Special Report Rationale of the Diet-Heart Statement …atvb.ahajournals.org/content/atvbaha/2/2/177.full.pdfSpecial Report Rationale of the Diet-Heart Statement of the American Heart

Special Report

Rationale of the Diet-Heart Statement of theAmerican Heart Association

Report of the AHA Nutrition Committee

A fundamental goal of the American Heart Association(AHA) is to prevent cardiovascular disease and, in

particular, to reduce the incidence of coronary heart dis-ease (CHD) and other atherosclerotic diseases in our soci-ety. A question of great concern to the AHA has beenwhether the American diet is a significant factor in thegenesis of atherosclerosis. This question has been thesubject of continuous dialogue within the Association for atleast 25 years. Although many and frequently diverse opin-ions have been voiced within the several committees of theAHA, a concerted effort has been made to integrate thebest available evidence on the subject. It is clear that the"diet-heart question" is complex, and new information isemerging continually. Still, a broad base of knowledge hasbeen accrued, and because of the urgency engendered bythe high incidence of CHD in the U.S., the AHA has felt aresponsibility to provide the best possible guidance to themedical community and the American public on the dietquestion. For this reason the policy of the AHA has been tofrequently update recommendations on the basis of thebest currently available evidence. The rationale and docu-mentation for particular recommendations of the AHA arediscussed below.

I. History of AHA Diet Statements, 1957 to Present

In 1957 a group chaired by Dr. Irvine H. Page prepared areport to the AHA summarizing the evidence on the rela-tionship between diet and atherosclerosis.1 A review of theavailable data led these investigators to the following con-clusions: (1) diet may play an important role in thepathogenesis of atherosclerosis; (2) the fat content and to-tal calories of the diet may be the dominant contributingfactors; and (3) the type of fat, or the balance betweensaturated and certain unsaturated fats, also may be impor-tant. In 1961 an ad hoc committee authorized by the AHA

This report was prepared by the AHA Nutrition Committee: ScottM. Grundy, M.D., PhD., Chairman; David Bilheimer, M.D., HenryBlackburn, M.D., W. Virgil Brown, M.D., Peter O. Kwiterovjch, Jr.,M.D., Fred Mattson, Ph.D., Gustav SchonfeW, M.D., William H.Wekdman, M.D., Members

Approved by the AHA Steering Committee on Medical andCommunity Program, December 11, 1981.

A reprint copy may be obtained from the American Heart Asso-ciation, 7320 Greenville Avenue, Dallas, Texas 75231 (ask for No.72-202-A). (Republlshed from Circulation 1982:65(4):839A-854Aby permission of American Heart Association.)

(Arteriosclerosis 4:177-191, March/April 1982)

prepared an updated report on the possible relation of di-etary fat to heart attacks and strokes.2 The authors of thisreport reached the following conclusions: (1) overweightpersons should decrease their caloric intake and attempt toachieve their desirable body weight; (2) weight reductionshould be facilitated by regular, moderate exercise; (3) thecomposition of the diet should be altered by reducing in-takes of total fats, saturated fats and cholesterol, and byincreasing polyunsaturated fats; (4) particular attentionshould be given to dietary alteration by men at increasedrisk for CHD (e.g., those with a previous atheroscleroticevent, a strong family history of CHD, elevated plasmacholesterol or hypertension); and (5) for those at high risk,dietary changes should be carried out under medical su-pervision.

Since the time of these recommendations, the AHA hascontinuously reviewed new data pertaining to the possiblerelationship between diet and CHD. The responsibility forupdating recommendations about diet has rested primarilywith the Nutrition Committee. This committee consists of agroup of scientists representing several disciplines (espe-cially epidemiology, pathology, nutrition and metabolic dis-eases). It meets two to three times per year to evaluatenew information, to reexamine its position in the light of thisinformation and, when appropriate, to develop new rec-ommendations. Tenure of membership on the committee is3 years, and care is taken to recruit new members who areexperts in their fields. The Nutrition Committee worksclosely with the various councils of the AHA (for example,Arteriosclerosis, Hypertension, Epidemiology, Cardiovascu-lar Disease of the Young, and Clinical Cardiology) to obtaintheir views. The product of these deliberations is subjectedto systematic review by the Council Affairs Committee andSteering Committee before it is finally approved. The de-velopment of a new statement on diet and CHD usuallytakes 3-5 years.

The AHA issued another statement on diet in 1965.3 Itrecommended caloric restriction to achieve desirableweight, substitution of polyunsaturated fats for saturatedfats where possible, reduction in cholesterol intake and in-clusion of the whole family in dietary changes. The 1965report did not specify precise quantities or set limits for thedifferent dietary constituents to achieve these goals. In1968 a new statement followed the same general recom-mendations,4 but it attempted to define more precisely de-sirable intakes of different nutrients. Weight reduction wasagain recommended. Intake of fat was set at 30-35% oftotal calories with a distribution of one third saturates, one

177

by guest on July 1, 2018http://atvb.ahajournals.org/

Dow

nloaded from

http://atvb.ahajournals.org/

178 ARTERIOSCLEROSIS VOL 2, No 2, MARCH/APRIL 1982

Table 1. Ten-Year CHD Deaths per 10,000 Males by Level of Total Serum Cholesterol (Standardized atAge 50)*

Population (male)

U.S. railroad

Finland

Mediterraneant

At risk

2246

1450

3321

Events

93

64

41

Total

200

247

270

103

serum cholesterol

250

399

381

178

(mg/dl)

300

637

534

306

*Data adapted from Keys,1421980.fGreece, Yugoslavia and Italy.

third monounsaturates and one third polyunsaturates. A re-duction of cholesterol intake to less than 300 mg/day alsowas recommended.

Since 1968 two statements have been issued, one in19735 and the most recent in 1978.6 The AHA has con-tinued to recommend that the general public consume adiet containing no more than 30-35% of calories as fat;further, the decrement in saturated fats should be replacedby complex carbohydrates and polyunsaturated fats, butintake of polyunsaturates should not exceed 10% of totalcalories. Cholesterol intake again was limited to 300 mg/day. Finally, relatively tow intakes of salt were suggestedbecause evidence exists that current levels of sodium in-take may raise blood pressure in many people.

II. Current Dietary Recommendations

The current recommendations of the AHA are based onthe concept that modification of risk factors should de-crease the danger of CHD. These risk factors include (1)elevated plasma cholesterol, (2) increased blood pressure,(3) smoking, (4) diabetes mellitus and (5) marked obesity.The interrelationships among diet-related risk factors andCHD must be reviewed briefly for an understanding of therationale behind the dietary recommendations. Particularattention should be given to the relationship between dietand plasma cholesterol.

