BULLETIN OF

THE NEXV YORK ACADEMY

OF MEDICINE

VOL. 58, No. 3 APRIL 1982

LIPOPROTEINS AS RISK FACTORS FORCORONARY HEART DISEASE*

HOWARD A. EDER, M.D.Director

Division of Arteriosclerosis Research and Human NutritionAlbert Einstein College of Medicine

Bronx, New York

A LMOST 30 years ago I stood on this same platform and discussed whetherf6 the elevated plasma cholesterol levels found in patients with coronaryheart disease have any role in causing heart disease and whether measuresdesigned to lower serum cholesterol are of any value. Since then we havecome a long way. Not only have epidemiologic studies accumulated a largebody of data demonstrating associations between plasma lipids and lipopro-teins and atherosclerosis, but, perhaps more important, we have begun toachieve an understanding of mechanisms of atherogenesis and of the role oflipoproteins in this process.

Atherosclerosis is a disease whose causes are multifactoral, and alterationsin plasma lipids and lipoproteins are one facet of the problem. A major

* Presented as part of a Symposium on Assessing Therapeutic Dietary Claims held by the Section onClinical Nutrition of the New York Academy of Medicine at the Academy February 28, 1981.

This research was supported by grants from the New York Heart Association and the Barbra StreisandFoundation.

220 HA. EDER

achievement of research has been to show how the various risk factors interactin atherogenesis. To focus solely on lipids and specifically on dietarycholesterol provides a narrow view of the problem. Various factors arecooperative and operate together to produce the disease. While this discus-sion will be largely limited to consideration of lipids and lipoproteins, I shallattempt to show some of these interrelationships. As is now well known,plasma lipids are carried as constituents of lipoproteins, shown in the table.The low density lipoprotein is undoubtedly the lipoprotein most clearlyrelated to atherosclerosis, and I shall use this to illustrate the role oflipoproteins in atherogenesis. Very low density lipoproteins have beendescribed in the Food and Nutrition Board Report' as neutral, but themeaning of neutrality is not clear. High density lipoproteins are now in vogueand are unique in that their levels in the plasma are negatively correlated withcoronary heart disease.Low density lipoprotein levels in the plasma are highly correlated with

levels of plasma cholesterol. Since the latter is far easier to measure, mostclinical and epidemiologic data are in terms of total cholesterol which,however, reflects low density lipoprotein levels. I have taken data from theLipid Research Clinics Prevalence Study2 and plotted low density lipopro-tein cholesterol versus plasma cholesterol. There is a high degree of correla-tion, except at the lower levels of plasma cholesterol where high densitylipoproteins make a relatively greater contribution to total cholesterol. I shallspeak about low density lipoproteins and serum cholesterol interchangeably,but when I speak about serum cholesterol, remember that I am really talkingabout low density lipoproteins.

There is now abundant evidence demonstrating that low density lipopro-tein is a risk factor3. How do we define a risk factor? First, we look at patientswith the disease to see how this factor operates in these patients. We study ex-

perimental animals where this factor can be altered and changes correlatedwith atherosclerosis. We look at this risk factor in various areas throughoutthe world to compare populations with high and low levels of plasmacholesterol. In a homogenous population, such as in Framingham, the lateDr. Cornfield pointed out that food intake is so similar as to make itimpossible to determine correlation. Epidemiologic studies can be retrospec-tive, or preferably prospective, such as the Framingham Study4. Clinical trialis the most definitive, but also the most difficult and expensive. Study of theinteraction of low density lipoprotein with cells provides understanding of itsrole in atherogenesis.

Bull. N.Y. Acad. Med.

220 H.A. EDER

LIPOPROTEINS AND CORONARY HEART DISEASE2

CORONARY HEART DISEASE

Now, first, let us look at patients with coronary heart disease. In studiesthat we conducted at Cornell 30 years ago5, we found that in the 28individuals with coronary heart disease plasma cholesterol values werehigher, as was the cholesterol in the fraction containing low density lipopro-teins and very low density lipoproteins, whereas high density lipoproteincholesterol was reduced. More recently, Goldstein and colleagues' studied500 survivors of myocardial infarction and found that 157 of these subjectshad elevations of serum lipids. That study found that four times as many ofthe patients with myocardial infarction were above the 95th percentile forcholesterol than in the control population. Clearly, in patients who havecoronary heart disease, the prevalence of hypercholesterolemia is higher thanin the control population. Incidentally, this same study demonstrated thatelevated lipid values are much more apparent in the survivors of myocardialinfarction in younger age groups. In general, young populations with coro-nary heart disease have more of the usual risk factors than do older popula-tions, because age by itself is a risk factor. The one possible exception is highdensity lipoproteins. In an older individual, such factors as age and factorsthat we do not know anything about obviously play a greater role.Now what about experimental animals? It has been known for more than

