High-density lipoprotein cholesterol and angiographic coronary artery disease in black patients

Sabbah

et al.

Regional left ventricular function in patients with segmental early relaxation and normal coronary arteries. J Am Co11 Cardiol4:45, 1984. 9. Gelberg HJ, Brundage BH, Glantz S, Parmley WW: Quantitative left ventricular wall motion analysis: A comparison of area, chord and radial methods. Circulation 59:991, 1979. 10. Miller RJ: Simultaneous statistical inference. 2nd ed. New York, 1981, Springer-Verlag, pp. 37. 11. Gaasch WH, Blaustein AS, Bing OHL: Asynchronous (segmental early) relaxation of the left ventricle. J Am Co11 Cardiol 5:891, 1985. 12. Sabbah, HN, Stein PD: Early segmental thinning of the left ventricular wall following regional &hernia. Cathet Cardiovast Diagn 9:473, 1983. 13. Theroux P, Ross J Jr, Franklin D, Kemper WS, Sasayama S: Regional myocardial function in the conscious dog during

High-density angiographic patients

14.

15. 16. 17.

acute coronary occlusion and responses to morphine, propranolol, nitroglycerine and lidocaine. Circulation 53:302, 1976. Sabbah HN, Khaja F, Brymer JF, Stein PD: Temporal pattern of regional left ventricular wall motion in patients with segemental early relaxation. Cathet Cardiovasc Diagn 10:629, 1984. Sigwart U, Grbic M, Essinger A, Fischer A, Morin D, Sadeghi H: Myocardial function in man during acute balloon occlusion (abst). Circulation 66:11, 1982. Wiegner AW, Allen GJ, Bing OHL: Weak and strong myocardium in series: Implications for segmental dysfunction. Am J Physiol 235:H776, 1978. Ilebekk A, Lekven J, Ku1 F: Left ventricular asynergy during intracoronary isoproterenol infusion in dogs. Am J Physiol 239:H594, 1980.

lipoprotein cholesterol and coronary artery disease in black

The clinical epidemiology of coronary artery disease (CAD) among black patients has not been well described, particularly in relation to the impact of the major risk factors. Lipoprotein profiles and other risk factors were measured in 114 black patients undergoing cardiac catheterization for probable CAD. Patients (coronary narrowing of 50% or greater; n = 63) were compared to those without significant stenoses (n = 51). Total cholesterol and low-density lipoprotein cholesterol (LDLC) were both significantly associated with the presence of CAD in men and women; however, high-density lipoprotein cholesterol (HDLC) had discriminatory value only for women. The ratio of total cholesterol to HDLC (TC:HDL) separated patients from control subjects in both sexes, most efficiently among women, and was the only lipid variable associated with CAD in the age group over 55 years. Hypertension and angina were frequent in both groups and did not identify those with disease. In summary, LDLC was the best predictor among the lipids for men (likelihood ratio = 9.4) and TC:HDL was the best indicator of disease among women (likelihood ratio = 15.7). Low HDLC levels may, in part, account for the increased incidence of CAD among black women. Further population studies of factors leading to reduced HDLC, namely, obesity and diabetes, are needed. (AM HEART J 110: 1006, 1965.)

Richard Cooper, M.D., Christopher Sempos, Ph.D., Jalal Ghali, M.D., and Jack Ferlinz, M.D. Chicago, Ill., a& Washington, DC.

Relatively few studies of coronary artery disease (CAD) have included sufficient numbers of black subjects to allow meaningful analysis of the relationFrom the Division of Adult Cardiology and the Section of Clinical Epidemiology Cook County Hospital, and the Clinical Nutrition Branch, Division of Nutrition, Food and Drug Administration. Received accepted

for publication June 3, 1985.

Richard Cooper, M.D., tal, 1835 West Harrison

1006

Jan.