A high plasma cholesterol is a well-established riskfactor.7 However, cholesterol in plasma is not homoge-nous but is transported in combination with specificaggregates of lipids and proteins called lipoproteins. Nor-mally, most plasma cholesterol is carried in low densitylipoprotein (LDL), and a high LDL-cholesterol is a signifi-cant risk factor for CHD.89 The most atherogenic of allthe lipoproteins appears to be LDL. A smaller fraction ofthe plasma cholesterol which is present in the tri-glyceride-rich lipoprotein, very low-density lipoprotein(VLDL), is increased in patients with hypertriglycer-idemia. Whether cholesterol In normal VLDL is athero-genic, and if so, to what extent, remains to be deter-mined. The remaining plasma cholesterol is transportedin high-density lipoprotein (HDL). Recent studies suggestthat HDL-cholesterol is not atherogenic. Indeed, an in-verse relationship appears to exist between HDL concen-trations and risk for CHD.710 To summarize: high con-centrations of total cholesterol, and particularly LDL, in-crease CHD risk; a low HDL may further enhance risk;and evidence on VLDL is incomplete.

For the purpose of modifying the above risk factors,especially high plasma cholesterol, the AHA makes thefollowing dietary recommendations.

Reduction In Saturated Fatty Acids

Saturated fats repeatedly have been shown to raiseplasma cholesterol and LDL levels.11"15 The currentAmerican diet contains about 40% of total calories as fat,of which about 15 - 17% are derived from saturated fats.The major sources of saturated fats are meat, animalfats, some vegetable oils (palm oil, coconut oil, cocoabutter, and heavily hydrogenated margarines andshortenings), dairy products (whole milk, cream, butter,ice cream and cheese) and bakery goods. The AHA rec-ommends a reduction of saturated fat intake to less than10% of total calories. Equations developed in carefullycontrolled clinical studies13"15 have shown that a reduc-tion in saturated fats from 17% to 10% of total calorieswill decrease plasma cholesterol about 20 mg/dl.

A reduction in saturated fats requires consideration ofwhat should be substituted in their place. For an over-weight person, removal of saturated fat can decrease to-tal calories and promote weight reduction. For people ofnormal weight, saturated fats can be replaced by unsatu-rated fats and complex carbohydrates.

Substitution of Unsaturated Fats

The fatty acids of these fats can be either monounsatu-rated or polyunsaturated. Oleic acid is the major repre-sentative of the former; linoleic acid is the dominantpolyunsaturated fatty acid. Both types of fatty acids lowerplasma cholesterol when substituted for saturated fats.Linoleic acid is somewhat more effective in cholesterolreduction than oleic acid1113"15 and, consequently, therehas been a widely held belief that saturated fats shouldbe replaced as much as possible by polyunsaturatedfats; this belief may be one explanation for a gradual in-crease in polyunsaturates in the U.S. diet.16 Still, at pres-ent, they only provide 5-6% of total calories. The AHAhas been cautious about recommending marked in-creases in polyunsaturates; the consequences of pro-longed ingestion of large quantities of these fats are notknown. To be on the safe side, the AHA has not recom-mended very high intakes of polyunsaturated fats for thegeneral population. While some replacement of saturatedfats by polyunsaturates seems safe, it may be prudentnot to exceed 10% of total calories.

Another substitute for saturated fatty acids could bemonounsaturated fatty acids. The latter do not reduceplasma cholesterol quite as much as polyunsaturates,but they have certain potential advantages. The humanbody synthesizes considerable quantities of monoun-saturated fatty acids, and thus these fatty acids probably


Dow

nloaded from


SPECIAL REPORT OF THE AHA NUTRITION COMMITTEE 179

do not possess any inherent "toxicity." Large amountsof monounsaturates are ingested as olive oil in theMediterranean region without evidence of adverseeffects. Indeed, the prevalence of CHD in this region ofthe world is relatively low compared to countries whereingestion of saturated fats is high.1718 At present, no in-expensive sources of natural monounsaturated fats simi-lar to olive oil are readily available in the U.S., but thereare several potential sources of monounsaturates whichcould be explored were they deemed a useful substitutefor saturated fats. Actually, a major portion of monoun-saturated fatty acids in the U.S. diet are consumed withanimal fats, and a reduction of saturates to 10% of caloricintake would produce a similar decrease in monounsatu-rates. The AHA does not at present recommend increas-ing these monounsaturated fatty acids since total fatcalories would then exceed 30%. Instead, the recom-mendation is to replace saturated fats principally withpolyunsaturates and carbohydrates.

Increase In Carbohydrates

Reducing caloric intake from fats to a maximum of30% would require an increase in carbohydrates from45% to 55% of total calories. This increase probablyshould take the form of complex carbohydrates: the long-chain carbohydrates, as contained in vegetables, beans,cereals and some fruits, seem preferable to simplesugars because they impart less risk for dental caries,avoid surges of hyperglycemia and provide more rough-age to the diet. When carbohydrates are substituted forsaturated fatty acids, plasma cholesterol and LDL usuallyfall; this decline is in the range of that with monoun-saturated fatty acids and somewhat less than with poly-unsaturates.15 An increase in dietary carbohydratecauses a rise in the plasma triglyceride level of fastingpatients,1920 but in persons without abnormal lipid levelsthe rise is usually small and transient.21|22 Furthermore, areduction in fat intake decreases the rise in plasma trig-lycerides that normally follows fat ingestion. Finally, inpopulations habitually consuming high-carbohydratediets, plasma triglycerides are not increased notice-ably.2122 Thus, the low total cholesterol on a low-fat,high-carbohydrate diet would seem to outweigh anyslight increase in plasma triglycerides.

Substantial Reduction In Dietary Cholesterol

Most carefully controlled metabolic studies haveshown that dietary cholesterol increases total plasmacholesterol.131523"30 In recent years the cholesterol in-take of the U.S. public may have declined, due mainly toa decreased consumption of eggs3132; at present, in-takes probably average 450 - 500 mg per day,1631 butundoubtedly there is much individual variation. The AHArecommends a further reduction to less than 300 mg ofcholesterol per day. For each 100 mg per day decreasein dietary cholesterol, the total plasma cholesterol falls anaverage of about 7 mg/dl.15'30

Caloric Intake Adjusted To Achieve andMaintain Desirable Weight

When diet-related risk factors are present they usuallyare accentuated by obesity and its associated increasedcaloric intake. Excess body weight frequently is associ-ated with high VLDL, moderately high LDL, low HDL,hypertension and diabetes mellitus. Furthermore, in

young and early middle-aged men, obesity seems to en-hance risk for CHD independent of its effect on theseother risk factors.7 Several reports indicate that correc-tion of obesity can reduce plasma lipid levels33 increaseHDL,34 decrease blood pressure and improve glucosetolerance. Thus, avoidance of obesity beginning early inlife, or a supervised weight reduction for those abovetheir desirable weight, is strongly recommended. Moder-ate exercise is a useful adjunct to caloric restriction in aprogram for weight control.

III. Rationale and Documentation ofAHA Position on Diet

Evidence for a Dlet-CHD Relationship

Early investigations demonstrated that dietary cho-lesterol can cause atherosclerosis in several animal spe-cies.35"37 This finding led to the concept that excess diet-ary cholesterol also may accelerate atherosclerosis inman. In animals the atherogenic influence of dietarycholesterol appears to be mediated to a large extentthrough the plasma cholesterol. When plasma cholester-ol is raised by dietary cholesterol, atherogenesis is en-hanced; in animals resistant to hypercholesterolemia,atherosclerosis usually is minimal. Therefore, in animals,cholesterol and perhaps other dietary components seematherogenic in proportion to their hypercholesterolemiceffect. The same could be true for man, but before thevalidity of this hypothesis can be considered, it seems aprerequisite to review evidence relating plasma cho-lesterol, or lipoproteins, to human atherosclerosis.

