The Prognosis of an AbnormalElectrocardiographic Stress Test

By JosEPH T. DOYLE, M.D., AND SANmRA H. KiNciH, M.S.

SUMMARYTreadmill walking for 10 min at 3 mph against a 5% grade has been used as an elec-

trocardiographic (ECG) stress test as part of a prospective epidemiologic study of 2,437men. From 1953 through 1966, 22,223 tests have been done without untoward event.The criteria of an abnormal postexercise ECG were ischemic flattening or coving

of the S-T segment, T-wave changes consistent with focal left ventricular epicardialischemia, and paroxysmal left bundle-branch block.Of the 2,003 men exercised two or more times, 264 developed some manifestation

of ischemic heart disease (IHD) and in 75 (30%) this was an abnormal ECG response

to exercise. They had higher blood pressures and were more often heavy smokersthan normal responders. Body weight and serum cholesterol were similar in the twogroups. Over the next 5 years there was an 85% probability that these abnormal re-

sponders would develop angina pectoris or experience a myocardial infarction.These relatively insensitive but highly specific and reproducible ECG criteria accu-

rately identify men with clinically silent but far-advanced coronary atherosclerosis,attested by the poor prognosis of an abnormal response. An abnormal postexerciseECG is valid evidence of IHD. A submaximal ECG stress test is useful in clinicaland epidemiologic studies and might be useful in assessing the effectiveness of effortsto reduce the risk of IHD.

Additional Indexing Words:Ischemic (coronary) heart disease Exercise

IT IS generally, although not universally,accepted that transient abnormalities of

intraventricular conduction and of ventricularrepolarization appearing during or after exer-cise are manifestations of reversible myocardi-al ischemia. In the absence of myocardial orvalvar dysfunction, anemia, electrolyte dis-turbance, or digitalis, these electrocardio-graphic abnormalities are imputed to theinability of coronary arteries narrowed byatheroma to deliver enough blood to meet theimmediate energy requirements of the in-creased myocardial work load. The electro-cardiographic criteria of induced myocardial

From the Department of Medicine, Division ofCardiology, and the Cardiovascular Health Center,Albany Medical College, and the New York StateDepartment of Health, Albany, New York.

Received May 19, 1969; accepted for publicationNovember 11, 1969.

Ciculation, Volume XLI, March 1970

Treadmill Epidemiology

ischemia and their sensitivity and specificityhave been long and inconclusively argued.Many types of exercise tests have been used,mostly submaximal but some maximal.Whether exercise tests can be standardized or,indeed, whether standardization is reallynecessary is a still unresolved question.1-9 Thisreport is concerned not, however, with theseproblems but with the frequency with whichischemic heart disease (IHD) can be inferredfrom arbitrary electrocardiographic criteria inmiddle-aged men subjected to a moderatelystiff exercise test. The topic of exerciseelectrocardiography has been extensively andauthoritatively reviewed, most recently byBruce and Hornsten.9

MethodsThe participants in a prospective study of

degenerative circulatory disease, known as theAlbany Cardiovascular Health Center, have beenroutinely subjected to an electrocardiographic

545

by guest on May 29, 2018

http://circ.ahajournals.org/D

ownloaded from

DOYLE, KINCH

stress test. The Cardiovascular Health Centerproject was established at the Albany MedicalCollege by the New York State Health Depart-ment in 1952 and has been fully describedelsewhere.10 In brief, 1,913 male employees ofNew York State born in the years 1898 to 1913and working in or near Albany (89% of thoseeligible), have agreed to place themselves undercontinuous medical surveillance. On entry intothe study a comprehensive lifetime medical andsocioeconomic history was obtained, a completephysical examination was done, and a number oflaboratory tests were perfonned. The testsincluded measurements of hemoglobin andhematocrit readings, urinalysis, anteroposteriorand left lateral roentgenograms of the heart andlungs, an electrocardiogram, and the blood lipidprofile. No participant was excluded because ofpreexisting heart disease. At least biennially, andusually annually, since he has come underobservation the interim medical history of eachparticipant has been ascertained, and the physicalexamination as well as the battery of laboratorytests already enumerated has been repeated. Theannual attrition rate from all causes has been lessthan 5%. Only 33 men are completely lost tofollow-up. An additional 621 men, mostly mem-bers of the New York State Legislature andappointed State officials have been graduallyadded to the base study and have been examinedone or more times. Since this group differs in noimportant respect from the core group, they havebeen included in this report.A major objective of the Cardiovascular Health