28, 1985; revision

Division of Adult Cardiology, St., Chicago, IL 60612.

received

May

Cook County

2, 1985; Hospi-

ship between risk factors and clinical disease.’ The limited data available on black patients with CAD have made it difhcult to resolve a number of important questions. Although it is clear that black women suffer high rates of CAD relative to white women, the data on incidence and death from CAD among black men are sufficiently inconsistent and sparse to make it difhcult to reach clear-cut conclusions.1-5 The specific causes of the high rate of CAD among black women also require detailed study,

Volume 110

Lipoproteins

Number 5

both as the basis for effective treatment and prevention among this group and as a source of new knowledge about the pathophysiology of atherosclerosis. Although data from Evans county suggested that the black population is protected against CAD by higher levels of high-density lipoprotein cholesterol (HDLC), it is not clear whether these findings apply to northern, urbanized, or middle-class black populations!, 7 The higher prevalence of hypertension and left ventricular hypertrophy may alter the impact of other risk factors, not only in terms of the occurrence of disease but for the prognosis of symptomatic patients. In order to address some of these questions, the Division of Adult Cardiology at Cook County Hospital established a Heart Disease Registry in the fall of 1983. The purpose of this registry was twofold, first to collect cross-sectional data on patients with CAD as a basis for descriptive and case-control analyses, and second to provide the basis for prospective studies of symptomatic patients to learn more about survival rates and factors which determine prognosis. This report examines cross-sectional data for an 11-month period from the catheterization laboratory. METHODS Patient selection. Cook County Hospital is a 1300 bed municipal institution primarily serving the poor and working class black population of Chicago. Over 11,000 patients/year are admitted to the medical wards and 350,000 adults are seen as outpatients. Overall, approximately 75% of the patients are black, with Hispanics making up the next largest group, and additional small numbers of white patients and recent immigrants from Poland, India, or the Far East. From October 17, 1983, to September 17, 1984, a total of 234 patients underwent catheterization for the evaluation of possible CAD. Of these patients, complete data were available for 159 and of that group 45 were not black. The analysis presented here is thus based on 114 black patients accrued over an 11-month period. Of the patients omitted because of missing data, the majority were excluded because a complete lipoprotein profile was not determined. Eight patients were excluded because of an incomplete catheterization or unavailability of historic data. Cases were defined as patients with narrowing by at least 50% of one of the following segments: (1) left main, (2) either the left anterior descending, the circumflex or the right coronary arteries, or (3) a proximal segment of a major branch of the three major coronary vessels. Noncases included all patients with normal coronary arteries or stenoses of less than 50%. Vessel narrowing was estimated by consensus of a panel of cardiologists during the clinical conference. Data collection. Historic data were collected by the

Table

and coronary

disease

in blacks

1007

I. Characteristics of the sample

Men Variable

CAD

Non-CAD

/~“onlc~n

(‘Al)

Lyon-(‘AD

Abbreviations: CAD = coronary artery disease; BMI = lmdy mass index (based on data from the National Health and Nutritwn Examination Survey) BMI for black men, ages 45 to 64 years, was 26.G and for women, 29.5. These data were provided by Dr. Christopher Sempw. All values are mean i standard deviation.

*p < 0.05. tp < 0.01.

cardiology fellow at the time of the precatheterization workup on a preceded questionnaire. Laboratory specimens for lipids and lipoproteins were drawn at the beginning of the catheterization from the venous catheter. Specimens were transported to the laboratory and processed on the same day. Except in emergency situations patients had fasted for 12 to 14 hours when blood was sampled. Assays were performed in serum. Total cholesterol was measured by an adaptation of the enzymatic method* and HDLC was determined in the supernate after lower density lipoproteins were precipitated with Mg++ dextran sulfate, as proposed by Warnick et al.!’ Based on a sample of 16 duplicates submitted blindly to the laboratory over the course of the study the following estimates for intrabatch technical error were obtained: total cholesterol (TC) = 5.7%) HDLC = 1.9, and triglycerides = 2.1. (Technical error was calculated by the formula CCI~ and the resulting value divided by the sample d- 2n mean and multiplied times 100 to obtain a percentage.) Interbatch reproducibility, as estimated by the coefficient of variation of repeat determinations on lypholized specimens, was 4.7% for TC and 2.3% for HDLC. Low-density lipoprotein cholesterol (LDLC) was calculated as TC minus HDLC and triglycerides/5.1° Two values for triglycerides were greater than 300 mg/dl, where this indirect estimate is unreliable, and they were excluded. In the clinical setting it is difficult to standardize precisely the sampling protocol. Blood pressure measurement was particularly vexing. In patients with evolving and sometimes unstable medical conditions, many of whom are receiving antihypertensive treatment, a basal state is impossible to define. For purposes of this study the first recorded blood pressure taken by the auscultatory method in the antecubital fossa at the time of the catheterization was used. The history of hypertension was regarded as the primary variable reflecting hypertensive