Plasma Cholesterol, Lipoproteins and CHD

Epidemiologic StudiesSeveral large epidemiologic surveys have shown a sig-

nificant correlation between total plasma cholesterol andthe incidence of CHD.71738"49 These surveys have beencarried out both in the U.S. and other countries. One ofthe more germane to the U.S. population is the PoolingProject,7 in which the data from several similar epi-demiologic studies in the U.S. were pooled. The projectcompared CHD risk to serum cholesterol concentrationsthrough five steps of increasing cholesterol levels; thepopulations under study were men aged 40-64 years.The five quintiles of serum cholesterol were I (< 194 mg/dl), II (195-218 mg/dl), III (219-240 mg/dl), IV (241-268mg/dl) and V (> 268 mg/dl). There were no significantdifferences in CHD rates between quintiles I and II, butrisk for CHD increased in the higher quintiles (risk ratios= 1.5, 1.64 and 1.99, respectively, for quintiles III, IV andV as compared to I and II). These differences in risk weregreater at younger ages. For instance, at ages 45 to 49the risk ratio of IV/I + II was 2.9 and of V/l+ll was 3.6.With advancing years the differential risk between quin-tiles declined; for example, at ages 60 to 64 the relativerisk for quintiles V/l+ll was only 1.5. This does not meanthat elevated plasma cholesterol is less dangerous in old-er people; however, the absolute number of patients withhigher cholesterol levels rises with age so that more peo-ple are in the higher quintiles and thus are at greater riskin the older age groups.

Generally, total cholesterol concentrations are a goodreflection of LDL levels, and the high correlation betweenplasma cholesterol and CHD risk thus suggests that highconcentrations of LDL is atherogenic. In support, recentepidemiologic studies indicate that elevated LDL is an in-


Dow

nloaded from



dependent factor.8950 A direct relatonship between LDLconcentrations and extent of coronary atherosclerosishas been shown in several angiographic studies.51'52

Epidemiologic studies also show an inverse correlationbetween CHD rates and HDL levels in populations withrelatively high levels of total plasma cholesterol.7'10'53'54

On thjB other hand, low HDL may not increase risk forCHD in populations in which LDL concentrations also arelow.55 Thus, in the absence of relatively high LDL, mildlydecreased HDL may not be particularly dangerous.

Finally, most epidemiologic studies have found astronger correlation between plasma cholesterol andCHD than for plasma triglycerides56; there are excep-tions, however. Clinically, many patients with CHD haveraised triglyceride levels, and hypertriglyceridemia maydenote increased risk in some patients. Still, it may bequestioned whether high triglycerides are an indepen-dent (or causative) risk factor; large amounts of triglycer-ides are not found in atherosclerotic plaques. A highplasma triglyceride may accompany some other abnor-mality in lipoprotein metabolism that is associated withgreater risk. For example, high triglycerides can beassociated with low HDL,5657 higher LDL58 or p-VLDL,59

all of which seemingly have atherogenic potential. Ele-vated triglycerides, at least for some patients, may be amarker for increased CHD risk even if they are not acausative factor.

What Is the Ideal Plasma Cholesterol?Few investigators doubt that high concentrations of

plasma cholesterol (and LDL) contribute to coronaryatherosclerosis. However, the relation between CHD andcholesterol levels has not been shown to be linear over abroad range of cholesterol concentrations. We might in-quire as to the ideal plasma cholesterol level. In a recentworkshop60 a group of epidemiologists, clinical investiga-tors and experimental pathologists attempted to developa consensus on the ideal cholesterol level from data ofseveral kinds. They concluded that the ideal level foradults would be in the range of 130 — 190 mg/dl (mean160 mg/dl).

A question of importance is whether the relationshipbetween plasma cholesterol and CHD is linear over abroad range of cholesterol levels. The results of the Pool-ing Project7 might be taken as evidence against such arelationship; in this study the risk for CHD in men withcholesterol levels in the first quintile (<194 mg/dl) wasnot clearly reduced compared to those in the secondquintile (195-218 mg/dl). Men with especially low plasmacholesterol may have had less CHD, but the evidencewas weak. Evidence for a linear association, on the otherhand, comes from multi-country epidemiological studiesin which CHD risk is extremely low. In these populationscholesterol levels are also very low. In reality the rela-tionship probably is curvilinear, so that above a certain"threshold" region, risk accelerates with rising cholester-ol levels. If the results of the Pooling Project7 are takenas evidence, this threshold range for total cholesterolprobably is in the range of 200-220 mg/dl, and desirablecholesterol concentrations would be below this level.This conclusion is consistent with the above-proposedideal range of 130-190 mg/dl. This desirable range isconsiderably below the cholesterol distribution in U.S.adults today; at least 50-60% of adult American men cur-rently have concentrations greater than 200 mg/dl andthus are in the zone of accelerating risk.

Although the concept of a desirable range for plasma

cholesterol probably has validity, it may have to be mod-ified under certain circumstances. For example, a reduc-tion of plasma cholesterol to 200 mg/dl might not reduceCHD risk to an acceptable level if a person smokes orhas either hypertension or diabetes mellitus. Further-more, plasma total cholesterol is not the only lipid factorrelated to CHD. Concentrations of HDL and other lipopro-tein subspecies, or their components, may affect CHDrisk Independent of total cholesterol. The same may betrue of the lipid composition of the diet, as is discussedsubsequently. Thus, the concept of a desirable range fortotal cholesterol probably pertains more to populationthan to individuals.

Genetic HyperiipidemiaAlthough dietary factors undoubtedly raise plasma

cholesterol, genetic influences on lipoprotein concentra-tions must not be overlooked. Genetic factors probablypredominate in people whose plasma LDL concentra-tions are above the 90-95th percentile of the population.Some patients at these high levels have monogenicforms of hypercholesterolemia. For instance, at least onein 500 people has familial hypercholesterolemia,61 '65 adisorder associated with severe elevations of plasmaLDL, defective clearance of LDL60"70 and very prematureCHD. Another monogenic disorder is familial combinedhyperiipidemia6'162; this disease is characterized by in-creases in LDL, VLDL or both, and it too is associatedwith premature CHD61'6271; indeed, up to 10% of patientswith CHD may have this lipoprotein disorder.6162 Yetanother monogenic disease is familial dysbetalipopro-teinemia which is manifested by the abnormal lipoproteinp-VLDL and premature atherosclerotic disease.72 Thesemonogenic hyperiipoproteinemias provide strong supportfor the concept that elevated plasma lipoproteins canaccelerate atherosclerosis. They also show that not allhyperiipidemias are dietary in origin, and if not, dietarychange alone may not normalize plasma lipids in thesedisorders: drug therapy may be required in addition todiet modification.

