58, No. 2, February 1986

The findings in our study are consistent with theseprevious reports.

In summary, the pressure and volume responsesof the left ventricle to progressive upright exercisein humans differ considerably from those previouslymeasured in exercising dogs and in man exercisingin the supine position. During low levels of uprightexercise, stroke volume increases primarily as a con-sequence of an increase in left ventricular fillingpressure and end-diastolic volume (Starling mecha-nism) with a small contribution from a decreasedend-systolic volume. During high levels of exercise,left ventricular end-diastolic volume does not in-crease, despite a further increase in filling pressure;on the contrary, end-diastolic volume decreases,possibly as a result of reduced diastolic filling athigh heart rates. Stroke volume at high levels ofexercise is maintained through a progressive de-crease in end-systolic volume.

We gratefully acknowledge James Stanfield, NMT, Debbie Repass,NMT, Jean Wilson, RN, and Jeanne Watkinson, RN, for technicalassistance, the Medical Media Production Service of the DurhamVeterans Administration Medical Center for the illustrations, andCathie Collins for preparation of the manuscript.

This study was supported by Grant HL-17670 from the NationalHeart, Lung, and Blood Institute, by the Geriatric Research, Educa-tion, and Clinical Center (GRECQ, and by General Medical ResearchFunds from the Veterans Administration.

Dr. Higginbotham was supported by a grant from the AmericanHeart Association, North Carolina Affiliate.

Address for reprints: Frederick R. Cobb, M.D., Division of Car-diology C111A), Durham VA Medical Center, 508 Fulton Street,Durham, North Carolina 27705.

Received February 25, 2985; accepted for publication August 29,1985.

References

Astrand I (1960) Aerobic work capacity in men and women withspecial reference to age. Acta Physiol Scand 49: (suppl 169): 1-92

Astrand P-O, Rodahl K (1977) Textbook of Work Physiology.New York, McGraw-Hill

Astrand P-O, Cuddy TE, Saltin B, Stenberg J (1964) Cardiacoutput during submaximal and maximal work. J Appl Physiol19: 268-274

Bevegard S, Holmgren A, Jonsson B (1960) The effect of bodyposition on the circulation at rest and during exercise, withspecial reference to the influence on the stroke volume. ActaPhysiol Scand 49: 279-298

Braunwald E, Sonnenblick EH, Ross J Jr, Glick G, Epstein SE(1967) An analysis of the cardiac response to exercise. Circ Res20/21 (suppl I): 44-58

Brutsaert DL, Sonnenblick EH (1973) Cardiac muscle mechanicsin the evaluation of myocardial contractility and pump func-tion: Problems, concepts and directions. Prog Cardiovasc Dis16: 337-361

Chapman CB, Fisher JN, Sproule BJ (1960) Behavior of strokevolume at rest and during exercise in human beings. ] ClinInvest 30: 1208-1213

Clausen JP (1976) Circulatory adjustments to dynamic exerciseand effects of physical training in normal subjects and patientswith coronary artery disease. Prog Cardiovasc Dis 18: 459-495

Crawford MH, White DH, Anon KW (1979) Echocardiographicevaluation of left ventricular size during handgrip and supineand upright bicycle exercise. Circulation 59: 1188-1196

DeBusk RF, Convertino VA, Hung ], Goldwater D (1983) Exercise

by guest on June 6, 2018http://circres.ahajournals.org/

Dow

nloaded from

Higginbotham et al./Regulation of Stroke Volume during Exercise 291

conditioning in middle-aged men after 10 days of bed rest.Circulation 68: 245-250

Dehmer GJ, Lewis SE, Hillis LD, Tweig D, Falkhoff M, ParkeyRW, Willerson JT (1980) Nongeometric determination of leftventricular volumes from equilibrium blood pool scans. Am JCardiol 45: 293-300

Ekblom B, Astrand P-O, Saltin B, Stenberg J, Wallstrom B (1968)Effect of training on circulatory response to exercise. J ApplPhysiol 24: 518-528

Ekelund LC, Holmgren A (1967) Central hemodynamics duringexercise. Circulation 20-21 (suppl I): 33-43

Epstein SE, Beiser CD, Stampfer M, Robinson BF, Braunwald E(1967) Characterization of the circulatory response to maximalupright exercise in normal subjects and patients with heartdisease. Circulation 35: 1049-1062

Gould KL (1982) Quantitative imaging in nuclear cardiology.Circulation 66: 1141-1146

Granath A, Jonsson B, Strandell T (1964) Circulation in healthyold men, studied by right heart catheterization at rest andduring exercise in supine and sitting position. Acta Med Scand176: 425-446

Hanson JS, Tabakin BS (1964) Simultaneous and rapidly repeatedcardiac output determinations by dye-dilution method. J ApplPhysiol 19: 275-278

Harris P, Heath D (1977) Normal variations in pressure and flow.In The Human Pulmonary Circulation. Its Form and Functionin Health and Disease. Edinburgh, Churchill-Livingstone, pp157-164

Harrison DC, Goldblatt A, Braunwald E, Glick G, Mason DT(1963) Studies on cardiac dimensions in intact, unanesthetizedman. III. Effects of muscular exercise. Circ Res 13: 460-467

Higginbotham MB, Morris KG, Conn EH, Coleman RE, Cobb FR(1983) Determinants of variable exercise performance amongpatients with left ventricular dysfunction. Am J Cardiol 51: 52-60

Horwitz LD, Atkins JM, Leshin SJ (1972) Role of the Frank-Starling mechanism in exercise. Circ Res 31: 868—875

