Effect of Ascending an Ordinary Flight of

Stairs on the Work of the Heart

Observations on Normal Individuals and on Patients withCoronary Heart Disease

By JAMES A. L. MATHERS, M.D., HAROLD I. GRIFFEATH, M.D., ROBERT L. LEVY, M.D.,AND JOHN L. NICKERSON, PH.D.

Observations of cardiac work were made in normal persons and in patients with coronary heartdisease after ascending an ordinary staircase at various rates of speed. Cardiac output was cal-culated from records obtained with the low frequency, critically-damped ballistocardiograph.The results were compared, for statistical significance, with those noted following descent and afterwalking for an equivalent distance on the level. It appears that the compensated coronary patientcan mount a flight of stairs leisurely without imposing a greatly increased burden of work on theheart. The optimal rate of ascent usually is the one chosen by the individual.

IT IS common medical practice to advisepatients with cardiac disease to avoid theuse of stairs when feasible; in the presence

of varying degrees of cardiac or coronary in-sufficiency many are cautioned against ascentat any time. It thus becomes necessary forthese persons either to live on the ground flooror to resort to an elevator. Such arrangementsmay be difficult or impossible to make and fre-quently are a source of inconvenience and an-noyance.That mounting an ordinary flight of stairs

places an added burden of work on the heartand circulation requires no demonstration. Butthe factors which determine the importance ofsuch an effort for the patient with heart diseaseare unknown. What is the increment of in-crease? How does this compare with the changesobserved after descending stairs or walkingon level ground? What effect on the result have

From the Departments of Medicine and Physiol-ogy, College of Physicians and Surgeons, ColumbiaUniversity, and the Presbyterian Hospital.

This investigation was carried out under a con-tract between the Army Research and DevelopmentBoard, Office of the Surgeon General, and ColumbiaUniversity, and was aided by a grant from the Di-vision of Research Grants and Fellowships of the Na-tional Institutes of Health, U. S. Public Health Serv-ice.

Presented at the First International CardiologicalCongress, Sept. 7, 1950, Paris, France.

W 224

varying rates of performance? What is theduration of the changes in circulatory dynam-ics? What is their magnitude and duration incardiac patients in comparison with those seenin normal individuals under similar conditions?These are some of the questions which must beanswered if sound principles of guidance areto be followed.

This study is the first of a series in which suchmatters are to be considered. The present ob-servations were made on normal persons andpatients with coronary heart disease after per-forming various types of exercise on a staircaseand walking on the level. It is planned to makesimilar studies in patients with other etiologictypes of cardiovascular disease and also afterother forms of effort, such as ascending a ramp.

CLINICAL MATERIAL AND PROCEDURE

Observations were made on 5 normal persons rang-ing in age from 41 to 77 years of age (average, 54.6years), and in 5 patients with coronary heart diseaseranging in age from 44 to 65 years (average, 55.2years). Each group was composed of 4 men and 1woman. Two of the patients had healed myocardialinfarcts; one of them had been in congestive heartfailure and was taking maintenance doses of digitoxinas well as a diet low in salt content. The diagnosis ofcoronary heart disease was confirmed in the otherpatients by the presence of a positive anoxemia test;in 2 of them, an electrocardiogram taken during aspontaneous attack of anginal pain showed alter-ations characteristic of acute coronary insufficiency.

Circulation, Volume III, February, 1951

by guest on July 3, 2018http://circ.ahajournals.org/

Dow

nloaded from

MATHERS, GRIFFEATH, LEVY AND NICKERSON

A complete history was taken and a physical exami-nation, electrocardiogram, 2 meter film of the heartand a hemoglobin estimation were made in eachsubject.