Most of the remainder of patients with cholesterollevels above the 90-95th percentile probably have poly-genic hypercholesterolemia.6^62 Again, dietary factorsmay affect plasma cholesterol in these patients, but thedominant cause of their hypercholesterolemia seeminglyis the interaction of multiple genes. Thus, drug therapymay be needed to reduce cholesterol levels in these pa-tients to the desirable range.

Finally, consideration should be given to the role ofheredity in raising plasma cholesterol into the 50-90thpercentile of the U.S. population, which probably isabove the desirable range. Polygenic factors likely are atplay here as well, but diet often may be the major factor.In the mildly elevated range, diet probably acts on agenetic foundation, but despite hereditary influence, di-etary change alone may be sufficient to reduce plasmacholesterol to desirable levels.

Diet and CHD

Evidence of several kinds strongly supports the con-cept that the composition and quantity of dietary constit-uents can influence rates of atherogenesis and thusaffect risk for CHD. These data can be reviewed briefly.

Diet and Atherosclerosis in Experimental AnimalsMany studies have shown that diet can induce athero-

sclerosis in animals.3537 As a general rule, the athero-


Dow

nloaded from



genie effect of diet appears to be mediated mainlythrough the plasma lipoproteins. In animals which are re-sistant to diet, cholesterol levels usually are not in-creased. On the other hand, in responsive animal spe-cies, even mild rises in cholesterol concentrations, to therange found in the U.S. population, are associated withdeposition of cholesterol in the arterial wall.73 While theanimals studies in general support the notion that inges-tion of saturated fats and cholesterol can accelerateatherosclerosis, it is important to note that hyper-cholesterolemia induced by diet in most species isassociated with unusual types of lipoproteins. Most typi-cally the increase occurs in cholesterol-rich particles, p-VLDL74'75 or large LDL76 which normally do not occur insignificant quantities in man. These abnormal lipopro-teins may be unusually atherogenic. Nevertheless, de-spite differences between experimental animals andman, the basic concept that dietary factors can induceabnormalities in cholesterol transport which in turn canaccelerate atherosclerosis almost certainly is valid. Note-worthy are investigations in primates showing the ather-ogenic potential of dietary cholesterol and saturated fatsin species not far removed from man.77"83

Epidemiologic StudiesTwo kinds of investigations have been carried out to

determine the influence of diet on CHD. These areepidemiologic surveys and intervention trials. Epidem-iologic surveys and population comparisons may be con-sidered "clinical trials of nature." As such, they havedefinite advantages over man-made trials. First, they en-compass large numbers of subjects, whereas interven-tion trials frequently have too few participants to drawdefinite conclusions; and second, they eliminate theproblem of duration because the population under studyusually has maintained a constant dietary pattern formany years. On the other hand, they are more subject tomisinterpretation because of confounding variables. Ashort summary of the major epidemiologic studies thuswould seem worthwhile.

Geographic Pathology of Atherosclerosis. In 1968 anextensive pathological study reported the extent ofatherosclerosis in approximately 21,000 people in 15countries throughout the world.84 An attempt was madeto correlate the severity of aortic and coronary athero-sclerosis with particular dietary habits of different popula-tions under study. The data of this project revealed a highcorrelation between the estimated level of fat in the dietand the severity of atherosclerosis. A similar associationwas found between the percentage of fat in the diet andcholesterol concentrations in serum, and the latter in turnwas highly correlated with the extent of atherosclerosis.

The Seven Countries Study.17 This cross-populationstudy contrasted dietary composition and energy ex-penditure among men of different countries including theregions of northern and southern Europe, the UnitedStates and Japan. The data showed a high correlationbetween percent of calories as saturated fat and CHDdeaths. A similar relation was noted between saturatedfats and serum cholesterol and between the latter andCHD deaths. Thus, the results are strongly suggestivethat amounts of dietary saturated fats affect the incidenceof CHD, at least in part through their effect on plasmacholesterol. Since the intakes of dietary cholesterol werenot measured in this study, its contribution is unknown.

The Ni Hon-San Study.85'91 This project comparedCHD deaths among Japanese living in Japan, Hawaii

and San Francisco. Intakes of saturated fat calories forthe three populations were estimated to be about 7%,12% and 14%, respectively. With increasing intakes ofsaturated fats, dietary cholesterol and mean bodyweights also rose. Serum cholesterol was higher in pro-portion to the greater intakes of saturated fats and cho-lesterol; they were 12% greater in Hawaii and 21% higherin San Francisco than in Japan. Compared to Japan,death rates from CHD were 1.7 times higher in Hawaiiand 2.8 times greater in San Francisco. Thus, In apopulation containing a high degree of genetichomogeneity, the composition of the diet correlated signi-ficantly with both serum cholesterol and CHD mortality.

Seventh-Day Adventist Study.92'96 This project com-pared CHD mortality rates in Seventh-Day Adventistswho were predominately lacto-ovo-vegetarians with anage-matched group in the general population of Califor-nia. Relatively low levels of serum cholesterol have beenshown repeatedly in this population. Although Seventh-Day Adventists may be protected from CHD in part by anabsence of other risk factors (e.g., lack of smoking), theirlow cholesterol levels, apparently due to dietary habits,probably contribute to their low CHD death rates relativeto the control population.

The Western Electric Study.97 Recently, the results ofanother survey were reported in which diet, serumcholesterol and other variables were examined in 1900middle-aged men. These men were examined at entryinto the study, one year later and, finally, twenty yearslater. Scores summarizing each participant's dietary in-take of cholesterol, saturated fatty acids and polyunsatu-rated fatty acids were calculated according to standardequations relating diet to plasma cholesterol.1314 Therewas a positive association between diet score and serumcholesterol concentration at the initial examination, apositive correlation between change in diet score andchange in serum cholesterol concentration from the initialto the second examination, and a positive associationprospectively between mean base-line diet score andthe 19-year risk of death from CHD. These correlationswere maintained after adjustment for potentially con-founding factors.

These epidemiologic studies are consistent with thetheory that intakes of saturated fats and cholesterol affectincidence of CHD. So are a host of other studies not dis-cussed here.98'113 It must be pointed out that sever-al epidemiologic studies done within the United Stateshave failed to observe significant correlations amongdietary fat, serum cholesterol concentrations and CHDrates114115; however, because of considerable geneticheterogeneity and dietary homogeneity in the U.S.population, it is not surprising that correlations were notfound, and because of limitations of design these nega-tive findings do not necessarily mean that correlations donot exist. It is important to note that the Western ElectricStudy97 was able to detect a correlation between diet andCHD rates within a large group of U.S. men.

Despite the many observations of a relationship be-tween dietary factors and CHD, as described above, thevarious studies have been criticized because all the fac-tors that might affect CHD risk (other than diet) were notheld constant. Therefore, we might consider the resultsof dietary intervention trials to test the hypothesis thatdietary change per se may not affect atherogenesis.