Center has been to evaluate the worth of variousdiagnostic methods in the detection of preclinicalIHD. An electrocardiographic stress test has,accordingly, been a part of the examinationroutine since the inception of the study. Althougha maximal stress test should have a higherdiagnostic yield, the fear of medicolegal compli-cations has dictated the use of a submaximal test.The first 617 participants did a double two-steptest exactly as specified by Master.1 All subse-quent tests have consisted of walking for 10 minat 3 mph on a motor-driven treadmill set at a 5%grade.4 11

Tests are performed only at the discretion ofthe physician who has reviewed the history, donethe physical examination, and examined theresting electrocardiogram. Electrocardiographicabnormality does not exclude the subject from theexercise test. Usually, although not invariably, awell-documented history of previous myocardialinfarction or of typical Heberden's angina hasbeen considered a contraindication to the exercisetest. The only other contraindication has beenorthopedic disability. No attempt is made tocontrol eating or smoking. The test is supervised

by a technician although a physician is always inthe immediate vicinity. The participant is cau-tioned to stop exercise should he develop pain,breathlessness, apprehension, or unusual fatigue.The technician is instructed to stop the testimmediately if, in her judgment, the subjectappears to be in trouble.

All electrocardiograms are recorded in recum-bency on a direct-writing multichannel instru-ment at a paper speed of 50 mm/sec. The limb-lead electrodes are left in place during exercise.Suction leads are used on the chest. Immediatelyafter exercise and again 3 min later the lateralprecordial leads and leads III, II, and I arerecorded. Leads taken at 6 min were found tocontribute no further diagnostic information andare no longer recorded. Continuous direct-wiremonitoring of the precordial electrocardiogramhas revealed no changes confined to, or peculiarto, the exercise period in a small group of subjectsin the study and in 100 medical students whoperformed a maximal exercise test. Indeed,abnormalities considered to be evidence of acutemyocardial ischemia induced by the exercise testsare typically persistent for some minutes.From February 1953 through December 1966,

2,437 of 2,534 men have had one or more exercisetests. A total of 22,223 exercise tests has beendone on these men. Two hundred four tests on154 men were incomplete. The most commonreasons for stopping tests before completionincluded dyspnea, fatigue, leg pain or cramps,dizziness, intense anxiety and, in a very fewindividuals, the appearance of anginal pain.There have been no recognized complications ofthe test and no fatalities.

Initially, the criteria of Master were rigorouslyapplied to the postexercise electrocardiogram.' Asthen promulgated, by these criteria almost anychange in the postexercise electrocardiogramother than an increase in heart rate was construedas evidence of coronary insufficiency. Moreover,the absence of such changes in the postexerciseelectrocardiogram was held to exclude coronaryheart disease. Fewer than 50% of our symptom-less, vigorous, and apparently healthy men couldpass the Master's test so evaluated. Although itwas conceded that most of these men indubitablyhad some degree of coronary atherosclerosis, itseemed implausible that so large a number hadclinically significant impairment of coronary bloodflow under the conditions of the double two-steptest.

