Maximal exercise stress testing in normal and hyperlipidemic children

85

Atherosclerosis, 25 (1976) 85-94 @ Elsevier/North-Holland Biomedical

Press, Amsterdam

MAXIMAL EXERCISE STRESS TESTING HYPERLIPIDEMIC CHILDREN

- Printed in The Netherlands

IN NORMAL

FREDERICK W. JAMES, CHARLES J. GLUECK, RONALD FRANK MILLETT and SAMUEL KAPLAN

AND

W. FALLAT,

Department of Pediatrics (Cardiology), The Lipid Research Center of the Children’s Hospital Medical Center and the General Clinical Research Center, College of Medicine, University of Cincinnati, Cincinnati, Ohio (U.S.A.) (Received 27th January, 1976) (Revised received 12th April, 1976) (Accepted 12th April, 1976)

summary Electrocardiographic and cardiovascular responses during maximal exercise were evaluated in 103 normal children and in 82 children with familial hyperlipoproteinemia. The normal and hyperlipidemic children were comparable in regards to age, weight-height index, resting and exercise blood pressures, and maximal working capacity indices. The cohort of 82 hyperlipidemic children included 61 children (29 boys and 32 girls) with well defined “monogenic” familial hyperlipoproteinemia. Segmental ST depression on the exercise electrocardiogram occurred in 8 of these 29 boys (27.6%) as compared to 4 of 55 normal boys (7.3%), P < 0.025 and in 6 of the 32 girls (19%) as compared to 7 of 48 normal girls (14.6%), P > 0.1. Segmental ST depression was present in 14 of 61 (23%) children with “monogenic” hyperlipoproteinemia, as compared to 11 of 103 (10.75%) normals (x2 = 4.47, P< 0.05). An assessment of the clinical significance of an abnormal exercise electrocardiogram in male children with “monogenic” hyperlipoproteinemia must await the following: (1) two to four decades of observation and study of the development of morbid or mortal coronary disease, or (2) the future development of improved invasive or noninvasive techniques for the early detection of covert coronary occlusive disease. Currently, maximal exercise electrocardiography cannot be contemplated as a useful indicator of eventual premature coronary artery disease in asymptomatic hyperlipidemic children. Supported in part by the American Heart Association, Southwestern Ohio Chapter Ad the General Clinical Research Center Grant No. RR-00068-13. A portion of this work was done during Dr. Glueck’s tenure (1971-1976) as an Established Investigator. American Heart Association. Address for reprints: Frederick W. James. M.D., Division of Pediatric Cardiology, Children’s Hospital Medcal Center. Cincinnati, Ohio 45229, U.S.A.

86

Key words:

Children - Familial hypercholesterolemia - ischemic heart disease - Maximal exercise electrocardiogmphy -Premature atherosclerosis

Introduction Certain hyperlipoproteinemias are closely associated with the development of premature ischemic heart disease (IHD) [l-3]. Adults, heterozygous for of IHD befamilial hypercholesterolemia (FHC) may develop manifestations fore age 30, but most commonly before age 50 to 60, and 50 to 70% of these events are lethal [l-3]. FHC occurs in l/200 to l/600 unselected American infants and is highly penetrant in the progeny of affected kindreds [4-71. Children heterozygous for FHC do not have clinical signs of IHD [ 81, whereas, those with homozygous FHC have angina, myocardial infarction (usually before age 20) and aortic ring atheromata [ 31. In autopsy studies in children and young adult males, there is suggestive evidence that atherosclerotic cardiovascular disease of the adult population may have its genesis during childhood [g-11]. Exercise may induce electrocardiographic signs of myocardial ischemia in children with certain congenital heart diseases [ 12-141. Since early demonstration of myocardial ischemia is of clinical significance in certain hyperlipoproteinemias, we analyzed the exercise electrocardiogram to determine the prevalence of abnormal segmental ST depression in asymptomatic hyperlipidemic and normal children. Materials and Methods Eighty-two children with well characterized familial hyperlipoproteinemia and 103 children with normal fasting cholesterol and triglyceride levels were ascertained by the Lipid Research Center during various lipid and lipoprotein sampling studies [ 57,151. At least 2 fasting blood samples were obtained while weights were stable on a habitual diet intake, and diagnoses of hyperlipidemia were made following well characterized analytical methods [ 21. After measurement of plasma lipids, we then established the following criteria for 2 study groups (normal and hyperlipidemic children), in which all subjects had a normal resting physical examination and electrocardiogram. The normal children also had normal fasting levels of cholesterol and triglyceride and normal resting supine blood pressures as compared to values reported by Londe [16], Masland et al. [17] and Heyden et al. [18]. Using these criteria, 103 subjects, 55 boys and 48 girls between the ages of 5 and 21 years were classified as normal. The average age for both males and females was 13. All hyperlipidemic children had primary elevation of cholesterol or triglyceride or both in 2 blood samples. To study as nearly a homogeneous group of children as possible, we established the following additional requirements for entry into the lipid study group: (a) Exclusion of secondary hyperlipidemia [2,4,8]. (b) Documentation of primary hyperlipidemia in at least one additional first degree relative of the index child. (c) When possible, documentation of “monogenic” familial hypercholesterolemia [7,8] by family studies which revealed either 3 generation vertical transmission or presence of tendon xan-