Dietary Intervention TrialsSeveral clinical trials to test whether dietary alteration


Dow

nloaded from



would change CHD risk were started in the 1950s and1960s. Two such studies were carried out in Chicago andNew York using free-living men following a prudentdiet.116"118 These "first generation" trials were inade-quate because of their small sample size and lack of ran-domized control groups. Several other large-scale clinicaltrials have been attempted and while they have limita-tions, the findings are nevertheless of interest. Generally,these trials have tested diets higher in polyunsaturatedfat content than that recommended by the AHA for thegeneral public. Nonetheless, they addressed the ques-tion of whether lowering of plasma cholesterol by dietarymeans can reduce risk of CHD.

One trial involved institutionalized men and women intwo mental hospitals in Helsinki, Finland.119 The studyemployed a crossover design with two periods of about 6years each. A diet low in saturated fats and cholesteroland high in polyunsaturated fats produced a decrease inserum cholesterol and an apparent decrease in CHDmortality rates. A similar trial was carried out in a Minne-sota mental hospital; it also compared a diet low in satu-rated fats and cholesterol and high in polyunsaturatedfats with a control group on a diet high in saturated fats.A preliminary report of this trial120 indicated that rates ofcoronary events were less in men under 40 years on thecholesterol-lowering diet, but they were not different incontrol and diet-treatment groups for older men andwomen.

Another dietary trial was conducted at the VeteransAdministration domiciliary facility in Los Angeles121; adouble-blind experimental design with a randomized con-trol group was used for the study. The results suggestedthat a diet low in saturated fats and cholesterol and highin polyunsaturated fats reduced coronary events; in spiteof the improved experimental design, the study has beencriticized because it was a mixed trial (i.e., partly primaryand partly secondary prevention), had too few subjectsand the subjects were too old at entry (average age 65.5years). Additionally, total death rates in the two ex-perimental groups were the same.

Finally, two intervention trials have been carried out inOslo, Norway. The first, reported in 1966,122 was asecondary prevention trial using a low-cholesterol, high-polyunsaturated fat diet. The results suggested that useof this diet reduced recurrent rates of myocardial infarc-tion. The findings of the second trial have been reportedrecently.123 This primary prevention trial examinedwhether lowering of serum lipids by a polyunsaturated fatdiet and reduction of smoking would decrease the inci-dence of new CHD in 1,232 high-risk patients studied for5 years. At the completion of the trial, myocardial infarc-tion and sudden death were 47% lower in the treatedgroup than in the controls. Statistical analysis revealedthat reduction in CHD in the intervention group was duemainly to a decrease in plasma cholesterol by diet and,to a lesser extent, to a reduction of smoking.

Although all the above studies were flawed in one oranother aspect of experimental design, they were uni-form in reporting a favorable trend toward decreasedCHD risk with cholesterol-lowering diets. Still, a majorproblem was that too few patients were studied for tooshort a time to be certain about this conclusion. In theview of some investigators, a definitive answer to thediet-heart question cannot be obtained until an adequateprimary-prevention trial has been canned out. Indeed, thelack of such a trial is the major criticism against givingrecommendations about diet to the U.S. public. For this

reason, the National Institutes of Health undertook a thor-ough review of the requirements for an adequatetrial,124125 but concluded that a major, large-scale diettrial should not be undertaken. The following were someof the reasons for rejecting the trial:

First, the most direct approach to the diet-heart issuewould be an induction trial, preferably beginning early inlife, to test whether a diet high in saturated fats andcholesterol does indeed yield a higher rate of CHD thanone low in saturated fats and cholesterol, all else con-trolled. Since such a trial obviously is out of the question,the next best approach would be a preventive trial. Toinsure a definite answer, any primary-prevention trialwould be extremely costly. A sample size of at least60,000 participants, and perhaps more, would be re-quired. At least a decade would be needed to completethe study. Furthermore, a diet-heart trial of adequate sizewould be impossible in a closed population — the num-ber of people required are not available — so that theuse of an open population would be necessary. The latterimmediately raises many difficult questions: (1) Howwould a double-blind trial using different diets in a largenumber of people, who are rightly concerned about theirrisk for CHD, be possible? (2) How could other risk fac-tors — blood pressure, smoking, and obesity — be reg-ulated and controlled? (3) How could motivation andadherence be insured for a such a long study? (4) Howwould it be possible to maintain a constant populationunder investigation in a society of great mobility such asthat of the U.S.? Even if these problems could be over-come, the cost in dollars would be prohibitive under thepresent system for funding clinical research. In a word, amajor diet-heart prevention trial probably cannot and like-ly never will be carried out in the U.S., and a decisionabout dietary recommendations will have to be madefrom currently available data. Perhaps the forthcomingdata of the Multiple Risk Factor Intervention Trial(MRFIT)126127 and the Lipid Research Clinic CoronaryPrevention Trial128129 will provide additional support forthe diet-heart theory, but these trials cannot substitute fora major diet-heart study.

Effects of Diet on Plasma Upoproteins in ManResults from experimental animals, epidemiologic sur-

veys and clinical trials all provide highly suggestive evi-dence that diet can affect risk for CHD. Beyond thesestudies, however, are two well-established facts: dietarysaturated fats and cholesterol directly raise the plasmatotal- and LDL-cholesterol, and high total- and LDL-cholesterol contribute directly to atherosclerosis andCHD. While these two facts may not prove a direct linkbetween diet and CHD, the existence of such a link isalmost inescapable. Thus, the combined data implicatingthe U.S. diet to the high incidence of CHD in our societypresent a strong challenge to the dietary habits of mostAmericans. Because of this challenge, the AHA hasdeemed it necessary to make a judgment on the prob-able best diet for the U.S. public. It recognizes that failureto make a recommendation is to condone the currentAmerican diet, i.e., a high intake of total calories, satu-rated fats and cholesterol.

Documentation of AHA Diet Statements

The above comments underlie the rationale for makinga diet recommendation for prevention of CHD. Specificrecommendations also were delineated. These recom-mendations have emerged through years of deliberations


Dow

nloaded from



within the Association and, obviously, universal agree-ment has not existed on every aspect of the final state-ment. The statements, furthermore, have been criticizedfrom outside the Association by several prominent inves-tigators1 30"136 and groups.137'38 Several of these criti-cisms are well considered and deserve a consideredreply.

Nutritional Adequacy

In some countries of the world where a low-fat, high-carbohydrate diet is consumed, malnutrition is common.This has led some to ask whether the modified-fat dietrecommended by the AHA can be made nutritionallyadequate. Extensive analysis of possible menus hasshown that this diet can easily be made adequate inessential nutrients — protein, vitamins and minerals —and need not be deficient in any required nutrient. Clear-ly, a diet with 25-30% of calories as fat need not predis-pose to infection. This level of fat intake is higher than inthe Japanese diet which has only 10-15% of calories asfat; of interest, life expectancy in Japan exceeds that ofthe U.S. at every age.1617 A somewhat better life expec-tancy also is found in Italians who eat less saturated fatsand cholesterol than Americans.1617 In a word, a diet lowin total fat, saturated fats and cholesterol need not benutritionally inadequate. Current cost estimates suggestthat most people can meet the AHA diet recommenda-tions, which are nutritionally adequate, with no more ex-pense than the current U.S. diet which contains sizablequantities of relatively expensive animal products.