It was, therefore, decided early in 1954 torequire as evidence of induced acute coronaryinsufficiency one or more of the following:(1) flattening or downward inclination for 0.10sec or longer or cove-plane deformity of the S-Tsegment irrespective of the presence or absence of

Ci7rculation, Volume XLI, March 1970

546

by guest on May 29, 2018

http://circ.ahajournals.org/D

ownloaded from

ECG STRESS TEST

junctional (J) displacement; (2) a change in thedirection of the T wave consistent with focal leftventricular epicardial ischemia; or, (3) paroxys-mal left bundle-branch block. These criteria wereadmittedly arbitrary and were based on acombination of empirical observation and theoret-ical assumption.Not long afterwards at a conference on March

8 and 9, 1956, at the Walter Reed Army Hospital,Mattingly and Robb presented rather similarcriteria used by them for some years andvalidated on the basis of a prospective study of alarge group of military personnel followed for 10years.'2 The QX:QT ratio has occasionally beenhelpful in verifying a suspected ST-segmentdeformity.5 Equivocal results of electrocardio-graphic stress tests are for clinical and statisticalpurposes considered to be normal. The degree ofabnormality of abnormal exercise responses wasalso classified by the method of Robb and Marks.8All exercise tests have been interpreted by thesame reader and reviewed blind at least once.The repeatability of readings has been 96%.The life table technic has been used to assess

the prognostic implication of the response to theelectrocardiographic stress tests. This is a well-known statistical device which has advantagesover the commonly used cumulative or averageannual incidence rate. The actual period ofexposure to risk is taken into account in theanalysis, a particularly important procedure since,as the period of observation increases in prospec-tive studies, increasing numbers of the group maybe lost to observation or succumb to some diseaseother than the one under study. As the method isused here, actual months of observation ratherthan the customary one-half interval estimate areused in obtaining person years at risk. Age andcalendar time have been confounded on thefundamental assumption that the age-specific riskof ischemic heart disease (IHD) has remainedrelatively constant over calendar time. Length offollow-up is calculated by subtracting the age atentry time 0 from the year of withdrawal.

For each year of age and for cumulativeperiods of 5, 10, 15, and 20 years after age 40,the probabilities of developing the event understudy, in the presence of the competing risk, arecalculated. The primary event and the competingevent are specifically defined for each comparison.This method, its assumptions, and a detaileddescription of the calculations appear in thetechnical appendix of an earlier publication.'3Two sets of probabilities have been prepared:

The first sets forth the probabilities of anindividual's developing an abnormal electrocar-diographic response to exercise as the primarymanifestation of IHD. The competing risk is anyother manifestation of IHD, death from any causeCirculation, Volume XLI, March 1970

other than IHD, or loss to observation. Theindividual enters the life table at the time of hisfirst exercise test. The second sets forth theprobabilities of developing an additional manifes-tation of IHD following an abnormal exerciseresponse. The competing risk is death from anycause other than IHD or loss to observation. Theindividual enters the life table experience at thetime of his first abnormal response and iswithdrawn at the first event of IHD or at his lastvisit to the Center.To evaluate the risk factor status of the

individuals with an abnormal exercise responsecompared with those who had no evidence ofIHD, a sample of 75 controls was selected fromthe total disease-free population matched to thecase group by date of birth and date ofobservation. The date of observation was the dateof the first diagnosis of an abnormal electrocardio-graphic stress test. The variables compared wereserum total cholesterol concentration, systolic anddiastolic blood pressures, total body weight, andcigarette smoking. A matched-pair analysis wasapplied to the actual measurements for the firstfour variables while the proportion of currentcigarette smokers in the two groups wascompared.