87

thomas or both (55 children, 23 kindreds). (d) Documentation of “monogenic” familial combined hyperlipidemia [ 31 by 3 generation vertical transmission of mixed II-A, II-B and IV lipoprotein phenotypes (4 children, 4 kindreds). (e) Documentation of “monogenic” familial hypertriglyceridemia [ 31 by 3 generation vertical transmission of Type IV lipoprotein phenotypes and endogenous hypertriglyceridemia (2 children, 2 kindreds). Using these criteria, 82 children, 42 boys and 40 girls, between the ages of 5 and 21 years, had primary hyperlipoproteinemia (Table 1). The average age of the hyperlipidemic males and females was 13. These 82 children came from 56 separate kindreds. Exercise procedure The 103 normolipidemic children and the 82 hyperlipidemic children underwent a continuous graded maximal bicycle exercise test. The exercise test was performed on a Quinton Instrument Ergometer (Model 844). Each subject was tested in 1 of 3 exercise programs which were based on body surface area (BSA). The limits of BSA for each program were as follows:
levels for the children

with

1

50

18

50

42

29

18

37

17

237

220

245

134

224

115

80

62

80

58

275

148

275

164

50th

AND

133

86

110

85

288

172

320

177

75th

TRIGLYCERIDE

170

120

166

102

336

198

363

199

90th

IN NORMAL

9 5th

208

128

188

123

358

204

389

360

137

254

132

363

214

420

213

99th

HYPERLIPIDEMIC

211

AND

42

55

40

48

36 a

55

38’

48

n

CHILDREN

99f

68+-

89+-

62+-

274 +

156*

291 c

167+

F+

10d

5

gd

5

7’

4

9 c

4

SE

aExcludes 2 females with primary hyperlipidemia who had normal cholesterol (160, 170 mg/lOO ml) but elevated triglyceride levels (254, 220 mg/lOO ml). bExcludes 6 mates with primary hyperlipidemia who had normal cholesterol (168. 178, 180, 200. 209. 215 mg/lOO ml) but elevated triglyceride levels (146.162. 164.170, 178, 248 mg/lOO ml). CP < 0.001, cholesterol levels in hyperlipidemic females vs. normal females. hyperlipidemic males. vs. normal males. dP < 0.02. triglyceride levels in hyperlipidemic females vs. normal females, hyperlipidemic males vs. normal males. a = Mean; +SE = standard error.