The question of nutritional adequacy is particularly im-portant when diets are recommended for children. Todate, the AHA has not singled out children or adoles-cents for specific recommendations, but the current dietstatement stresses the need for diet modification for thewhole family. There is no evidence that the AHA dietwould be deleterious to children or adolescents. A radicalchange in basic nutrients is not proposed. Adequateamounts of high-quality protein can be obtained fromskimmed milk, fish, chicken and a variety of plant foods.These foods can provide sufficient amounts of calcium,vitamins and other nutrients needed for growth. Again,the example of the Japanese diet demonstrates the pos-sibility of nutritional adequacy of a low-fat diet.

Effectiveness of Diet in Plasma LJpid Loweringand Potential Impact on CHD

Since dietary change cannot drastically reduce plasmacholesterol levels, the potential impact of the AHA diet onCHD rates becomes an important issue. If the standardequations derived from metabolic ward investigations areemployed, the proposed AHA diet should decrease plas-ma cholesterol by 30-40 mg/dl as compared to the cur-rent U.S. diet.13"15 A somewhat greater reduction mightbe achieved in obese subjects by weight reduction, andat the same time their HDL levels would be increased.34

The effect of such a change in cholesterol levels on CHDundoubtedly would depend on a variety of factors includ-ing the duration of the change, the age at which it oc-curred and the initial level of plasma cholesterol. Al-though a change in plasma cholesterol of 30 to 40 mg/dlmay have a relatively small impact on risk for CHD wheninitial levels of cholesterol are near 200 mg/dl or below,some reduction in risk might still be possible. For peoplewhose cholesterol is above this desirable range, how-ever, a substantial dietary change should reduce plasma

cholesterol sufficiently to decrease risk of CHD signifi-cantly.

Furthermore, the full impact of dietary change on CHDmay not be predicted from equations derived from short-term diet studies conducted on the metabolic ward. Ufe-long adherence to a diet low in total calories, saturatedfats and cholesterol actually may produce a greaterlowering of cholesterol levels than observed in short-termdietary studies. Evidence for a long-term effect of diet isfound in special groups of people whose mean choles-terol levels are lower than would be predicted fromequations developed on the metabolic ward.13"15 One ex-ample is vegetarians of a Boston commune who eat prin-cipally whole grains, beans and fresh vegetables.139

When plasma cholesterol levels in the Boston groupwere compared to age-matched controls in Framingham,Massachusetts, their levels of total cholesterol and LDL-cholesterol were 31% and 38% lower, respectively, thanthe controls. Another report140 found plasma totalcholesterol and LDL-cholesterol to be 33% and 45% low-er in men and women vegetarians (ages 30-40) living ina Tennessee community compared to age-and sex-matched non-vegetarians In St. Louis. The findings inthese two groups of American vegetarians suggest thatthe much lower cholesterol concentrations in people onvery low intakes of saturated fats and cholesterol (suchas the Japanese), as compared to most Americans, maybe related more to diet than to racial or genetic factors.This lack of genetic influence is also supported by theresults of the Nl-Hon-San study.85"91 Of interest in the di-etary trial at the Los Angeles Veterans Administrationdomiciliary,121 prolonged ingestion of a diet rich inpolyunsaturated fats resulted in gradual but continuousdecline in total plasma cholesterol over several years.Finally, another phenomenon possibly reflecting the long-term effect of diet is the accentuated rise in plasmacholesterol which occurs with aging in the U.S. popula-tion.60 Although this rise may be universal, it is lessmarked in other populations,14 and the pronounced risein Americans could be related to weight gain with agingor to prolonged intakes of saturated fats and cholesterol.

Finally, serious consideration must be given to the pos-sibility that the influence of diet on CHD may not be link-ed only with plasma cholesterol and LDL levels. Otherpoorly understood mechanisms may be involved. Thefindings in the Seven Countries Study are in accord withan independent effect of diet. In table 1 1 4 2 the 10-yearprobability of CHD death by serum cholesterol level inCHD-free men at age 50 are compared for U.S. railroadworkers, Finnish men, and men from the Mediterraneanregion (Greece, Yugoslavia and Italy). At similar cho-lesterol levels, risk was about the same in U.S. and Fin-nish men, but rates were much lower in Mediterraneanmen. In the former two groups, intakes of fats averaged40% of total calories. Saturated fats and cholesterol con-stituted a relatively large proportion of the calories fromfat. In Mediterranean men, intakes of total fats andcholesterol were, lower, and monounsaturates were themajor type of fat ingested. A discrepancy between plas-ma cholesterol and CHD risk might have several ex-planations. Obviously, factors unrelated to lipid andlipoprotein metabolism could affect the relative risk forCHD among the different countries. On the other hand, asingle measurement of plasma cholesterol may not ade-quately reflect CHD risk, or saturated fats and cholesterolmight affect lipoprotein-arterial wall interactions in a waynot revealed by estimation of total plasma cholesterolalone.143144


Dow

nloaded from



Treatment of High Plasma Cholesterol andOther Risk Factors

Since atherosclerosis is a multifactorial disease, it canbe asked whether the most effective approach to preven-tion of CHD might be to concentrate on other risk factors(e.g., smoking and hypertension). This is a reasonablequestion, but evidence of several types supports the con-cept that a relatively high level of LDL is a crucial factorin atherogenesis; furthermore, epidemiological studiessuggest that other risk factors may not contribute greatlyto CHD until the concentration of LDL has reached a cer-tain level.17145 Therefore, if LDL concentrations can bereduced through dietary means, the risk from other fac-tors might be lessened significantly. This is not to say thatmodification of other risk factors is not important and,from a practical point of view, their elimination could havea profound influence on risk in the U.S. population.

Practicality for U.S. Population

One criticism of the AHA diet has been that it would bepoorly accepted by the U.S. public. This criticism may notbe valid because the American diet is changing.18 In-takes of cholesterol and possibly saturated fats have de-creased; in addition, ingestion of polyunsaturated fatshas increased, although not to levels above the limitsproposed by the AHA. Many individuals have consciouslychosen to alter their diets to one low in saturated fats andcholesterol, and these changes may be partly responsi-ble for the apparent reduction in plasma cholesterol ofthe American population over the past few years.

The AHA contends that, at the very least, the publicshould be given the option of choosing foods consonantwith its recommended diet. These foods should be readi-ly available in both groceries and restaurants. Labeling offoods is strongly advocated so that the public can moreeasily choose the desired composition. Many peopleseemingly desire to alter their eating habits but have in-sufficient knowledge of dietary principles to accomplishthe change. Thus, before dietary recommendations canhave a great impact on eating patterns, they must beassociated with a program of public education.

Weight Reduction vs Change in Diet Composition

Although the debate continues whether American peo-ple should change the composition of their diets, the pru-dence of maintaining body weight in the desirable rangehas received almost universal acceptance. Indeed, a re-duction in body weight may decrease risk for CHDthrough (1) reduction of total plasma cholesterol (espe-cially VLDL-cholesterol), (2) an increase in HDL-cholesterol and (3) a reduction in blood pressure. If calor-ic restriction is carried out by decreasing the intake ofsaturated fats and cholesterol, LDL-cholesterol alsoshould fall in most people. This reduction of LDL can beachieved only if saturated fats and cholesterol are de-creased and will not occur with some reducing diets thatstress significant quantities of animal fats at the expenseof other constituents; indeed, the latter actually may raiseLDL.