80 years that atherosclerosis can be induced by feeding cholesterol. Anatherosclerotic coronary artery obtained from a rhesus monkey with experi-mental atherosclerosis is indistinguishable from the lesion found in humancoronary artery disease. The interesting thing is that such lesions can beproduced in many nonhuman primates, and these lesions are reversible.Lesions in rhesus monkeys can be induced by feeding monkeys the averageAmerican diet. After resuming a monkey chow diet for periods from threemonths to a year the lesions regress.7 Such data suggest the possibility thatatherosclerosis in man may be reversible. It is of interest that in examining le-sions both in experimental animals and also in humans, one finds proteinscharacteristic of the lipoproteins including apo B, A- 1, and apo C. These arepresent in the lesions, both in the fatty core and in the fibrous plaque. Inpatients with hyperlipidemia twice as much apo B was found in vessel wallsobtained at surgical operations than in patients with low levels of low densitylipoproteins. Lewis8 has injected radio-labeled low density and very lowdensity lipoproteins and has recovered the labeled lipoproteins from vesselsremoved surgically 24 to 48 hours later.

Vol. 58, No. 3, April 1982

221_

22 H. DE

Studies have been done throughout the world and the prevalence ofcoronary heart disease correlated with serum cholesterol.9 In populationswith higher cholesterol intake, heart disease is more prevalent; in populationswith low intakes of cholesterol the prevalence of coronary heart disease islower. The Framingham data'0 showed a high correlation of serum cholester-ol with coronary heart disease. It is of interest that the highest correlation wasin the early years of the study, when the population was younger. In theWestern Electric Study," the coronary death rate was correlated with dietusing a score developed by Keys,'2 which is determined by the dietarycontent of saturated and unsaturated fatty acids and cholesterol. The Hegstedscore,'3 quite similar to the Keys score, showed that the death rate correlatedpositively with saturated fat intake and negatively correlated with polyunsat-urated fatty acid intake. Is the incidence of coronary heart disease reversiblein populations? Data from the United States show that the death rate fromcardiovascular disease and the death rate from noncardiovascular diseasebegan to decrease in the late 1950s, but the rate of decrease became muchgreater about 1967. Since then, the death rate from cardiovascular disease,largely of coronary heart disease, has decreased by approximately 20%. In athick volume published by the National Heart, Lung and Blood Institute,'4an attempt is made to explain this trend. Unfortunately, no definitive conclu-sion was reached. Those interested in smoking note the decreased numbers ofsmokers. Those interested in blood pressure note the change in attitude ofphysicians toward the treatment of hypertension. The lipidologists point to

decreases in serum cholesterol in various large United States populationsfrom the period of 1960-1975. As far as can be ascertained, the changesobserved are not due to changes in cholesterol assay methods which were

monitored by the Communicable Disease Center in Atlanta during the periodof these studies. With each successive study, the curve of serum cholesterolwith age fell. While I would like to believe that this accounts for thedecreasing mortality from coronary heart disease, at the present time no

definitive statement can be made, but the association does exist. Other dataare consistent with this conclusion. In war-time Europe, when Scandinaviancountries'3 and later Germany was blockaded, their food intake was marked-ly altered and the incidence of coronary heart disease fell markedly during the

period of a blockade and then rose steadily thereafter. In Germany,Schettler'6 showed that coincident with the blockade and the change in thefood supply there was a decrease in the amount of saturated fat in the diet anda drop in serum cholesterol which coincided with this decrease in mortality.

Bull. N.Y. Acad. Med.

222 H.A. EDER

LIPOPROTEINS AND CORONARY HEART DISEASE

Further, the Seven Country Study by Keys9 showed a correlation betweendiet and serum cholesterol and with prevalence of coronary heart disease.