November,

1008 Table

Cooper et al.

American

II. Medical history and medication use Men

Table

Ill. Lipids in univariate analysis: Cases vs non-

cases

Women

Men Variable

CAD

Non-CAD

CAD

Abbreviations: trophy. *p < 0.01. tp < 0.05.

65 TO 71 17 74

13* 80 7 64

25 93 30 90

78

25 58 88

50 47 86

63 57 71

39 57 53

40 70 17

36 64 57*

29 75 12

51 23

MI = myocardial

infarction;

LVH

Women

Non-CAD Lipid

Medical history MI Hypertension Diabetes Angina Medications Diuretics Beta blockers Smoking ECG Q wave ST-T change LVH

1985

Heart Journal

= left ventricular

17 78

(mgldl)

TC LDLC HDLC Triglycerides Ratio TC:

19

HDLC

Non-CAD

CAD 204 148 35 110 6.3

+ + * & t

45 44 10 55 2.3

176 114 39 123 4.8

Abbreviations: TC = total cholesterol; cholesterok Values are mean + standard lipoprotein cholesterol. *p < 0.05.

6t

k f * ? *

CAD 40* 35t 10 76 1.4t

231 166 37 136 6.6

Nan-CAD

f 41 + 35 f

205 ?44* 141 k 40t

10

45 + 12t

_t 73 t 1.9

103 f 64 4.8 + 1.3t

LDLC = low-density lipoprotein deviation. HDLC = high-density

tp < 0.01.

hyper-

status. Angina was assessed with the London School of Hygiene Cardiovascular Questionnaire. As in all casecontrol studies the impact of the disease process itself on the risk factors being studied must be taken into account. Thus 16 patients had sustained a myocardial infarction in recent months and other patients received some dietary or pharmacologic treatment for hypercholesterolemia. In general, these factors would reduce lipid values preferentially among cases, relative to control subjects, since they were on the average a sicker group of patients. As in most case-control studies this would tend to create a false negative bias.“, ‘* Statistical analysis. Data were collected onto preceded computer forms, entered into a data file, edited, and analyzed with programs available on SAS.13In two group comparisons a two-tailed t test was used. Logistic regression models and likelihood ratios were computed.14 RESULTS

Descriptive statistics by sex and disease category are presented in Table I. Although more women than men underwent catheterization, they were much less likely to have significant coronary narrowing. The mean age of men was similar in cases and control subjects, but women with CAD were modestly yet significantly older. Women in both groups were more obese than the men and there were no important differences between the two groups. Blood pressure was lower in cases than in control subjects in both sexes, demonstrating most likely the effect of myocardial damage from prior infarction among the cases. Nutrient intake was not measured. Data on medical history and current medications demonstrate that cases were most likely to have any