Clearly, the AHA recommends weight reduction inobese subjects and the maintenance of desirable weightin general. However, many people whose weights are inthe desirable range still have a relatively high LDL, and inthese only a change in diet composition (e.g., a decreasespecifically in saturated fats and cholesterol) can reduce

LDL concentrations toward an acceptable range. For thisreason the AHA does not accept the premise that weightcontrol should be the only dietary recommendation forthe general public.

Dietary Change in Low-risk Groups

A common criticism of the AHA position is that univer-sal dietary change would lead to "treatment" of manypeople who are already at low risk for CHD. Severalpoints can be made in response to this criticism. First,the possibility of a curvilinear relation between plasmacholesterol and CHD risk is still hypothetical, and a re-duction in plasma cholesterol through dietary modifica-tion may further decrease the risk for CHD even in thosewho already have cholesterol levels in the so-called"ideal range."60 For example, CHD rates apparently arelower in countries with severely restricted intakes of satu-rated fats and cholesterol than in people of the U.S.population whose plasma cholesterol is in the first quin-tile (i.e., less than 194 mg/dl). This difference could bedue to lower mean levels of cholesterol in these otherpopulations as compared to the first quintile,7 but as indi-cated above an excess intake of cholesterol and satu-rated fats may contribute to atherosclerosis throughmechanisms not revealed by plasma cholesterol levelalone in the fasting patient.

Second, unless eating habits of Americans are widelychanged, people in the higher-risk categories probablywill not alter their eating patterns. The important pointshould be made that there is nothing sacred about thecurrent American diet high in fat and total calories. TheAHA diet, which emphasizes variety (especially fruits,vegetables, whole grains, legumes and animal proteinsources such as chicken, fish, skim milk and egg whites),can certainly be as palatable as one containing largequantities of fat-rich meats, egg yolks and dairy fats. Asthe U.S. public is becoming more diet-conscious, there isan increasing interest both in health value and eatingpleasure that can be derived from diets patterned alongMediterranean and Far Eastern lines. In addition to thevaluable AHA cookbook, several others have recentlybeen published which offer many tasty recipes that are inline with the AHA recommendations.

Third, in the absence of universal screening for CHDrisk, it is impossible to identify specific individuals whomight benefit most from the AHA diet. Therefore, sinceCHD risk may be reduced to some degree even in thosein low-risk catagories, the diet can be recommended toall who have not had careful screening for plasma lipidsand other risk factors. Finally, cholesterol levels usuallyrise with age, and even if screening is carried out at anappropriately early age, future cholesterol concentrationscannot be predicted with accuracy. Indeed, adherence toa diet low in cholesterol and saturated fats may actuallydiminish the increase in plasma cholesterol that occurscommonly with aging.141 There is no evidence that therecommended changes could result in an elevation inplasma cholesterol.

AHA Diet and Individual Variability of Response

Another common criticism of the recommendation foruniversal dietary alteration is that changes in plasmalipids vary from person to person. In other words, somepeople appear more diet-responsive than others. Thosewho do not respond significantly may not derive any


Dow

nloaded from



direct benefit from dietary alteration. This is a logical crit-icism, but the extent to which it holds in man has notbeen determined. Variability in responsiveness can bemarked in some animals fed cholesterol.146'147 Some de-gree of variability almost certainly exists for man as well.However, carefully controlled studies on metabolic wardssuggest that variability is less than might be anticipatedfrom animal studies. For instance, responses to dietarycholesterol over a considerable range of cholesterol in-takes in normal men are reasonably constant.30 Further,exchange of polyunsaturated fats for saturated fats pro-duces a fairly uniform alteration in cholesterol concentra-tions in the metabolic ward.11"15 These findings suggestthat some of the reported variability in outpatient studiesmay be due to differences in adherence to diet. Obvious-ly, more investigation is needed to define the extent towhich individual responses to diet differ. But even if thedegree of change in plasma cholesterol varies consider-ably, this does not rule out the validity of recommendinggeneralized change in diet to bring about a net loweringof plasma cholesterol in the whole population.

AHA Diet and Genetic Hyperiipidemias

The question has been raised whether the publichealth approach of the AHA may not overlook or inade-quately treat the large number of people in the U.S.population who have genetic hyperiipidemias. Thesepeople are at high risk for CHD,6162 and dietary changealone may be inadequate to bring their plasma lipids todesirable levels. This is an important point, and it bringsup the issue of how much emphasis should be put on thesearch for patients with significant risk factors. The AHAhas never advocated screening the total population on auniversal and recurring basis because of impracticality.On the other hand, the attempt has been made to edu-cate the medical community on the importance of check-ing plasma lipids in individuals suspected of being atgreater risk (i.e., patients who have existing CHD, hyper-tension, diabetes mellitus, obesity, who are smokers orhave a family history of CHD or hyperlipidemia). Thequestion of who should be screened for hyperlipidemiaand under what conditions is still a subject of debatewithin the Association, but the need to seek out high-riskpatients is clearly recognized. The advantage of identify-ing patients with hyperlipidemia should be twofold. First,if a patient is found with a distinct elevation in plasmacholesterol, he can be informed of his increased risk andbe urged to correct the abnormality. (It is also importantto find other family members who may have hyper-cholesterolemia.) Second, if a person clearly has a lowcholesterol he can be reassured not to be inappropriatelyconcerned. Theoretically, screening for lipid and lipopro-tein levels has the advantage of allowing for individual-ized modification of any abnormality. However, from apractical standpoint, detection or screening would not besufficient as the only means of controlling the high preva-lence of relatively elevated cholesterol levels in our soci-ety. First, the medical community is not organized to car-ry out universal screening for hypercholesterolemia.Second, and of great importance, a single determinationof plasma cholesterol is not sufficient to define an indi-vidual's status with respect to plasma lipids and lipopro-teins; instead, multiple determinations are needed atseveral different ages to obtain an accurate assessmentof risk. Because of biological and methodologicalvariability, many patients would be misled by a single de-

termination into believing that their plasma cholesterolwas in a "safe" range when, in fact, a cholesterol levelintegrated over a long period of time might not be. Andthird, even if patients with "dangerous" and "safe" levelsof cholesterol could be differentiated, no means are cur-rently available to adequately counsel large numbers ofpatients on proper diet and to provide long-term nutrition-al guidance. Therefore, the public health approach of dietmodification is needed along with case detection.Through the combination of both approaches, maximalmodification of abnormal plasma lipids should beachieved.