Normal males Hyperlipidemic males

Triglyceride Normal females Hyperlipidemic females

Normal males Hyperlipidemic males e

145

124

6th

25th

OF CHOLESTEROL

Percentile

DISTRIBUTION

Cholesterol Normal females Hyperlipidemic females

PERCENTILE

TABLE

89

primary hyperlipidemia and for normals are summarized in Table 1. Of the 82 children with primary hyperlipidemia, 74 had predominant hypercholesterolemia, with essentially no overlap in the disparate distributions of cholesterol levels in affected and normal children. Since 8 of 82 children with primary hyperlipidemia had predominant hypertriglyceridemia, the distribution of triglycerides for normal and hyperlipidemic children has considerable overlap, although there were notable distinctions in group means which were higher for P < 0.02 (by Student’s t-test). Fiftyboth boys and girls with hyperlipidemia, five of the 82 hyperlipidemic children came from kindreds thought to have “monogenic” familial hypercholesterolemia, 4 from kindreds with familial combined hyperlipidemia, and 2 from kindreds with familial hypertriglyceridemia. Age, weight-height index, blood pressure and maximal working capacity The normal and hyperlipidemic groups were further subdivided by sex (Table 2). Comparing the males to males and females to females from both groups, there were no significant differences in the mean ages and weightheight indices. Also, the resting blood pressures and maximal systolic blood pressures were not different in either group except for the maximal diastolic blood pressure being significantly higher in the hyperlipidemic males than in the normal males. There were no significant differences in the mean exercise times and maximal

TABLE 2 THE MEAN AGES, WEIGHT-HEIGHT PERLIPIDEMIC CHILDREN

INDICES,

AND BLOOD

PRESSURES

IN NORMAL

AND HY-

_

N

Age

W-HI

RSBP

RDBP

MSBP

MDBP

77 16.1 2.2 83.2 a 12.4 1.9

(Yrs)

M&S Normal

55

HyperIipidemic

42

8 +SD +SE 8 +SD +SE

13.1 4 0.5 13.1 4 0.6

30.9 8.1 1.1 30.3 8.1 1.3

112.8 11.9 1.6 116.1 14.6 2.2

71.2 10.4 1.4 73.5 9.5 1.5

167.7 33.6

8 +SD *SE a *SD *SE

13 4.2 0.6 13.2 4 0.6

30.2 11.3 1.7 31.9 11.6 1.8

112.6 11.9 1.7 117.3 13.4 2.1

72.2 10.1 1.5 73.4 10 1.6

153.2 27.2 4 162.2

4.5 174.8 33.1 5.1

Femde.S

Normal

48

Hyperlipidemic

40

25.2 4

N = number of children W-HI = weight-height index RSBP = resting syst.oIic blood pressure RDBP = resting diastolic blood pressure MSBP = systolic blood pressure at maximal exercise MDBP = diastolic blood pressure at maximal exercise SD = standard deviation, SE = standard error. x = mean. a The MDBP is higher in the hyperlipidemic males than in the normal maks, P < 0.05. Other comparisons within the sex groups were not significantly different, P > 0.1.

82.8 9.4 1.4 83.8 11.1 1.8

90

TABLE

3

THE MEAN

EXERCISE

HYPERLIPIDEMIC

TIMES

AND

MAXIMAL

WORKING

CAPACITY

INDICES

IN NORMAL

AND

CHILDREN N

ET

MWC/M2

MWC/kg

x +SE s tSE

12.2 0.7 11.15 0.6

5146 414 4316.2 334

160 13.5 128.4 6.4

x +SE K *SE

9.02 0.49 9.2 0.5

M&S

Normal

55

Hyperlipidemic

42

Fe?tlde.S

Normal

46

Hyperlipidemic

40

3240 211 3111 204

99.9 1.4 97.3 6

Within the sex groups compared variables were not significantly different, P > 0.1. N = number, ET = exercise time, MWC/m2 or /kg = maximal working capacity per body surface area (m2) or body weight (kg).

working capacity indices (MW&) in the comparison of males to males and females to females from each group (Table 3). However, the mean MWCI were higher in the males than in the females. Exercise electrocardiograms Normal children ST depression was not present in any child before exercise. During exercise, ST depression > 1 mm was recorded in 4 boys and 7 girls (Table 4). No arrhythmias were recorded. TABLE