HYPERCHOLESTEROLEMIA

Now I would like to turn to biological data. Familial hypercholesterolemiais, as its name implies, a hereditary form of hypercholesterolemia. In thehomozygous form, low density lipoprotein levels are extremely high, andmortality from coronary heart disease is very high. Survivors past age 25 arerare, and myocardial infarction is not at all infrequent in children six or sevenyears of age. In the more common heterozygous form, low density lipopro-tein levels are not as high, but nevertheless the incidence of coronary heartdisease in these subjects is high. Fifty percent of men with heterozygousfamilial hypercholesterolemia have myocardial infarctions before age 50.'"In the course of studies of cholesterol metabolism by cultured cells, Brownand Goldstein'" found that cells from normal subjects bind low densitylipoproteins when they are grown in their absence. However, cells frompatients with homozygous familial hypercholesterolemia do not bind lowdensity lipoproteins, and heterozygotes are intermediate in their binding.These studies led to the conclusion that cells have specific receptor sites forlow density lipoproteins and that these are absent in patients with homozy-gous familial hypercholesterolemia. The failure to bind and to catabolize lowdensity lipoproteins leads to elevation of their plasma levels. Originally,receptors were found only in peripheral cells, but more recently they havebeen found in liver cells. It appears that about half of the low densitylipoproteins is cleared from the plasma through the liver, although notnecessarily by the high affinity receptor mechanism. '9 These studies clearlydemonstrate that a genetic defect in receptor production leads to elevatedplasma low density lipoprotein levels.

Our current model for atherogenesis is based on multiple factors workingin concert. One factor is increased low density lipoproteins, another isendothelial injury."' Elevated levels of low density lipoproteins can induceendothelial injury, which then exposes the intima to high concentrations oflow density lipoproteins that are taken up by receptors in smooth muscle cellsand into macrophages by other mechanisms. Uptake by arterial tissue leads tocholesteryl ester accumulation. While endothelial injury can occur as theresult of high low density lipoproteins concentration, there are other causessuch as hypertension. Platelets aggregate on injured endothelium and this

Vol. 58, No. 3, April 1982

223

224 H.A. EDER

causes the platelet growth factor to be released, a factor that causes smoothmuscle proliferation. These various events lead to the development of thecomplicated lesion of atherosclerosis. It is evident that low density lipopro-tein plays a role in atherogenesis and that measures to maintain lower levels inthe population or lowering levels in subjects with elevated levels have arational basis.

Only three of the clinical trials Dr. Olson discussed were for primaryprevention."2 The rest were secondary prevention trials in patients withestablished coronary heart disease. These populations probably have faradvanced and irreversible lesions. In the three primary prevention trials therewere a total of 1,700 subjects, most of them more than 50 years of age. Ofthese trials, two showed positive but not significant differences in incidenceof coronary heart disease. Dr. Olson has mentioned that no recommendationcould be made until there is definitive proof from an adequate clinical trial.The Lipid Research Clinics Primary Prevention Trial22 may answer thisquestion. The Diet Heart Study was proposed to test this hypothesis. It wasconcluded that in 1971 this study would be so expensive that it was notfeasible.23 In the political and economic climate of 1981 it is even less likelyto be performed so that no definitive answer will be available and we must de-pend on the data that we have at hand. It should be recalled that the WorldHealth Organization Study24 with clofibrate did show a 20% reduction in theincidence of new myocardial infarctions, something often overlooked inconcern over the toxicity of the drug. The evidence that low density lipopro-tein is a risk factor is overwhelming, and it is reasonable to institute measuresto reduce it.How do we reduce it? It has already been pointed out that low density

lipoprotein levels are affected by levels of saturated fatty acids in the diet so

that reducing saturated fatty acids in the diet will lead to reduction of lowdensity lipoproteins. In the Indians studied by Dr. Connor in Mexico,25 hefound a good correlation between dietary cholesterol and plasma cholesterol.It is pointed out that these subjects ate diets with a high P:S ratio, and it couldwell be argued that this may not be applicable to the American population.Dr. Ahrens and his associates demonstrated clearly that substitution ofpolyunsaturated fat for saturated fat in the diet lowered serum cholesterol.Substitution of coconut oil, which does not contain saturated fatty acids,causes the cholesterol level to return to baseline levels and corn oil causedcholesterol to decrease. This is an unequivocal proof that by altering the dietwe can change levels of low density lipoproteins. Studies have beenperformed in free living Scandinavian populations on a diet low in

Bull. N.Y. Acad. Med.