one of several manifestations of cardiovascular disease, yet were taking similar medications (Table II). A history of hypertension was present in approximately 80% of patients in all four groups. Women with CAD were more likely to have been smokers, although this difference was not significant. Among the ECG findings Q waves were more common among cases while left ventricular hypertrophy was observed more frequently in control subjects. In univariate analysis a consistent lipoprotein pattern emerged (Table III). TC and LDLC contained important discriminatory information for both sexes. It should be noted, however, that the mean levels among men without CAD are lower than would be anticipated in the population at large, thus accentuating the case-control difference. Among men, although HDLC was lower among cases, the difference was modest (4.2 mg/dl) and not statistically significant. (The fact that this difference would be biologically significant on a population basis provides evidence of the potential beta, type II, error in this relatively small clinical series.) Although women with CAD had higher HDLC levels, in cases of both sexes they were virtually the same. The ratio of TC: HDLC separated cases and noncases well for both men and women, and the ratios were again remarkably similar in both sexes. Logistic regression models confirm the univariate comparison (Table IV), that is, for men most of the information was contained in the LDLC fraction, while for women HDLC was more important. The TC: HDLC ratio was useful in both groups. Since body mass index and history of hypertension, smoking, or medication use was not different between cases and control subjects, they were not included in the model. These models were adjusted for age, however, which was different in women. The likelihood ratios associated with various lipoprotein fractions are shown in Table V. In this analysis LDLC

Volume 110

Lipoproteins

Number 5

Table

IV. Lipids in age-adjusted logistic regression analy-

Table

and coronary

disease

V. Age-adjusted likelihood

in blacks

1009

ratio associated with

lipids

sis Regression coefficient Lipid

Men

TC LDLC HDLC Triglycerides TC: HDLC For ahhreviations, *p < 0.05.

0.016 0.022’ -0.034 -0.004 0.433* see Table

Women 0.010 0.014 -0.osot 0.007 0221t

Lipid

Men

TC LDLC HDLC Triglycerides TC:HDLC

4.21 9.41 1.23 0.70 5.78

For abbreviations, see Table Ratio >3.84; p < 0.05.

III.

III.

tp < 0.01.

was the most sensitive variable for men, while the TC:HDLC was highly discriminatory for women. Previous reports from the Framingham study indicated that TC was not predictive of CAD over the age of 55, while HDLC became even more highly related.15 When our sample was divided at age 55, there were 29 cases and 26 control subjects who were younger and 34 cases and 25 control subjects in the older group. With covariance analysis, both sexes combined, under the age of 55 all four lipid variables were significantly related to the presence of disease (p < 0.05), while only the TC:HDLC ratio was among those 55 or older. DISCUSSION

The relationship of the major lipoprotein fractions to symptomatic CAD is confirmed in this group of black patients. Based on angiographically proved cases, TC, LDLC, and HDLC were all predictive-with varying degrees of consistency-of the presence of CAD. Among men the LDLC fraction contained most of the information, while for women HDLC was a more sensitive discriminator than LDLC. Overall the ratio of TC: HDLC was the most consistent single way to express the lipid findings as correlates of CAD. In general, these findings confirm the sex-specific predictive value of lipoproteins while indicating a somewhat more important role for HDLC among black women.15 Based on age-adjusted rates, the ratio of CAD deaths among black patients in the United States is 1.6: 1.0 men to women compared to 2.2 for white patients.16 Among the younger age groups black women have CAD death rates up to twice as high as those of white women. The basis for the increased susceptibility to CAD among black women is not known; although hypertension must certainly play a role it should not increase rates among black women relative to black men. Recently, differences in lipoproteins