ResultsOf a total of 2,534 men examined at the

Cardiovascular Health Center, 2,437 wereexercised at least once. The following cate-gories have been excluded from analysis: (1)97 men who were not exercised, mostlybecause of orthopedic disabilities; (2) 45 menwith antecedent IHD (myocardial infarction,23; angina pectoris, 22); and (3) 361 menwith normal postexercise electrocardiogramswho continued under surveillance but werenot again stress tested. Two hundred andninety-one men of this third group were notreexamined either because they refused toreturn or because they had entered the studywithin less than 2 years of the cutoff date. The28 men who had an abnormal electrocardio-graphic response to their first exercise test areconsidered separately. Of the remaining 2,003men, 75 have subsequently exhibited abnor-mal exercise responses, 189 have had anothermanifestation of IHD, 99 have died fromother causes, and 33 have been lost to follow-up. Severe myocardial or valvar disease,anemia, and digitalis have been excluded aspossible causes of exercise-induced electrocar-

547

by guest on May 29, 2018

http://circ.ahajournals.org/D

ownloaded from

DOYLE, KINCH

STUDY POPULATIONDECEMBER, 1966

l EXCLUDED FROM ANALYSIS

Figure 1Population subjected to an electrocardiographic stressbetween February 1953 and December 1966. Ab-breviations: AX=abnormal ECG response to exer-cise; IHD = ischemic heart disease.

diographic abnormalities. Serum electrolyteshave been assumed to have been normal.Hyperventilation has not been recognized.The first manifestation of IHD has been anabnormal electrocardiographic response toexercise in nearly one third of all cases. Figure1 summarizes this enumeration.

Seventy-five men had an abnormal electro-cardiographic response to exercise. Thirty-four(45%) subsequently developed other evidenceof IHD in the period covered by this report.Fifty-seven had an unequivocally normalresting electrocardiogram. Twenty-four ofthese men (41%) subsequently developedother evidence of IHD. Only six (11%) had apostexercise electrocardiogram with a junc-tional depression of 2 mm or more (Robb,grade 3) .8 The other 17 men had restingelectrocardiographic abnormalities due aboutequally to left ventricular enlargement conse-quent to arterial hypertension or to nonspe-

cific T-wave changes. One third of this smallergroup had Robb grade 3 changes in thepostexercise electrocardiogram and nearly twothirds10 later developed other manifestationsof IHD.

Figure 2 shows the 5-year cumulativeprobabilities after age 40 of not developing anabnornal exercise response as the first mani-festation of IHD. This probability declinedfrom 998 per 1,000 persons (99.8%) during thefirst year to 940 per 1,000 (94.0%) over a 20-year span.

Figure 3 shows the 5-year cumulativeprobabilities of not developing angina pectorisor a myocardial infarction or dying from IHDfollowing an abnormal response to exercise.The rate declines from 136 per 1,000 duringthe first 5 years to 6 per 1,000 within 20 yearsin the 75 men who developed abnormal testswhile under observation. The- probability of-surviving IHD over the same period in theabsence of an abnormal stress test declinesfrom 985 per 1,000 in the first 5 years to 820per 1,000 by age 59. The 28 men with anabnormal electrocardiographic response toexercise on admission also fared badly. Thecumulative probability of their not developing

1.00-

5 0.75-

-4

E: 0 50-

0.25

0.00-_0 5 10 15

YEARS AFTER AGE 4020

Figure 2

Probability of a continued normal electrocardiographicresponse to an exercise test after age 40. There was a99.8% probability for the first year and a small declineto 94.0% after 20 years.

Circulaion, Volume XLI, March 1970

C5A,

by guest on May 29, 2018

http://circ.ahajournals.org/D

ownloaded from

ECG STRESS TEST

1.00-

< 0.75-

E: 0.50-

0.25-

0.00-0 5 10 15

YEARS AFTER AGE 4020 25

Figure 3

Probability of not developing (that is, remaining freeof) another manifestation of ischemic heart diseaseaccording to electrocardiographic response to exercise.

another manifestation of. IHD over 20 years

was 4%.Data for the total 103 men with abnormal

tests have not been combined since the twogroups are not comparable. The 28 men withan abnormal test at entry represent survivorsof a cohort whose person years of exposure torisk of IHD after an abnormal electrocardio-graphic response to exercise is unknown sinceno data on experience prior to entry are

available. Furthermore, the limited numbersof person years of experience at the 5-yearintervals after age 40 are so few as to makethe probabilities relatively unstable. For simi-lar reasons persons with myocardial infarction

and angina pectoris at entry were excludedfrom the present analysis.An attempt has been made to evaluate the

relationship of the degree of abnormality toprognosis. Unfortunately, subdivision of thedata results in very small frequencies in eachabnormality group. Disease rates based onthese are subject to large standard errors.Therefore, it is very difficult to draw reliableconclusions.