224 H.A. EDER

LIPOPROTEINS AND CORONARY HEART DISEASE 225

saturated fat and high in polyunsaturated fatty acids. In almost all in-stances, serum cholesterol levels in these individuals fell.We must therefore make a decision on the basis of numerous types of

evidence but without the definitive evidence of an adequate trial whetheraltering the diet to lower low density lipoprotein levels is rational at this time.My feeling is that one has to make a choice and that the evidence indicatesthat intervention is desirable.

There is considerable disagreement as to the significance of very lowdensity lipoprotein as a risk factor. (Since very low density lipoproteins infasting subjects and in patients with elevated levels of very low densitylipoproteins without hyperchylomicronemia have a high correlation withplasma triglyceride concentrations, it is reasonable to use triglyceride as ameasure of very low density lipoproteins.) Epidemiologists using multivar-iate analysis state that hypertriglyceridemia is not a primary risk factor, andsome do not believe it useful to measure plasma triglycerides.26 However, ithas long been known that patients with coronary heart disease have, onaverage, higher levels of triglyceride. Thus, in the study of Goldstein et al.6the patients with coronary heart disease had higher levels of triglycerides thandid the control subjects. In some populations, such as in Sweden,27 prospec-tive studies have shown a high correlation between plasma triglycerideconcentrations and the incidence of myocardial infarction.

In an analysis of the Western Collaborative Study, Hulley et al.26 foundthat an elevation of triglycerides by one standard deviation increased risk by36%. However, by performing multivariate analysis using multiple riskfactors, the increased risk from triglyceride was no longer apparent. Fromthis it was concluded that the evidence of a causal relationship betweentriglyceride and coronary heart disease is so inadequate that hypertriglyceri-demia should not be treated and plasma triglycerides should not be measured.It is noteworthy that by a similar multivariate analysis obesity cannot bedemonstrated to be a primary risk factor. However, Hulley judged that "in thelight of prior biologic knowledge, " obesity is a causal risk factor. By the samereasoning, very low density lipoprotein could be judged to be a causal factor,and there is considerable biologic knowledge about its metabolism. It ismetabolized to low density lipoprotein, which is a strong risk factor. Interme-diates in the metabolism of triglyceride-rich lipoproteins, i.e., chylomicronremnants and intermediate density lipoprotein, may be atherogenic,28 andfinally the level of very low density lipoprotein has a significant negativecorrelation with the level of high density lipoprotein. Thus, if the samerationale as applied to obesity is applied to very low density lipoprotein, it is

Vol. 58, No. 3, April 1982

225LIPOPROTEINS AND CORONARY HEART DISEASE

EDE

difficult to agree with Hulley 's conclusion. Further, it may be recognized thatthe use of this measurement by epidemiologists to determine risk in apopulation is far different from that of physicians treating a patient. In apatient an elevated triglyceride is a cause for concern in that it may indicategenetic forms of hyperlipidemia such as familial combined hyperlipidemia ortype III hyperlipoproteinemia, which carry a high risk of coronary heartdisease.'7 Elevated triglycerides may be secondary to other diseases, such asdiabetes or hypothyroidism. Further, markedly high levels of triglyceridescause pancreatitis. Finally, the level of triglycerides is important in assessingthe significance of a reduced level of high density lipoprotein. Although verylow density lipoprotein may not be a primary risk factor, it may have a causalrelationship to coronary heart disease, and physicians should certainly in-clude its measurement when determining lipid profiles. In most instances,patients are overweight and, with caloric restriction, triglyceride levelsdecrease dramatically. In most patients, except those with levels of triglycer-ide so high as to put them at risk for pancreatitis, the use of drugs is notnecessary.

In 1951 we presented data showing that high density lipoprotein-cholester-ol was reduced in patients with coronary heart disease. ' Interest in thislapsed until 1975, when Drs. George and Norman Miller29 proposed thehypothesis that high density lipoprotein was a protective factor againstcoronary heart disease. Subsequently, a number of studies were done. Aretrospective study by Castelli et al.° (The Cooperative Phenotyping Study)showed that the prevalence of coronary heart disease was much higher inindividuals with low levels of high density lipoprotein and with higher levelsthere were fewer cases of coronary heart disease. The Framingham popula-tion was examined more than 20 years after the start of the study.3' Thisprospective study showed a strong negative correlation between the level ofhigh density lipoprotein-cholesterol and coronary heart disease. It was foundthat measurement of high density lipoprotein had greater predictive valuethan measurement of low density lipoprotein or total cholesterol. However, itshould be noted that this study was done in an older population, where lowdensity lipoprotein has less predictive value. We have recently shown thatmeasurement of subclasses of high density lipoprotein32 may be useful andthat HDL2 is the more variable component.33 In men, where high densitylipoprotein is lower than in women, HDL2 is considerably lower than inwomen, whereas HDL3 levels are similar in men and women. Joggers have

higher levels of high density lipoprotein and HDL2 . Men with coronary heart

Bull. N.Y. Acad. Med.