have been proposed

as an explanation

of

this phenomenon. I6 As repeatedly shown in epidemiologic studies,17-‘* HDLC levels in women exceed those of men by 5 to 10 mg/dl at various ages, and in Framingham the ratio of TC: HDLC for women without CAD was 4.4 whereas for men it was 5.1.15 Glueck et alI6 have suggested, on the basis of the survey data from National Health and Nutrition Examination Survey that the widespread problem of obesity among black women may in part deprive them of the protective value of femininity by lowering mean HDLC levels. In fact, a relative “masculinization” of the lipoprotein pattern was observed among the black women studied here. Men and women with CAD had similar levels of HDLC, and the TC: HDLC ratios in all pairs (men and women; cases and control subjects) were similar. In the Framingham data, even women with CAD had slightly lower ratios of TC: HDLC than did men (5.3 vs 5.8).15 Further evidence of the importance of HDLC in influencing sex differences in CAD is evident in the magnitude of the likelihood ratios associated with various Iipoproteins. While HDLC is associated with a likelihood ratio of 14 for men in the Framingham study, the comparable value for women is 21.15 A similar trend exists for the TC: HDLC ratio by sex. l5 This pattern is even more apparent in our data; the likelihood ratio for HDLC for women was 9.1, while for men it was not even significant (1.2). In the only prior study on angiographic CAD and black patients, reported in abstract form by Pearson et al.,‘” HDLC was a risk factor for women but not for men. The data presented here of course do not reflect population trends. As outlined by Pearson,12 a series of powreful selective factors and other sources of bias determine who undergoes cardiac catheterization. In our series black women were more likely than men to be studied, yet only about a third as likely to be found to have the disease. Furthermore, there is no direct evidence in these data that obesity

1010

Cooper

et al.

was a risk factor for HDLC or CAD. In fact, women without CAD were slightly more obese than were cases. Both groups, nonetheless were overweight, and very close to population means for their respective age-sex-race groups.’ A small clinical series such as this, however, might well fail to identify these secondary relationships. The prevalence of hypertension in patients drawn from a predominantly white population undergoing cardiac catheterization for CAD has usually been reported to be in the range of 20% to 30% .lg,20,21 Among all groups in our study a history of hypertension was present in four out of five, and there was no difference between cases and control subjects. The finding of a lower mean blood pressure in cases has been reported in 10 of 12 previous studies and undoubtedly reflects prior myocardial infarction.12 In a population where such a large proportion of patients already has one major risk factor, the relationship of the other risk factors to the disease may be altered. Certainly the mean level of total cholesterol among cases is low, particularly for the men. Assuming roughly comparable laboratory standards, in the Framingam study, for example, mean TC in symptomatic patients was reported to be 240 mg/dl,15 and 230 mg/dl in the Coronary Artery Surgery Study registry. 2oDiabetes and smoking were more common among cases, but the differences were not significant. Angina, as assessed by the London School of Hygiene questionnaire, was also not significantly more frequent in cases. Although this instrument has been previously validated to some extent in the black popu1ation,22,23 among women it was not particularly useful. 23 In a clinical setting this questionnaire is not very sensitive and the physician can probably offer a more accurate opinion based on clinical judgement. Although the patients were evaluated before catheterization, many had a prior diagnosis of CAD and a strong bias to record a positive history of angina would have been likely for all groups. The high prevalence of left ventricular hypertrophy may also have accounted for some of the chest pain. Coronary arteriography has been widely used as the basis for case-control studies in examining the relationship between risk factors and CAD.12 In general, an inverse and graded association with levels of HDLC has been noted.12s24-28HDLC is often the best of the lipids as a predictor24-26 and in a relatively young group of Air Force personnel was associated with an odds ratio of 172:1.24 With increasing age the sensitivity declines. Whether apolipoproteins yield additional information beyond the lipid findings is an area of controversy and intense study at the present time.2gs30