4

A COMPARISON NORMAL

OF SEGMENTAL

AND HYPERLIPIDEMIC ST Depr

ST DEPRESSION

> 1 mm DURING

No ST Depr

x2

Males

Hyperlipidemic Normal

MAXIMAL

EXERCISE

IN

CHILDREN ST Depr

No ST Depr

x2

Females

9 4

33 51

4.12a

6 7

34 41

0.003NS

8 4

21 51

6.4 b

6 7

26 41

0.25NS

“Monogenic” hyperlipidemia Normal

Males and

females

Hyperlipidemic Normal

15 11

61 92

2.2NS

“Monogenic” hyperlipidemia Normals

14 11

41 92

4.47 a

“Monogenlc” = hyperlipoproteinemia with vertical transmission in 3 generations, or in 2 generations in kindreds with familial hypercholesterolemia and tendon xanthomas in adult subject. Of 82 children with primary hyperlipidemia, 61 were shown to have “monogenic” hyperlipidemia. Depr = depression, NS = not significant, x2 = Chi-square. a P < 0.05. b P < 0.025.

91

Hyperlipidemic children In order to focus on those children with the best defined, and generally quantitative most severe primary hyperlipidemia, studies of the frequency of ST depression were made in the 61 children with well defined “monogenic” familial hyperlipoproteinemia (Table 4). Segmental ST depression was present in 8 of 29 hyperlipidemic boys (27.6%), which was 3 times more frequent than in normal boys, 4 of 55 (7.3%), P < 0.025 by CHI square analysis (Table 4). Segmental ST depression was present in 6 of 32 (19%) girls with “monogenic” familial hyperlipoproteinemia, but this was not more frequent than in normal girls (7/ 48, 14.6%), P > 0.1. When all children with monogenic hyperlipoproteinemia were considered together, and compared to all normal children, ST depression was more common in hyperlipidemic children (Table 4). Discussion Our study revealed that abnormal ST depression was more prevalent in boys with “monogenic” hyperlipoproteinemia than in normolipemic boys. At the present time, we are unable to interpret the significance of our findings of abnormal ST depression in 7.3% of the normal males and 14.6% of normal females. There is, however, a parallel finding of a high incidence of “false positives” in normal adult females [26-281 as compared to normal adult males. The results of the exercise electrocardiogram are reproducible, since each of 20 hyperlipidemic and 31 normolipemic children has demonstrated identical ST segment changes on both, the initial and follow up, exercise tests. A preliminary analysis of post-exercise systolic time intervals in these same children has revealed that the duration of the pre-ejection period and the ratio of pre-ejection time:left ventricular ejection time were significantly increased in the hyperlipidemic boys as compared to the normolipidemic boys, P < 0.05. [29] In the girls, the post-exercise systolic time intervals were similar in both patient and control groups. In the presence of a relatively large group of presumably false positive exercise tests in normal children it is difficult to make firm conclusions regarding the significantly increased incidence of abnormal tests in hyperlipidemic boys. We could not contemplate coronary arteriography to diagnose coronary occlusive disease in asymptomatic children who have an abnormal maximal exercise test and familial hyperlipoproteinemia, particularly since similar studies would also have to be done in normal controls for comparison. There are then no immediate prospects to obtain direct objective visualization of the coronary arteries in these children. Follow-up studies to provide evidence to link the ECG changes with the clinical development of morbid or mortal premature coronary disease will require at least 2-5 decades in this group of children. The significance of an abnormal exercise electrocardiogram in these patients will then have to await further clinical follow-up studies accompanied hopefully in the future by sophisticated non-invasive or improved invasive techniques for detecting and measuring coronary artery occlusion. Several investigative groups have analyzed the exercise electrocardiogram in asymptomatic children between the ages of 5 and 21 years. Goldberg [30] recorded segmental ST depression of l-2 mm during exercise in 5% of normal