226 H.A. EDER

LIPOPROTEINS AND CORONARY HEART DISEASE 227

LIPOPROTEINS OF NORMAL HUMAN PLASMA

Density Electrophoretic Diameter MolecularDesignation (g. Iml.) mobility (A) weight

Chylomicrons < 0.95 Origin 1_-104 >4 x I0OVLDL 0.95 1.006 pre-,/ 250-800 5-10 x 106IDL 1.006 1.019 83 or pre-/3 250-300 4-5 x 10"LDL 1.019 1.063 /3 200-250 2-3 x 10"HDLHDL, 1.063 1.125 a, 70 120 3.9 x 1O'HDL1 1.125 1.210 a, 50 100 1.9 x 10'

disease have low levels of high density lipoprotein and very low levels ofHDL2. HDL2 is low in coronary heart disease and diabetics withouttreatment. Insulin administration raises high density lipoprotein, especiallyHDL2.

SUMMARY

Low density lipoprotein is clearly a major risk factor for coronary heartdisease. There are less definitive data regarding the role of very low densitylipoprotein, but it does have a strong association with coronary heart disease.Studies of the role of tr'glyceride-rich lipoproteins in atherogenesis supportthis view and further data on its role are needed. While there is considerableevidence that high density lipoprotein levels are negatively correlated withrisk, there are as yet no data on the effects of altering these levels inindividuals or in populations.

REFERENCES

1. Food and Nutrition Board, National Re-search Council: Toward HealthfulDiets. Washington, D. C., Nat. Acad.Sci., 1980.

2. Lipid Research Clinics Program Epide-miology Committee: Plasma lipid distri-butions in selected North American pop-ulations: The lipid research clinicsprogram prevalence study. Circulation60:427-39, 1979.

3. Steinberg, D.: The rediscovery of highdensity lipoprotein: a negative risk factor

in atherosclerosis. Eur. J. Clin. Invest.8:107-09, 1978

4. Dawber, T. R.: The Framingham Study.Cambridge, Harvard University Press,1980.

5. Barr, D. P., Russ, E. M., and Eder, H.A.: Protein-lipid relationships in humanplasma II. In atherosclerosis and relatedconditions. Am. J. Med. 11:480-93,1951.

6. Goldstein, J. L., Hazzard, W. R.,Schrott, H. G., et al.: Hyperlipidemia in

Vol. 58, No. 3, April 1982

228 H.A. EDER

coronary heart disease I. Lipid levels in500 survivors of myocardial infarction.J. Clin. Invest. 52: 1533-43, 1973.

7. Armstrong, M. L.: Regression ofAtherosclerosis. In: Atherosclerosis Re-views, vol. 1, Paoletti, R. and Gotto, A.M., editors. New York, Raven, 1976,pp. 137-82.

8. Onitiri, A. C., Lewis, B., Bentall, H.,et al.: Lipoprotein concentrations in se-rum and in biopsy samples of arterialintima. A quantitative comparison.Atherosclerosis 23:513-19, 1976.

9. Keys, A.: Coronary heart disease in sev-en countries. American Heart Associ-ation Monograph No. 29. Circulation41 (Suppl. 1): 1-199, 1970.

10. Kannel, W. B., Castelli, W. P., andGordon, T.: Cholesterol in the predic-tion of atherosclerotic disease. Ann. Int.Med. 90:85-91, 1979.

11. Shekelle, R. B., Shryock, A. M., Paul,O., et al.: Diet, serum cholesterol, anddeath from coronary heart disease: TheWestern Electric Study. N. Engl. J.Med. 304:65-70, 1981.

12. Keys, A., Anderson, J. T., and Grande,F.: Serum cholesterol response tcchanges in the diet. Metabolism 14:747-87, 1965.

13. Hegsted, D. M., McGandy, R. B.,Myers, M. L., and Stare, F. J.: Quanti-tative effects of dietary fat on serumcholesterol in man. Am. J. Clin. Nutr.17: 281-95, 1965.