The observations consisted of the following: (1)determination of cardiac output four times at 3minute intervals during a resting period and duringthe period of recovery after the exercise, at intervalsof 1, 2, 3, 5, 7, 9, 12, 15, 20, 25 and 30 minutes;(2) recording of systolic and diastolic blood pressuresand of heart rate at one minute intervals throughout.The cardiac output was calculated from records ob-tained with the use of the low frequency, critically-damped ballistocardiograph designed by Nickersonand Curtis' and tested by comparison with the directFick method by Nickerson, Warren and Brannon.2Lead II of the electrocardiogram was recordedsimultaneously with the ballistocardiogram. Therecords were made without any attempt to controlrespiration. In the course of the work, more than 4000ballistocardiograms were taken.The exercises employed were of three types: (1)

walking up one flight of stairs; (2) walking down one

flight of stairs; (3) walking down the hall and backto the laboratory for an equivalent distance. Alltests were carried out under basal conditions andafter the subject had rested for 30 minutes in therecumbent position on an inflated air mattress on theballistocardiographic table. Following the restingperiod, the subject was transported by wheelchairand elevator to the bottom or top of the staircase forthe test of the day. The blood pressure cuff andelectrocardiograplhic electrodes remained on the ex-

tremities during the exercise to facilitate the speedof making the observations during the recoveryperiod. All obstructions, such as closed doors andotherpersons, wereremoved from the path of the sub-ject being tested.

The three types of exercise were repeated twice atthe individual's own chosen rate of performance. Theeffect of increasing or decreasing this rate was deter-mined twice for each type of exercise in the 5 normalpersons and in 3 of the cardiac patients. In no

instance was more than one test given in the course

of 24 hours.The staircase was of the conventional variety and

was in the Medical School building. It consisted of2 sections of 11 treads each separated by an 8 footlanding. Each tread measured 11 inches in width,with a rise of 7 inches. The distance from the firststep to the 'ballistocardiographic table was 70 feetand from the last step to the table was 34 feet. Thetime elapsing between the conclusion of the exerciseand the beginning of the observations on the tablewas less than 30 seconds.A control maneuver was performed in duplicate

during which the same observations were made as

already described. However, instead of performingthe exercise, the subject was transported by wheel-chair and elevator up or down one floor and returned

to the ballistocardiograph without leaving the chair.He then stood erect for 30 seconds before resumingthe recumbent position on the table. In this way theeffects of all factors minus the prescribed exercises,could be determined.The cardiac work was calculated according to the

formula

CW = CO X MP

where CW = cardiac work, CO = cardiac output(liters per minute), and MP = mean pressure (dia-stolic blood pressure + i pulse pressure, in mm. Hg).

NORMALS- ASCENT

leart Rate

RECOVERY o

S 10 IS 20 25 30TIME IN MINUTES

FIG. 1. Averages of changes in circulatory dy-namics in 5 normal persons after ascending a flight ofstairs at each individual's own chosen rate. Arrowmarks ascent.

RESULTSThese are shown in figures 1 to 6. They are

expressed as averages for the various groupsunder consideration; the figures have been sub-mitted to statistical analysis.* It is clear that,although the number of observations runs intothousands, the number of subjects is relativelysmall. In order that the changes associatedwith exercise shall be of significant magnitudeit is necessary that they should be compara-tively large so as to cover the great variationswithin individual performances as well as varia-

* Dr. John W. Fertig, Professor of Biostatistics,School of Hygiene and Public Health of the Collegeof Physicians and Surgeons, Columbia University,furnished helpful advice in making the statisticalanalyses.

225

by guest on July 3, 2018http://circ.ahajournals.org/

Dow

nloaded from

EFFECT ON HEART OF ASCENDING STAIRS

tions between individuals. When significant dif-ferences are apparent they are of importance;when the differences are not significant although

CORONARY HEART DISEASE-ASCENT

A8OrR Heart Rate

70

CCONTROL RECOVERY

0 5 t00 5 10 Is 20 25 30

TIME IN MINUTES

FIG. 2. Averages of changes in circulatory dy-namics in 5 patients with coronary heart disease afterascending a flight of stairs at each individual's own

chosen rate. Arrow marks ascent.

NORMALS- DESCENT

5:? 70 r R Heart Rates, 60' - vS CONTROL RECOVERY

0 5 10

5 10 Is 20 25 30TIME IN MINUTES

FIG. 3. Averages of changes in circulatory dy-namics in 5 normal persons after descending a flightof stairs at each individual's own chosen rate. Arrowmarks descent.

manifest, the lack of significance cannot bestressed, for negative conclusions might notbe verified in a larger experience. In a few in-

stances, borderline values were obtained andattention will be called to them in the proper

CORONARY HEART DISEASE- DESCENT

FIG. 4. Averages of changes in circulatory dy-namics in 5 patients with coronary heart disease afterdescending a flight of stairs at each individual's own

chosen rate. Arrow marks descent.