November. 1985 American Heart Journal

Because of the method of recruitment all catheterization-based studies suffer from significant limitations.12 Clearly, a strong bias exists in sampling, since as a rule symptomatic cases who are thought to be surgical candidates will be studied. At the same time, the possibility of measurement bias exists in various forms, as evidenced in our data on angina. In general, these factors would tend to create a false negative outcome, that is, diminish the difference between cases and noncases and weaken the chance of observing a true relationship if it exists. A recent prior myocardial infarction or treatment of elevated lipids would reduce lipid values selectively in CAD cases. Additionally, as noted in a recent study, patients with minimal disease (stenosis less than 50%) have more risk factors for CAD than patients with truly normal coronary arteries, and inclusion of this group among our control subjects would further reduce the case-noncase differences31; we chose to retain this group for sample size considerations. Many cardiovascular drugs alter serum lipids, and this factor has been generally ignored in previous studies. Fortunately, these medications were prescribed equally to both cases and control subjects in our study. Despite these methodologic problems, inherent in all such studies, a consistent series of findings in relation to lipoproteins and coronary atherosclerosis was observed. Since the weight of known biases favors a false negative outcome, it is possible to have greater confidence in positive findings when they are observed. Efforts to define the relationship among risk factors and CAD in black patients must be extended beyond patients undergoing catheterization. Additional data on myocardial infarction from the Cook County Hospital Heart Disease Registry confirm the extraordinary importance of hypertension in this population and suggest a role for diabetes in contributing to the relatively high mortality rate among women.32 Data from two large cooperative trials on treatment of cardiovascular disease have reported similar logistic coefficients for the three major risk factors among black and white patients, with the possible exception of hypertension.33s34 Additional analyses of population-based data are sorely needed. The role of the lipoprotein fractions is less well studied among black patients. In fact, despite the widespread view that the black population has higher levels of HDLC, this conclusion is based mainly on a sample of less than 100 adults studied as part of the Lipid Research Clinic Prevalence Study.‘* Further data on the urban working class and poor black groups now comprising the bulk of the population is urgently needed. These data will help form the basis for intervention against CAD in

"Ql"nw

110

Number

5

this high-risk population and teach us more about the etiologic processes of this widespread disorder. We would like to thank the Fellows in the Division of Adult Cardiology, Cook County Hospital for their excellent cooperation. Lipids were analyzed in the lipid laboratory of the Division of Biochemistry, under the supervision of Dr. Ivan Salamon. We would like to thank Denotra Allen for preparing the manuscript. REFERENCES

1. Report of National Heart, Lung, and Blood Institute Working Conference on Coronary Heart Disease in Black Populations. Johnson KW, Payne GH, editors: Proceedings of a symposium on coronary heart disease in black populations. Gillum RF, editor: AM HEART J lOB(Part 2). 1984. 2. Weisse AB, Abiuso PD, Thind IS: Acute myocardial infarction in Newark, NJ. Arch Intern Med 137:1402, 1977. 3. Hagstrom RM, Federspiel CF, Ho YC: Incidence of myocardial infarction and sudden death from coronary heart disease in Nashville, Tennessee. Circulation 44:884, 1971. 4. Gillum RF: Coronary heart disease in black populations. I. Mortality and morbidity. AM HEART J 104:839, 1982. 5. Cassel J, Heyden S, Bartel AG, Kaplan BH, Tyroler HA, Coroni JC, Hames CG: Incidence of coronary heart disease by ethnic group, social class and sex. Arch Intern Med 128:901, 1971. 6. Tyroler HA, Hames CG, Krishan I, Heyden S, Cooper G, Cassel JC: Black-white differences in serum lipids and lipoproteins in Evan County. Prev Med &541, 1975. 7. Wilson PWF, Savage DD, Catelli WP, Garrison RJ, Donahue RP, Feinlieb M: HDL-cholesterol in a sample of black adults: The Framingham minority study. Metabolism 32:328, 1983. 8. Allain CC. Poon LS. Chan CSG. Richmond W. Fu PC: Enzymatic’ determination of total serum cholesterol. Clin Chem 20:470, 1974. 9. Warnick RG, Benderson d, Albers JJ, Baillie EE, Sexton B, Schaefer EJ. Carlson D. Hill M. Brewer HB. Wiebe DA, Hazlehurst J, Cooper Gk: Dextran sulfate-M& precipita: tion procedure for quantitation of high-density-lipoprotein cholesterol. Clin Chem 28:1379, 1982. 10. Levy RI, Feinleib M: Risk factors for coronary artery disease and their management. In Braunwald E, editor: Heart disease. Philadelphia, 1980, WB Saunders Company, p 1246. 11. The case-control study: Consensus and controversy. J Chron Dis 32:(1/2), 19’79. 12. Pearson TA: Coronary arteriography in the study of the epidemiology of coronary artery disease. Epidemiol Rev 6:140, 1984. 13. SAS user’s guide: Statistics. Cary, NC, 1982, SAS Institute Inc. 14. Kleinbaum DG, Kupper LL: Applied regression analysis and other multivariable methods. Boston, 1978, Duxbury Press. 15. Gordon T, Castelli WP, Hjortland MC, Kannel WB, Dawber TR: High density lipoprotein as a protective factor against coronary heart disease. Am J Med 62:707, 1977. 16. Glueck CJ, Gartside P, Laskarzewski PM, Khoury P, Tyroler HA: High-density lipoprotein cholesterol in blacks and whites: Potential ramifications for coronary heart disease. AM HEART J lOI(Part 2):815, 1984. 17. Heiss G, Johnson NJ, Reiland S, Davis CE, Tyroler HA: The epidemiology of plasma high-density lipoprotein cholesterol levels. Circulation 62(Suppl IV):116, 1980.