14. Proceedings of the Conference on theDecline in Coronary Heart DiseaseMortality, Havlik, R. J. and Feinleib,M., editors. NIH Publication No. 79-1610.

15. Malmros, H.: The relation of nutritionto health - A statistical study of theeffect of the war-time on arteriosclero.sis, cardiosclerosis, tuberculosis an(diabetes. Acta Med. Scand. (Suppl.246: 137-53, 1950.

16. Schettler, G. Personal communication.17. Fredrickson, D. S., Goldstein, J. -L.

and Brown, M. S.: The Familial Hyperlipoproteinemias. In: The Metabolic Ba.sis ofInherited Disease, Stanbury, J. B.Wynagaarden, J. B., and FredricksonD. S., editors. New York, McGrawHill, 1978, pp. 604-55.

18. Brown, M. S. and Goldstein, J. L.:Familial Hypercholesterolemia: Modelfor Genetic Receptor Disease. In: Har-vey Lectures 1977-78, Series 73 NewYork, Academic, 1979, pp. 163-201.

19. Steinberg, D., Pittman, R. C., Attie, A.D., et al.: The Role of the Liver andLDL Catabolism. In: Atherosclerosis,vol. 5, Gotto, A. M., Jr., Smith, L. C.,and Allen, B., editors. New York,Springer Verlag, 1980, pp. 800-03.

20. Ross, R. and Glomset, J. A.: The patho-genesis of atherosclerosis. N. Engl. J.Med. 295: 369-77, 1976.

21. Ahrens, E. H., Jr.,: The management ofhyperlipidemia: whether, rather thanhow. Ann. Intern. Med. 85:87-93,1976.

22. The Lipid Research Clinic Program:The coronary primary prevention trial:Design and implementation. J. ChronicDis. 32: 609-31, 1979.

23. Arteriosclerosis. Report by NationalHeart and Lung Institute Task Force onArteriosclerosis, vol. 1. NIH Pub. No.72-137, 1972.

24. Oliver, M. F., Heady, J. A., Morris, J.N., and Cooper, J.: A co-operative trialon the primary prevention of ischaemicheart disease using clofibrate. Br. HeartJ. 40: 1069-1118, 1978.

25. Connor, W. E., Cerqueira, M. T., Con-nor, R. W., et al.: The plasma lipids,lipoproteins, and diet of the TarahumaraIndians of Mexico. Am. J. Clin. Nutr.31:1131-42, 1978.

26. Hulley, S. B., Rosenman, R. H.,Bawol, R. D., and Brand, R. J.: Epide-miology as a guide to clinical decisions.The association between triglyceridesand coronary heart disease. N. Engl. J.Med. 302:1383-89, 1980.

27. Carlson, L. A., Bottiger, L. E., andAhfeldt, P. E.: Risk factors for myocar-dial infarction in the Stockholm prospec-tive study. Acta Med. Scand. 206:351-60, 1979.

28. Zilversmit, D. B.: Atherogenesis: Apostprandial phenomenon. Circulation60:473-85, 1979.

29. Miller, G. J. and Miller, N. E.: Plasma-high-density-lipoprotein concentrationand development of ischaemic heart-dis-ease. Lancet 1: 16-19, 1975.

Bull. N.Y. Acad. Med.

LIPOPROTEINS AND CORONARY HEART DISEASE 229

30. Castelli, W. P., Doyle, J. T., Gordon,T., et al.: HDL cholesterol and otherlipids in coronary heart disease. The co-operative lipoprotein phenotyping study.Circulation 55:767-72, 1977.

31. Gordon, T., Castelli, W. P., Hjortland,M. C., et al.: High density lipoprotein asa protective factor against coronary heartdisease. Am. J. Med. 62:707-14, 1977.

32. Gidez, L. I., Miller, G. J., Burstein,M., and Eder, H. A.: Analysis of Plas-ma High Density Lipoprotein Sub-classes by a Precipitation Procedure:

Correlation with Preparative and Ana-lytical Ultracentrifugation. In: Report ofthe High Density Lipoprotein Method-ology Workshop, Lippel, K., editor.NIH Pub. No. 79-1661, 1979, pp. 328-42.

33. Eder, H. A., Feder, A., and Gidez, L.I.: Lipoproteins and Lipoprotein Sub-classes in Diabetes Mellitus. In: Diabe-tes 1979, Waldhausl, W., editor. Am-sterdam, Excerpta Medica, 1980, pp.600-04.

Vol. 58, No. 3, April 1982