Tables 1 and 2 show the 28 individuals withan initial abnormal electrocardiographic re-sponse to exercise and the 75 with subsequentabnormal responses according to the restingelectrocardiogram and the degree of abnor-mality of the postexercise tracing.

In table 3 are considered selected riskfactors in relation to the electrocardiographicresponse to exercise. The mean levels and thestandardc deviations for seran total choles-terol, systolic and diastolic blood pressures,relative body weight, and the proportion ofcigarette smokers in the two groups aredisplayed. The mean levels for the personswith abnormal ECG-stress tests are higherand more variable than for those of thecontrols. These differences are significant onlyfor systolic and diastolic blood pressures andsmoking. This finding is in accord with otherdata which show higher levels of these factorsin diagnosed cases of IHD as compared withindividuals free of IHD. There were too fewcases to detect any influence on prognosis

Table 1

Prognosis of Development of Additional Evidence of Ischemic Heart Disease According toSeverity of Response after Entry: Abnormal Electrocardiographic Stress Test

Abnormal ECG stress test after entry

Total Normal resting ECG Abnormal resting ECG

Prognosist Prognosist PrognosistDegree of

abnormality* Total No. % Total No. % Total No. %

Total 75 34 45 57 24 42 18 10 56Grade la 40 15 37 34 13 38 6 2 33

lb 2 2 100 2 2 100 0 0 02 17 9 53 14 8 57 3 1 333 13 6 46 6 1 17 7 5 71

Left bundle 3 2 67 1 0 0 2 2 100

* See reference 8.t Based on number of persons in whom additional evidence ofIHD developed during study period.

Cicsdation, Volume XLI, Match 1970

NORMALEXERCISETEST N= 1928

\ \s ABNORMAL\ EXERCISE TEST\ AT ENTRY\ "N = 28

ABNORMALEXERCISE TEST

xAFTER ENTRY "N=75"

V.-

549Q

by guest on May 29, 2018

http://circ.ahajournals.org/D

ownloaded from

DOYLE, KINCH

Table 2

Prognosis According to Severity of Response at Entry: Abnormal Electrocardiographic StressTest at Entry

Abnormal ECG stress test at entry

Total Normal resting ECG Abnormal resting ECG

Prognosist Prognosist PrognosistDegree of

abnormality* Total No. % Total No. % Total No. %

Total 28 17 61 21 13 62 7 4 57Grade la 8 4 50 6 3 50 2 1 50

lb2 15 9 60 11 7 64 4 2 503 4 4 100 3 3 100 1 1 100

Other 1 0 1

* See reference 8.t Based on number of persons in whom additional evidence of IHS developed during study period.

resulting from a conjunction between one ormore of the risk factors considered and theseverity of the induced electrocardiographicabnormality rated by Robb's criteria. Never-theless, it is reasonable to suppose that acompounding of risk factors adversely in-fluences prognosis. The effect of increasingage was likewise obscured by the smallnumber of cases and the multiple confoundingeffects of the several risk factors. There is noreason to suppose, however, that the influenceof rising age, as displayed in figure 4, inincreasing the likelihood of an abnormalelectrocardiographic response to exercisediffers in any respect from its influence onother risk factors. The longer and the moreintense the exposure to risk of the susceptible

Table 3

Level of Selected Risk Factors in Men with AbnormalElectrocardiographic Responses to Exercise Comparedto Age-Matched Men with Normal Responses

ECG response to exercise

Abnormal NormalVariable (Mean i4 SD) (Mean i SD)

Serum total cholesterol 250 ± 43 241 ±t 43(mg/100 ml)

Systolic blood pressure 150 i 30 134 4 18(mm Hg)

Diastolic blood pressure 92 + 14 83 + 12(mm Hg)

Obesity (% ideal) 119 ± 15 117 i 14Cigarette smokers (%) 69 37

subject, the greater is the likelihooddisease.