children, using single lead (CR6) electrocardiography. Bengtsson [ 311 analyzed the exercise electrocardiogram using multiple leads in 99 asymptomatic children between the ages of 5 and 21 years. During tachycardia, the ST segment position was measured at 50 msec or less after the J point. The exact percentage of patients with ST depression at maximal exercise is difficult to determine from this series. However, ST depression of l-l.5 mm was recorded in a few subjects during and after exercise. SjSstrand [ 321 studied the relationship between the ST segment level on the electrocardiogram and heart rate in 70 adult men and women during the administration of pharmacological agents and exercise. The ST segment level was measured in relationship to the p-q level at 20 msec after the J point. ST depression of 1 mm was recorded in more than one electrocardiographic lead during exercise. This author suggested that the shift in the ST level was due to a slow change of potential after the T wave which produced an elevation in the p-q level as diastole was shortened. As a result, the ST segment appears to be depressed. Our study differs from the above series in that we recorded the exercise electrocardiogram using multiple conventional and the Frank orthogonal lead systems. Our method of analyzing the ST segment and the interpretation of a positive change on the exercise electrocardiogram were different. Also, our data was segregated according to lipid levels and sex. Because of these differences, it is difficult to compare our study with the findings of Goldberg [ 301, Bengtsson [31] or Sjiistrand [32]. Rogers et al. [33] analyzed the E-ECG for J point displacement and ST segment slope in 73 asymptomatic adolescent males. J point or segmental ST depression alone or in combination was not recorded in any of their patients. The authors suggested that any J point depression on the E-ECG should be considered abnormal in the adolescent male. J point depression < 1.5 mm with an upward sloping ST segment did occur in many of our patients. However, J point depression combined with depression of the ST segment for 60 msec or more below the isoelectric line was more than we expected from tachycardia alone and was considered a positive finding in these children. In adults, segmental ST depression on the E-ECG has been observed in patients with hypertension, obesity and abnormal blood lipids [21]. Moreover, several studies [34-361 have suggested that an abnormal maximal treadmill test (MTT) response in adults has significant predictive value for subsequent development of clinical coronary disease and might be considered as an independent risk factor. Froelicher et al. [34] observed that adults with ischemic ST segments provoked by dynamic exercise have an increased risk of developing subsequent covert coronary heart disease [34]. Bruce et al. [35] reported that clinical coronary artery disease developed over a 5-year interval in 3 of 22 normal subjects (13.6%) who initially had an abnormal MTT. In 199 normal subjects who initially had a normal MTT, 2 (1%) developed clinical signs of coronary artery disease over the same 5 year interval. Aronow [36] reported the development of clinical coronary artery disease over a 30 months interval in 3 of 13 normal subjects with an initially abnormal MTI’. In 87 normal subjects with an initially normal MTT, 1 (1.1%) developed clinical coronary artery disease during the same 30 month interval. Carlson et al. [37]’ observed an increased frequency of abnormal ST depres-