NORMALS-LEVEL

FIG. 5. Averages of changes in circulatory dy-namics in 5 normal persons after walking on the levelat each individual's own chosen rate. Arrow marksthe walk.

places. For each group the changes are ex-

pressed as increases over the levels observed inthe resting period for that particular group;

226

h

by guest on July 3, 2018http://circ.ahajournals.org/

Dow

nloaded from

MATHERS, GRIFFEATH, LEVY AND NICKERSON

by so doing, the effects of the exercise aloneare measured.

Ascent of Stairs. In the normal subjects, thereoccurred a significant increase in cardiac work,stroke volume and heart rate. The cardiacwork increased 226 units* (54 per cent), thestroke volume 26 cc. (32 per cent) and the heartrate 8 beats (13 per cent). In the patients withcoronary heart disease there was a significantincrease in pulse pressure, cardiac work andstroke volume. The pulse pressure rose 18 mm.Hg (27 per cent), the cardiac work increased323 units (75 per cent) and the stroke volume30 cc. (44 per cent). The change in heart ratewas small, with an average increase of only 9beats per minute; this amount is not statis-tically significant.The notable increases occurred in cardiac

work and, to a somewhat lesser degree, in strokevolume in both normal subjects and cardiacpatients. The changes were greater in the car-diac group but the differences between the twogroups were not larger than might have beencaused by chance variation.

Descent of Stairs. Although there were in-creases in the figures obtained for all four fac-tors observed, these were not statistically sig-nificant in either the normal or the cardiacgroup; nor were there significant differencesbetween the two groups. It should be noted,however, that in the normal group cardiacwork increased 56 units (12 per cent) and inthe cardiac group it increased 148 units (42per cent). Similarly, in the normals, the strokevolume increased 5 cc. (2 per cent) and in thecardiacs 12 cc. (26 per cent). The increaseswere thus consistently greater in the cardiacgroup.

Walking on Level Ground. In the normalgroup, there was a borderline increase in pulsepressure and in the cardiac group, a similarlyequivocal increase in stroke volume. Thechanges were not striking, either by themselvesor on comparison of the differences betweenthe two groups.

Control Maneuver. There were no significantincreases in either group in any of the circula-

* The unit of cardiac work is obtained by multi-plying the figure for cardiac output, in liters, by thatfor mean pressure, in mm. Hg.

tory factors measured, nor were there note-worthy differences between the two groups.The changes throughout were numerically rela-tively small.

Varying the Rate of Performance of the Ex-ercises. The chosen rate of speed of carrying outascent, for the normals, was, on the average29.6 seconds, with a range of 25 to 37; for thecardiacs, the average was 36.4 seconds with arange of 25 to 47. For descent, the average timefor the normals was 26.7 seconds, with a rangeof 19 to 34; for the cardiacs, it was 31.1 sec-onds with a range of 23 to 37. For walking on

CORONARY HEART DISEASE-LEVEL

z

N,

0 5 10 IS 20 25 30TIME IN MINUTES

FIG. 6. Averages of changes in circulatory dy-namics in 5 patients with coronary heart disease afterwalking on the level at each individual's own chosenrate. Arrow marks the walk.

the level, the average for the normals was 33.4seconds, with a range of 22 to 42; for the cardi-acs, it was 37.3 seconds, with a range of 31 to47. Deliberately increasing or decreasing therate of performance approximately 20 to 60per cent did not significantly alter the results.For the normal group, in both ascent and

descent, the individual's chosen rate of per-formance was the most rapid. At the request ofthe observer, the time was slowed down; theaverage slow rate was 47 seconds. For the car-diac group, the fastest average rate for ascentwas 27.3 seconds; this was slowed, by request,in those whose original rate was faster than 30seconds, to an average of 44 seconds; for de-