Of

DiscussionNearly one third of the diagnoses of IHD

made in the Cardiovascular Health Centerhave rested on what has been construed as anabnormal electrocardiographic response towalking on a motor-driven treadmill at 3 mphagainst a 5% grade for 10 min. The very highsubsequent incidence of other manifestationsof IHD in men with abnormal postexerciseelectrocardiograms, namely, angina pectoris,myocardial infarction, and sudden death, is

40-

; 20O

040 45 50 55 60 65 70

AGE GROUP

Figure 4Incidence of abnormal electrocardiographic responseto exercise according to age; rate per 1000 person-years exposure.

C'rcsdl#i0os, Volum8 XLI, March 1970

550

by guest on May 29, 2018

http://circ.ahajournals.org/D

ownloaded from

ECG STRESS TEST

felt to confirm the validity of induced ischemicflattening of the S-T segment as evidence ofacute coronary insufficiency. Ischemic T-wavechanges and paroxysmal left bundle-branchblock were not useful criteria. Abnormalresponders weigh more, have higher bloodpressures, have higher serum total cholesterolconcentrations, and are heavier cigarettesmokers than men who have a normalelectrocardiographic response to exercise. Theprognosis for the 28 men who had anabnormal postexercise electrocardiogram onentry into the study has been similarlyunfavorable.

Brody's experience with the Master doubletwo-step test at the Greenbrier Clinic is mostnearly comparable with the present report.14He tested 756 men, of whom 23 hadelectrocardiographic responses considered ab-normal by criteria very similar to those used inthis study. Seventy per cent of these mendeveloped other manifestations of IHD in theensuing 3 to 10 years (six had myocardialinfarction and 10 "classical" angina pectoris).There is no other study with which the presentobservations can be cogently compared. Themilitary population followed by Mattingly12was originally stress tested because of com-plaints of chest pain. The insurance applicantsreported by Robb and Marks8 were exercisedin many instances because of complaints ofchest discomfort.In our opinion the observations presented

justify the acceptance of an abnormal electro-cardiographic response to exercise as anadditional category to the nosology of IHD.The intensity of the exercise test is probablymore important than the technic. Tread-mills are valuable research instruments, par-ticularly in applying work loads of progressiveseverity, but are not likely to be used outsidethe laboratory. On the other hand, the widelyknown and accepted two-step of Master caneasily, effectively, and safely be used in anyoffice. Even simpler is the single 12-inch steptest proposed by Simonson,"5 which couldeasily be utilized in field studies. The one-steptest, furthermore, is readily adaptable tocontinuous direct-wire recording of the elec-Circulation, Volume XU, Marcb 1970

trocardiogram; it is not only a cheaper but isalso a more reliable technic than teleme-try.'6The great likelihood of developing angina