Lipoproteins

and coronary

disease

in blacks

1011

18. The Lipid Research Clinics: Population studies data book, vol I, The prevalence study. United States Department of Health and Human Services, NIH Publication No. 80-1527, 1980. 19. Pearson AT, Bulkley BH, Kwiterovich PO, Achuff SC, Gordis L: Anatomically defined coronary disease in blacks: Importance of hypertension as a risk factor (abstr). Circulation 59-6O(Suppl II):14, 1979. 20. Principal Investigators of CASS and their Associates. National Heart, Lung, and Blood Institute Coronary Artery Surgery Study. Circulation 63(Suppl I):1981. 21. Proudfit WJ, Bruschke AVG, MacMillan JP. Williams GW, Sones FM Fifteen year survival study of patients with obstructive coronary artery disease. Circulation 68:986, 1983. 22. Oherman A, Cutter G: Issues in the natural history and treatment of coronary heart disease: Surgical treatment. AM HEART J 108(Part 2):688,1984. 23. Langford HG, Oberman A, Borhani NO, Entswisle G, Tung B: Black-white comparison of indices of coronary heart disease and myocardial infarction in the stepped-care cohort of the Hypertension Detection and Follow-up Program. AM HEART J 108(Part 2):797,1984. 24. Uhl GS, Troxler RG, Hickman JR, Clark D: Relation between high density lipoprotein cholesterol and coronary artery disease in asymptomatic men. Am J Cardiol 48:903, 1981. 25. Zampogna A, Luria MH, Manubens SJ: Relationship between lipids and occlusive coronary artery disease. Arch Intern Med 140:1067, 1980. 26. De Backer G, Rosseneu M, Deslypere JP: Discriminative value of lipids and apoproteins in coronarv heart disease. Atherosclerosis 42:197, 1982. 27. Kaldetzky RG, Asaman G, Walgenhach S, ‘I’auchect P, Hilb HD: Lipoprotein and apoprotein values in coronary angiography patients. Artery 7:191, 1981. 28. Tan MH, Macintosh W, Weldon KL: Serum high density lipoprotein cholesterol in patients with abnormal coronary arteries. Atherosclerosis 37:187, 1980. 29. Maciejko JJ, Holmes DR, Kottke BA, Zinsmeister AR, Dinh DM, Mao SJT: Apolipoprotein A-I as a marker of angiographically assessed coronary artery disease. N Engl J Med 309:385, 1983. 30. Glueck CJ: Future genetic-epidemiological research in coronary heart disease: The decade of the apolipoproteins, dysapolipoproteinemia, and dyslipoproteinmia. In: Genetic epidemiology of coronary heart disease: Past, present, and future. New York, 1984, Alan R. Liss, Inc, p 525. 31. Kahl FR, Crouse JF, Schey HM: Composition of the control group in coronary angiographic studies: Normal coronary status vs nonobstructive coronary disease (ahstr). Circulation 7O(Suppl 11):66, 1984. 32. Cooper R, Prasad R, Cotler RP, Franklin (:. Chali .I, Ferlinz