227

by guest on July 3, 2018http://circ.ahajournals.org/

Dow

nloaded from

EFFECT ON HEART OF ASCENDING STAIRS

scent, the fastest average rate was 26 seconds,which was slowed to 40 seconds for later ob-servations. The normals, then, all slowed up byrequest; the cardiacs were less uniform in theirchosen rates of performance and were askedto accelerate or retard as much as 20 per centin either direction.Although the average differences were not

statistically significant, in most instances car-diac work was minimized when, in the ascentof stairs, both normals and cardiac patientswere permitted to proceed without guidance.The individual, either as a result of habit or,in the case of the cardiacs, because of previousinstruction, instinctively chose what for himwas the optimal rate. Hickam and his group,3using both the Nickerson ballistocardiographand the Fick method after cardiac catheteriza-tion, and Wolff and his associates,4 using thesame ballistocardiograph, have observed thatordinary physical exertions of everyday lifeperformed during periods of relaxation imposelittle extra work on the heart, whereas, whencarried out under emotional stress, they causean increase in cardiac output. It appears fromour experience that a conscious effort to modifythe rate of performance of an exercise, evenslightly, also may result in an augmentationof cardiac work.

Duration of Changes in Circulatory Dynam-ics after the Exercises. In the normal group thereturn to resting levels was completed at theend of the third minute during the period ofrecovery. The three-minute observation waschosen, therefore, as a reference point formeasuring the duration of the increases noted.In the coronary group, after ascent, there wasa significant delay beyond three minutes inthe return to average resting values for cardiacwork and a similar lag, of borderline signifi-cance, for stroke volume. However, in bothinstances, the delay was only a matter of afew minutes (fig. 2). There was no significantdelay in the return to resting levels after theother exercises, in either group of subjects. Inthe normals, on descent and after walkingon the level, there was a tendency for cardiacwork to fall slightly below resting levels at thetime of the three minute observation; thisdid not occur in the cardiac patients.

DISCUSSION

Numerous studies have been made of cardiacoutput and cardiac work after effort, using avariety of methods.' For the most part, theexercises employed have been quite vigorousand have been performed at rates of speedfixed by the investigator. In the present study,the exercises used were moderate and, in thebasic experiments, were carried out at a ratechosen by the individual subjects. Previousresults and our own are not, therefore, strictlycomparable. In addition, as far as can be de-termined, in no other study have statisticalmethods been applied to determine the signif-icance of observed differences from controllevels or of variations between normal indi-viduals and patients with cardiovascular dis-ease.

In this investigation, also, the usefulness ofthe low frequency, critically-damped ballisto-cardiograph has been demonstrated for makingpossible a large series of observations of car-diac output in each subject. The determinationsof stroke volume and cardiac output were madeon ballistic recordings taken without any con-straint on the respiration. For some individualswith cardiac disease the ballistocardiographicpatterns deviate from the normal form and theresults for the stroke volume may differ some-what from the absolute values. However, therelative changes in stroke volume are givenquite accurately. The accuracy of the methodfor relative change was originally demonstratedby Nickerson, Warren and Brannon2 and isfurther supported, in the present study, by themanner in which the curve for stroke volumeparallels that for pulse pressure.

In dealing with group averages in a smallseries, it is inevitable that a wide range of indi-vidual variability will reduce the incidence ofstatistical significance of even comparativelylarge numerical differences between the groupsconsidered. Under the conditions of these tests,it was observed that there was considerableindividual variation.The levels for the averages for cardiac work

and for stroke volume during the resting periodwere lower for the cardiac group than for thenormals in every series of comparisons. While

228

by guest on July 3, 2018http://circ.ahajournals.org/

Dow

nloaded from

MATHERS, GRIFFEATH, LEVY AND NICKERSON

the difference was often considerable it cannotbe termed significant in view of the large indi-vidual variations exhibited by both groups.Thus, although no symptoms or signs of con-gestive failure are present, the functional capac-ity of the heart, even at rest, apparently issomewhat diminished in patients with coronaryheart disease. A similar observation was madeby Starr and Wood,6 using the ballistocardio-graphic method, and according to these ob-servers, cardiac work, in the chronic form of thisdisorder, tends to decrease as the condition ad-vances.