pectoris, of sustaining a myocardial infarction,or of dying suddenly not long after anabnormal electrocardiographic response toexercise is evinced is clearly indicative of far-advanced atherosclerosis of the coronaryarteries. Close questioning of these abnormalresponders, however, has consistently failed toelicit symptoms that could in retrospect beconstrued as angina pectoris. The more severetransient electrocardiographic deformitiesinduced in hypertensive men with the electro-cardiographic pattern of left ventricularenlargement or with nonspecific T-wave ab-normalities can reasonably be construed asevidence of advanced coronary arterial insuffi-ciency, due either to arterial stenosis, myocar-dial hypertrophy or, most probably, to bothfactors. This interpretation is supported by thesubsequent high incidence of clinically overtIHD. The lower rate of severe exercise-induced abnormalities in individuals withnormal resting electrocardiograms is consis-tent with the view that coronary irrigation wasless severely compromised in these subjectsand is supported, although scarcely proved, bythe apparently lower subsequent incidence ofovert IHD. A priori, it seems unlikely that theabnormal reactors have coronary vasculaturemuch different from that of men whopresented initially with angina pectoris andwho, incidentally, did not necessarily respondto exercise with an acutely abnormal electro-cardiogram. It is possible although unprovablethat the abnormal reactors had alreadyunconsciously restricted their activity short ofthat which might precipitate angina. The highincidence of silent myocardial infarction,attested only by the symptomless progressionof the electrocardiogram from normal to apattern typical of intercurrent myocardialinfarction, is a clear instance of the notuncommon dissociation between the subjec-tive and the objective manifestations ofIHD.17, 18

51

by guest on May 29, 2018

http://circ.ahajournals.org/D

ownloaded from

DOYLE, KINCH

The high rate of other manifestations ofIHD experienced by men with an abnormalelectrocardiographic response to exercise sug-gests that this group might be particularlyuseful for testing regimens designed to lessenthis risk. The elimination of cigarette smoking,the reduction of arterial blood pressure, and aprogram of graded exercise, for example,might be evaluated separately or concurrentlyby serial electrocardiographic stress tests. Asignificant reduction in the incidence ofangina pectoris, of myocardial infarction, andof sudden death could fairly be construed asevidence for the effectiveness of these inter-ventions.

Thus, of the 2,003 men who had at least twoECG-stress tests through December 1966, 264men developed some manifestation of IHDand 75 (30%) of them had an abnormalelectrocardiographic response to exercise astheir first manifestation of IHD. Over the next5 years there was an 85% probability that these75 men would develop another manifestationof IHD, namely, angina pectoris, myocardialinfarction, or sudden death. Men who onadmission to the study had an abnormalelectrocardiographic stress test also had a highsubsequent incidence of IHD. Only transientischemic flattening or coving of the S-Tsegment induced by exercise proved to be adiagnostically useful index of potentiallyinadequate coronary arterial perfusion. Thesemen also had higher blood pressures and weremore often heavier smokers than men whoconsistently had normal electrocardiographicresponses to exercise. Relative body weightand serum total cholesterol concentrationwere not significantly different in the twogroups. Increasing age was associated with anincreasing likelihood of an abnormal electro-cardiographic response to exercise.

ConclusionsIt is concluded that the electrocardiographic

stress test used in this study reliably identifiesmen who are at an exceptionally high risk ofovert IHD but whose coronary atherosclerosisis still clinically silent. The test employed inthis study is safe but has a low yield and is,

probably, too expensive for widespread fieldsurveys. A more strenuous exercise test wouldbe diagnostically more productive but alsomore hazardous and, possibly, less acceptable.Simple step tests should be considered forclinical and epidemiologic studies. The elec-trocardiographic criteria used in this study arerigid, relatively insensitive but highly specific,as attested by the bad prognosis of anabnormal test. It is suggested that an abnor-mal electrocardiographic response to exerciseshould be considered a valid addition to thenosology of IHD. Individuals in this categorymay be particularly appropriate subjects fortesting the effectiveness of risk factor interven-tion programs.

AcknowledgmentThe assistance of Phillip Quickenton, M.P.H.,

Associate Biostatistician, New York State Departmentof Health, is gratefully acknowledged.