93

sion in asymptomatic adults with primary hyperlipidemia and suggested that the increased frequency of the ST depression was positively related to the concentration of both low density and very low density lipoproteins. However, the limitations of exercise electrocardiography in hyperlipidemic adults have recently been demonstrated by Borer et al. [ 381. In hypercholesterolemic subjects who had an abnormal exercise test in the absence of other indications of coronary artery disease, only 37% had greater than 50% stenosis, 30% had less than 5096, and 33% had normal coronary arteries. Borer et al. concluded that “ . . . the diagnostic usefulness of exercise electrocardiography is limited. False-negative responses are frequent in patients with clinically suspected coronary disease, and false-positive responses frequent in asymptomatic patients” [ 381. Hence, even in adults where occlusive coronary artery disease is presumably well developed, the significance of abnormal electrocardiographic response is difficult to evaluate. Any attempt at this time to interpret the abnormal maximal exercise ECG as predictive for premature coronary artery disease in high risk children with “monogenic” hyperlipidemia would be purely speculative. Maximal exercise electrocardiography in children with familial hypercholesterolemia can be viewed as an experimental attempt to provide early information relative to coronary artery disease which can only be validated by long term follow-up for several decades or the development of improved invasive or non-invasive methods for providing objective evidence of occlusive coronary artery disease. Hence, conventional maximal exercise electrocardiography in asymptomatic children with “monogenic” familial hypercholesterolemia cannot then be contemplated as useful in the routine clinical evaluation of this disorder which is associated with a high risk of premature coronary artery disease. Acknowledgements We thank Mrs. Vera Naylor, Miss Mary Jo Sandker and Mrs. Anne Doerner for their technical assistance; and Mrs. Margaret DeHo for typing the manuscript. References 1 Stone. N.J.. Levy, RI.. Fredrickson, D.S. and Verter. J., Coronary artery disease in 116 kindred with familial type II hyperlipoproteinemia, Circulation, 49 (1974) 476. 2 Fredrickson. D.S. and Levy. RI.. Familial hyperlipoproteinemia. In: J.B. Stanbury, J.B. Wyngaarden and D.S. Fredrickson (Eds.). The Metabolic Basis of Inherited Disease, McGraw-Hill, New York, N.Y., 1972, p. 545. 3 Goldstein, J.L.. Schrott, H.R.. Hazzard, W.R.. Bierman, E.L. and Motulsky. A.G.. Hyperlipidemia in coronary heart disease, Part 2 (Genetic analysis of lipid levels in 176 families and delineation of a new inherited disorder, combined hyperlipidemia). J. Clin. Invest., 52 (1973) 1544. 4 Glueck. C.J.. Fallat. R.W. and Tsang. R.C., Hypercholesterolemia and hypertrlglyceridemia in children - Pediatric approach to primary atherosclerosis prevention, Amer. J. Dis. Child.. 128 (1974) 569. 5 Tsang. R.C.. Fallat. R.W. and Glueck. C.J., Cholesterol at birth and age, Part 1 (Comparison of normal and hypercholesterolemic neonates), Pediatrics. 53 (1974) 458. 6 Goldstein, J.L., Albers, J.J.. Schrott, H.G.. Hazzard, W.R., Bierman, E.L. and Motulsky. A.G., Plasma lipids levels and coronary heart disease in adult relatives of newborns with normal and elevated cord blood lipids, Amer. J. Hum. Genet., 26 (1974) 727. 7 Tsang. R.C., Glueck. C.J., Fallat. R.W. and Mellies. M.. Neonatal familial hypercholesterolemia, Amer. J. Dis. Child., 129 (1975) 83. 8 Kwiterovich. P.O.. Levy, R.I. and Fredrickson, D.S.. Neonatal diagnosis of familial type II hyperlipoproteinemia, Lancet, 1 (1973) 118.

94

9 Strong, J.P. (1969) 251.

and McGill.

Jr.,

H.C..

The pediatric

aspects

of atherosclerosis,

9

Pesonen. E., Norio. R. and Sama. S.. Thickenings in the coronary of genetic factors in coronary heart disease, Circulation, 51 (1975)

11

McNamara. J.J.. Molot. M.A.. Stremple, J.F. and Cutting, A.T., Coronary artery disease in combat casualties in Vietnam, J. Amer. Med. Ass., 216 (1971) 1185. Hzdioren, K.H., The telemetered exercise electrocardiogram in congenital aortic stenosis, Pediatrics, 47 (1971) 31. James, F.W., Effects of physical stress on adolescents with normal or abnormal cardiovascular function, Postgrad. Med., 56 (1974) 53. James, F.W. and Ka~ian. S., ST depression and systolic hypertension during exercise in aortic insufficiency, Circulation. 50 (1974) U-27.

13 14 15 16 17 18

19 20 21

22 23 24 25 26 27

28 29 30 31 32 33 34

35 36 37 38

in infancy

Res..

10

12

arteries 218.