Hossack KF, Bruce RA, Green B, Kusumi F, DeRouen TA, TrimbleS (1980) Maximal cardiac output during upright exercise: Ap-proximate normal standards and variations with coronary heartdisease. Am J Cardiol 46: 204-212

Julius S, Amery A, Whitlock LS, Conway J (1967) Influence ofage on the hemodynamic response to exercise. Circulation 36:222-230

Keul J, Dickhuth H, Simon G, Lehmann M (1981) Effect ofdynamic exercise on heart volume, contractility, and left ven-tricular dimensions. Circ Res 48 (suppl I): 162-170

Kjellberg SR, Rudhe V, Sjostrand T (1949) The amount of hemo-globin (blood volume) in relation to the pulse rate and heartvolume during work. Acta Physiol Scand 19: 152-158

Links JM, Becker LC, Schindledecker JG, Guzman P, Burow PD,Nickoloff EL, Alderson PO, Wagner HN (1982) Measurementof absolute left ventricular volume from gated blood poolstudies. Circulation 65: 82-91

Magorien DJ, Shaffer P, Bush CA, Magorien RD, Kolibash AJ,Leier CV, Bashore TM (1983) Assessment of left ventricularpressure-volume relations using gated radionuclide angiogra-phy, echocardiography, and micromanometer pressure record-ings. A new method for serial measurements of systolic and

diastolic function in man. Circulation 67: 844-853Martin WH, Berman Wl, Buckey JC, Snell PG, Blomqvist CG

(1983) Cardiovascular responses to large and small muscledynamic exercise in patients with ischemic heart disease andin normal subjects (abstr). Circulation 68: (suppl III): 149

Messin R, Degre S, Demaret B, Vandermoten P, Denolin H (1970)The effect of age on pulmonary circulation in normal subjects.Prog Resp Res 5: 385-394

Mitchell JH, Sproule BJ, Chapman CB (1958) The physiologicalmeaning of the maximal oxygen intake test. J Clin Invest 37:538-547

Morris KG, Higginbotham MB, Coleman RE, Shand DC, CobbFR (1983) Comparison of high-dose and medium-dose pro-pranolol in the relief of exercise-induced myocardial ischemia.AmJ Cardiol 52: 8-13

Nixon JV, Murray RG, Leonard PD, Mitchell JH, Blomqvist CG(1982) Effect of large variations in preload on left ventricularperformance characteristics in normal subjects. Circulation 65:698-703

Reeves JT, Grover RF, Blount SG, Filley GF (1961) Cardiac outputresponse to standing and treadmill walking. J Appl Physiol 16:283-288

Rerych SK, Scholz PM, Newman GE, Sabiston DC Jr, Jones RH(1978) Cardiac function at rest and during exercise in normalsand in patients with coronary heart disease: Evaluation byradionuclide angiocardiography. Ann Surg 187: 449-464

Robinson BF, Epstein SE, Kahler RL, Braunwald E (1966) Circu-latory effects of acute expansion of blood volume: Studiesduring maximal exercise and at rest. Circ Res 19: 26-32

Rowell LB (1974) Human cardiovascular adjustments to exerciseand thermal stress. Physiol Rev 54: 75-159

Rushmer RF (1959) Constancy of stroke volume in ventricularresponses to exertion. Am J Physiol 1%: 745-750

Saltin B, Astrand P-O (1967) Maximal oxygen uptake in athletes.J Appl Physiol 23: 353-358

Sarnoff SJ, Mitchell JH (1961) The regulation of the performanceof the heart. Am J Med 30: 747-771

Sonnenblick EH, Braunwald E, Williams JF, Glick G (1965) Effectsof exercise on myocardial force-velocity relations in intact un-anesthetized man: Relative roles of changes in heart rate,sympathetic activity, and ventricular dimensions. J Clin Invest44: 2051-2062

Sorenson SG, Ritchie JL, Caldwell JH, Hamilton GW, KennedyJW (1980) Serial exercise radionuclide angjography. Circulation61: 600-609

Thadani V, Parker JO (1978) Hemodynamics at rest and duringsupine and sitting bicycle exercise in normal subjects. Am JCardiol 41: 52-59

Varner SF, Pagani M (1976) Cardiovascular adjustments to exer-cise: Hemodynamics and mechanisms. Prog Cardiovasc Dis 19:91-108

Varner SF, Franklin D, Higgjns CB, Patrick T, Braunwald E (1972)Left ventricular response to severe exertion in untethered dogs.J Clin Invest 51: 3052-3060

Wildenthal K, Mitchell JH (1969) Dimensional analysis of the leftventricle in unanesthetized dogs. J Appl Physiol 27: 115-119

INDEX TERMS: Exercise • Hemodynamics • Stroke volume

by guest on June 6, 2018http://circres.ahajournals.org/

Dow

nloaded from

M B Higginbotham, K G Morris, R S Williams, P A McHale, R E Coleman and F R Cobbman.

Regulation of stroke volume during submaximal and maximal upright exercise in normal

Print ISSN: 0009-7330. Online ISSN: 1524-4571 Copyright © 1986 American Heart Association, Inc. All rights reserved.is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation Research

doi: 10.1161/01.RES.58.2.2811986;58:281-291Circ Res. 

http://circres.ahajournals.org/content/58/2/281World Wide Web at:

The online version of this article, along with updated information and services, is located on the

  http://circres.ahajournals.org//subscriptions/

is online at: Circulation Research Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. Further informationEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculation Research Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:

by guest on June 6, 2018http://circres.ahajournals.org/

Dow

nloaded from