References1. MASTER AM: The two-step exercise electrocardio-

gram: A test for coronary insufficiency. AnnIntern Med 32: 842, 1950

2. SCHERF D, SCHAFFER AI: The electrocardio-graphic exercise test. Amer Heart J 43: 927,1952

3. MYERS GB, TALMERS F: The electrocardiograph-ic diagnosis of acute myocardial ischemia. AnnIntern Med 43: 361, 1955

4. SIMONSON E, KEYs A: The electrocardiographicexercise test: Changes in the scalar ECG andin the mean spatial QRS and T vectors in twotypes of exercise: Effect of absolute andrelative body weight and comment on normnalstandards. Amer Heart J 52: 83, 1956

5. LEPESCHKIN E, SuRAwicz B: Characteristics oftrue-positive and false-positive results of elec-trocardiographic Master two-step exercise tests.New Eng J Med 258: 511, 1958

6. HELLERSTEIN HK, PROZAN GB, LIEBOW IM, ETAL: Two step exercise test as a test of cardiacfunction in chronic rheumatic heart disease andin arteriosclerotic heart disease with oldmyocardial infarction. Amer J Cardiol 7: 234,1961

7. SIMONSON E: Use of the electrocardiogram inexercise tests. Amer Heart J 66: 552, 1963

8. ROBB GP, MARKS HH: Postexercise electrocar-diogram in arteriosclerotic heart disease. JAMA200: 918, 1967

9. BRUCE RA, HoRNsTEN TR: Exercise stress testingin evaluation of patients with ischemic heart

Circulation, Volume XLI, March 1970

552

by guest on May 29, 2018

http://circ.ahajournals.org/D

ownloaded from

ECG STRESS TEST

disease. Progr Cardiovasc Dis 11: 371,1969

10. HIL.LEBOE HE, JAMEs G, DoYIE JT: Cardiovas-cular Health Center: I. Project design forpublic health research. Amer J Public Health44: 851, 1954

11. ERICKsON L, SIMoNsoN E, TAYLOR HL: Theenergy cost of horizontal and grade walking onthe motor driven treadmill. Amer J Physiol145: 391, 1946

12. MATrINGLY T: The postexercise electrocardio-gram: Its value in the diagnosis and prognosisof coronary arterial disease. Amer J Cardiol 9:395, 1962

13. KiNCu SH, GrrrELsoHN AM, DoYLE JT:Application of a life table analysis in aprospective study of degenerative cardiovascu-

lar disease. J Chron Dis 17: 503, 196414. BRODY AJ: Master two-step exercise test in

clinically unselected patients. JAMA 171:1195, 1957

15. SIMONSON E: Differentiation Between Normaland Abnormal in Electrocardiography. St.Louis, C. V. Mosby Co., 1961, pp 189-246

16. ROSENFELD I, MASTER AM, ROSENFELD C:Recording the electrocardiogram during theperformance of the Master two-step test.Circulation 29: 204, 1964

17. DOYLE JT: Unpublished data.18. JENKINS CD, ROSENmAN RH, ZYzANsx SJ:

Cigarette smoking: Its relationship to coronaryheart disease and related risk factors in theWestern Collaborative Group Study. Circula-tion 38: 1140, 1968

Cwc,dation, Volume XU, March 1970

553

by guest on May 29, 2018

http://circ.ahajournals.org/D

ownloaded from

JOSEPH T. DOYLE and SANDRA H. KINCHThe Prognosis of an Abnormal Electrocardiographic Stress Test

Print ISSN: 0009-7322. Online ISSN: 1524-4539 Copyright © 1970 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.41.3.545

1970;41:545-553Circulation. 

http://circ.ahajournals.org/content/41/3/545located on the World Wide Web at:

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

  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. and Rights Question and Answer

Permissionsthe Web page under Services. Further information about this process is available in thewhich permission is being requested is located, click Request Permissions in the middle column ofClearance Center, not the Editorial Office. Once the online version of the published article for

can be obtained via RightsLink, a service of the CopyrightCirculationoriginally published in Requests for permissions to reproduce figures, tables, or portions of articlesPermissions:

by guest on May 29, 2018

http://circ.ahajournals.org/D

ownloaded from