J. Atheroscler.

as an indication

Glueck. C.J., FaIIat. R.W., Tsar&. R.C. and Buncher. CR., Hyperhpidemia in progeny of parents with myocardiai infarction before age 50, Am. J. Dis. Child.. 127 (1974) 70. Londe, S., Blood pressure in children as determined under office conditions. Clin. Pediatrics, 5 (1975) 71. MasIand. Jr., R.P., HeaId. F.P., Goodaie, W.T. and Gallagher, J.R.. Hypertensive vascular disease in adolescence, New EngI. J. Med., 255 (1956) 894. Heyden. S., Bartel, A.G.. Hames. C.G. and McDonough. J.R., Elevated blood pressure levels in adolescents. Evans County, Georgia - Seven-year follow-up of 30 patients and 30 controls, J. Amer. Med. Ass., 209 (1969) 1683 Blomqvist, C.G.. Use of exercise testing for diagnostic and functional evaluation of patients with arteriosclerotic heart disease, Circulation 44 (1971) 1120. Sive, P.H.. Medaiie. J.H., Kahn, H.A.. Neufeld, H.N. and Riu, E., Correlation of weight--height index with diastolic and with systolic blood pressure, Brit. J. Prev. Sot. Med., 24 (1970) 201. Blackburn, H., The exercise electrocardiogram, technological procedural and conceptual developments. In: H. Blackbum (Ed.), Measurement in Exercise Electrocardiogram, Charles C. Thomas, Springfield, Iii.. 1969, p. 220. Linhart, J.W. and Turnoff, H.B., Maximum treadmill exercise test in patients with abnormal control electrocardiogram, Circulation. 49 (1974) 667. Kattus. A.A.. Exercise electrocardiography -Recognition of the ischemic response, false positive and negative patterns. Amer. J. Cardiol., 33 (1974) 721. Case. R.B.. Nasser, M.G. and Crampton. R.S., Biochemical aspects of early myocardiai &hernia. Amer. J. Cardiol.. 24 (1969) 766. Stuart, Jr., R.J. and Eiiestad, M.H., Upsloping S-T segments in stress testing-a six-year follow-up Correlation in 248 angiograms (Abstr.), Circulation, 50 (1974) 111-8. Astrand, I., Exercise electrocardiograms recorded twice with an 8-year interval in a group of 204 women and men 4843 years old, Acta Med. Scand., 178 (1965) 27. Lepeschkin. E., Physiological factors influencing the electrocardiographic response to exercise. In: H. Blackburn (Ed.), Measurement in Exercise Electrocardiography, Charles C. Thomas, Springfield, III., 1969, p. 363. Gumming. G.R.. Dufresne, C.. Kich, L. and Samm. J.. Exercise electrocardiogram patterns in normal women, Brit. Heart J.. 35 (1973) 1055. James, F.W.. Glueck. C.J.. Kaplan, S. and Naylor. V., Exercise electrocardiogram and postexercise systolic time intervals in hyperiipidemic children. Ciin. Res.. 23 (1975) 471A. Goldberg, S.J.. Exercise in cardiac patients. In: A.J. Moss and F.H. Adams (Eds.). Heart Disease in Infants, Children. and Adolescents, Wiiams and Wilkins. Baltimore, Md., 1968, p. 270. Bengtsson. E.. The exercise electrocardiogram in healthy children and in comparison with adults. Acta Med. Scand., 154 (1956) 225. Sjiistrand. T.. The relationship between the heart frequency and the ST level of the e1ectrocaxliogra.m. Acta Med. Scand.. 138 (1950) 201. Rogers, J.H., Strong, W.B. and HeRerstein. H.K.. The exercise electrocardiogram in trained and untrained adolescent males (Abstr.), Pediatric Res.. 9 (1975) 270. Froehcher. Jr., V.F.. Thomas, M.N., Pillow. C. and Lancaster, M.C., Epidemiologic study of asymptomatic men screened by maximal treadmill testing for latent coronary artery disease, Amer. J. Cardiol.. 34 (1974) 770. Bruce, R.A. and McDonough. J.R.. Stress testing in screening for cardiovascular disease, Bull. N.Y. Acad. Med., 45 (1969) 1288. Aronow, W.S., Thirty-month follow-up of maximal treadmill stress test and double master’s test in normal subjects. Circulation, 47 (1973) 287. Carlson. L.A.. Ekelund, L.G. and Olsson, A.G.. Frequency of ischemic exercise ECG changes in symptom-free men with various forms of primary hyperlipidemia, Lancet, 2 (1975) 1. Borer. J.S., Brensike, J.F., Redwood, D.R.. Itscoitz, S.D.. Passman, E.R., Stone, N.J.. Richardson, J.M.. Levy. R.I. and Epstein, S.E., ECG response to exercise in predicting coronary-artery disease, New En%. J. Med., 293 (1975) 367.