This study was undertaken primarily to ob-tain information of practical value in the guid-ance of patients with coronary heart diseasewith respect to the use of stairs. The inferencesto be derived from our observations are some-what at variance with current beliefs. It ap-pears that ascending an ordinary flight of stairsat an individual's own chosen rate of speed,which is commonly a leisurely one, does resultin a significant increase in the work of theheart. This is due chiefly to augmentation incardiac output, since there is but little changein the heart rate. The differences between as-cent on the one hand, and descent and walkingon the level for an equivalent distance on theother, are in general, rather large numericallybut are not, in all cases, statistically signifi-cant because of the individual variations. Theamount of cardiac work observed in descentis, as might be anticipated, intermediate be-tween that involved in ascent and that re-quired for walking on the level. The absence ofany notable increase in the rate of the heartfollowing all three types of exercise, in bothnormal and cardiac groups, is also worthy ofrepeated mention. None of the 5 patients ex-perienced discomfort of any sort during orfollowing the tests performed in the course ofthis study.

SUMMARY AND CONCLUSIONS

In a series of 5 normal persons and 5 pa-tients with coronary heart disease, observationswere made after ascent of an ordinary flight ofstairs, descent of these stairs and walking foran equivalent distance on the level. The low

frequency, critically-damped ballistocardio-graph was used to measure cardiac output.Cardiac work, stroke volume, pulse pressureand heart rate were recorded. The results weresubmitted to statistical analysis.The ascent of stairs does not impose a greatly

increased burden of work on the heart. Theresponse of the patient with compensated cor-onary disease is not significantly greater, fol-lowing this exercise, than after descent and isonly slightly greater than after walking for anequivalent distance on the level. The stress isof relatively brief duration. Obviously, in thepresence of congestive failure or if anginal painoccurs during the exertion, stairs should beavoided. Otherwise, the coronary patient maytake them leisurely and preferably at a pacewhich he finds agreeable. Deliberate retarda-tion, like hurry, may augment cardiac work.

ACKNOWLEDGMENTSFor technical assistance the authors are indebted

to Mr. Vincent Glaviano, Mrs. Bernadine Miller andMiss Isabella Clarke.

REFERENCESNICKERSON, J. L., AND CURTIS, H. J.: The design

of the ballistocardiograph. Am. J. Physiol. 142:1, 1944. : The low frequency, critically-damped ballistocardiograph. Federation Proc.4: 201, 1945.

2 , \WARREN, J. V., AND BRANNON, E. S.: Thecardiac output in man: Studies with the lowfrequency, critically-damped ballistocardiograph,and the method of right atrial catheterization. J.Clin. Investigation 26: 1, 1947.

3 HICKAM, J. B., CARGILL, W. H., AND GOLDEN,ABNER: Cardiovascular reactions to emotionalstimuli. Effect on the cardiac output, arterio-venous oxygen difference, arterial pressure, andperipheral resistance. J. Clin. Investigation 27:290, 1948.

4STEVENSON, I. P., DUNCAN, C. H., AND WOLFF, H.G.: Circulatory dynamics before and after exer-cise in subjects with and without structural heartdisease during anxiety and relaxation. J. Clin.Investigation 28: 1534, 1949.

5 LILJESTRAND, G., LYSHOLM, E., AND NYLIN, G.:The immediate effects of muscular work on thestroke and heart volume in man. Skandinav.Arch. f. Physiol. 80: 265, 1938.

6 STARR, I., AND WOOD, F. C.: Studies with theballistocardiograph in acute cardiac infarctionand chronic angina pectoris. Am. Heart J. 25:81, 1943.

229

by guest on July 3, 2018http://circ.ahajournals.org/

Dow

nloaded from

NICKERSONJAMES A. L. MATHERS, HAROLD I. GRIFFEATH, ROBERT L. LEVY and JOHN L.Observations on Normal Individuals and on Patients with Coronary Heart Disease

Effect of Ascending an Ordinary Flight of Stairs on the Work of the Heart:

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 1951 American Heart Association, Inc. All rights reserved.

is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation doi: 10.1161/01.CIR.3.2.224

1951;3:224-229Circulation. 

http://circ.ahajournals.org/content/3/2/224the World Wide Web at:

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

  http://circ.ahajournals.org//subscriptions/

is online at: Circulation 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 Further information about this process is available in therequested is located, click Request Permissions in the middle column of the Web page under Services.the Editorial Office. Once the online version of the published article for which permission is being

can be obtained via RightsLink, a service of the Copyright Clearance Center, notCirculationpublished in Requests for permissions to reproduce figures, tables, or portions of articles originallyPermissions:

by guest on July 3, 2018http://circ.ahajournals.org/

Dow

nloaded from