Polyunsaturated Fats and AHA Diet

A common misconception is that the AHA currentlyrecommends the intake of large quantities of polyunsat-urated fats. It is true that the recommendation is toincrease the mean intake for the U.S. public from itspresent 5-6% to an upper limit of 10%. The majorreasons for the recommended increase are twofold: (1)polyunsaturated fats have an inherent cholesterol-lowering property and (2) they provide a preferablesource of fat calories to cholesterol-containing saturatedfats. On the other hand, the AHA has been reluctant toadvocate a marked increase in polyunsaturated fats (i.e.,to 15-20% of total calories) for several reasons. First, nolarge population has consumed such large quantities ofpolyunsaturated fats for many years with demonstratedsafety. Second, high-fat diets promote weight gain. Third,polyunsaturated fats may increase the risk for gallstonesin some people.148149 Fourth, these fats have beenshown to have co-carcinogenic potential in experimentalanimals.150"152 And fifth, they can alter the composition ofcell membranes.153 Thus, it may be prudent for the U.S.public to avoid very large amounts of polyunsaturatedfats for the present time.

AHA Diet vs Very Low Fat Diets

Criticism of the AHA diet has also come from thosewho believe Americans should eat a very low fat diet(i.e., a fat intake of 10-20% of total calories). They citethe low incidence of CHD in populations in whom total fatintake is very low (e.g., Japanese in Japan, rural Guate-malans, Tarahumara Indians, etc.), and they concludethat the AHA diet is too liberal in total fat. However, acareful review of epidemiological studies, particularly theSeven Countries Study,17 suggests that the crucial ath-erogenic factors in the diet are saturated fats andcholesterol and not percent total fat. For example, CHDdeaths are very low in Greece and adjacent areas de-spite an intake of fat of 30-35% of calories17; in this re-gion, however, monounsaturated fatty acids and notsaturated fats make up the bulk of the fat intake. InCrete, for instance, fat intake is 40% of total calories, butsaturated fats make up only about 8%, and CHD mortal-ity rates are extremely low. Therefore, the existing datasuggest that the AHA diet with recommendations to sev-erely curtail saturated fats and cholesterol shouldachieve essentially the same results as a very low intakeof total fat. It is unlikely that the American public wouldaccept such a marked departure from its usual fat intake,and it probably is not necessary.

AHA Diet, Low Plasma Cholesterol and Cancer

During the past few years several epidemiological


Dow

nloaded from



studies have suggested an inverse relationship betweennaturally occurring serum cholesterol and the incidenceof malignant neoplasmas and other causes of death. Thefirst report of this type154 indicated a significantly loweraverage level of blood cholesterol among men in Eng-land who died of cancer of the colon. Following this re-port, a significant inverse relationship between serumcholesterol and overall mortality was reported for NewZealand Maoris, a group of Polynesians who haveadopted a Western life style.155 A subsequent study fromEngland of 18,000 men, aged 40-64 years, again notedan increased death rate from cancer and other causes inmen with lowest plasma cholesterol156; however, this in-verse association was limited to the first two years of fol-low-up; afterward, total mortality and cholesterol levelswere correlated evenly and positively. A similar phe-nomenon of early death in low-cholesterol subjects notconfirmed in long-term follow-up has been reported in theParis Prospective Study of Coronary Heart Disease.157

In a study of 3,102 people in Evans County, Georgia,followed for 14 years, new cancer cases had signifi-cantly lower mean levels of serum cholesterol at the timeof initial sampling than did the noncancer population; thisrelationship was stronger for men than women and wasobserved for various sites of cancer and cell types. In areport from Framingham, Massachusetts,159 691 casesof cancer were documented in 5,000 subjects followedfor 24 years. During this period, 691 cases of cancerwere observed and significant associations of differenttypes of cancer were observed with cigarette smoking,alcohol use, education, height-weight and parity. Further-more, serum cholesterol levels were found to be inverse-ly correlated with the incidence of colon cancer and pos-sibly other sites in men (but not women). In another re-port of this type from Malmo, Sweden, a follow-up of10,000 men revealed that deaths occurred in two peaks,one at the higher end of the serum cholesterol distribu-tion but the other at the lower end. A significant portion ofdeaths in the low-cholesterol range appeared to be re-lated to alcohol (acute intoxication, accidents and alco-hol-induced diseases — cirrhosis of the liver, esophagealvarices and pancreatitis).

Despite these reports, not all epidemiologic studieshave reported a significant inverse correlation betweenplasma cholesterol levels and cancer.160 Of particular in-terest for the U.S. population, the large Pooling Project7

did not show the low cholesterol-cancer association.Therefore, the available epidemiologic data are tentativeat best, and more prospective studies will be requiredbefore a site-specific relationship between low cholester-ol levels and cancer can been proven.

Furthermore, several investigators have held for sometime that high-fat, high-cholesterol, low-fiber diets alsomay play a role in the etiology of colon cancer and othermalignancies.161 In most industrialized countries in whichdiets contain large amounts of fat and cholesterol, theincidence of colon cancer is greater than in countrieswhere most people ingest a diet low in cholesterol and fatand high in fiber. Indeed, international comparisons sug-gest a significant relationship between mortality rates ofCHD and colon cancer.162 Thus, when all the epidem-iologic studies are taken into consideration, high intakesof fat and cholesterol do not seem to be protectiveagainst cancer; if anything, they appear to heighten therisk for malignancy. This has led to the suggestion thathigh intakes of fat and cholesterol may be associatedwith an increased influx of cholesterol and its conversion

products (bile acids) into the large intestine,163164 andthese steroids, or their secondary bacterial conversionproducts, may be carcinogenic.

Regardless of the explanation for the apparent asso-ciation between naturally occurring low plasma cholester-ol and cancer, there are no data arguing that changingthe diet toward that recommended by the AHA wouldheighten the risk for cancer. It might be noted that theauthors of the clinical trial carried out in the Veterans Ad-ministration Domiciliary in Los Angeles reported that thetherapeutic diet, which was rich in polyunsaturated fats,was associated with a greater number of cancer deathsthan the control diet. The results have been widelyquoted, but the statistical limitations have been stressedby many. In a review of the cancer experience from fourother clinical trials using similar cholesterol-loweringdiets,166 the conclusion was reached that diets rich inhigh-polyunsaturated fats did not influence cancer risk.Finally, it should be emphasized again that the AHA doesnot recommend high-polyunsaturated fat diets similar tothose used in the above clinical trials for the generalpublic.

AHA Diet and HDL

Recent reports suggest that cholesterol-lowering dietscan cause a decrease in plasma HDL as compared todiets containing large amounts of saturated fats and cho-lesterol.167168 To date, it has not been tested whether theAHA-recommended diet will alter HDL. However, the ma-jor causes of low HDL in our society seem to be obesityand sedentary life-styles. If desirable weight can beachieved along with an appropriate exercise program,these should cause a rise in the HDL concentration. Thisrise in HDL should offset any tendency for a lowering ofHDL by reducing saturated fats and cholesterol or in-creasing polyunsaturates.

The role of diet on the metabolism of HDL is poorlyunderstood, but it is a subject of much current interest.Several subspecies of HDL have been identified, and theimpact and clinical significance of diet on each remainsto be determined. The preliminary observations thatfeeding of saturated fats and cholesterol may raise HDL,which appears from epidemiological studies to be associ-ated with a lesser risk of CHD, is a paradox that needs tobe resolved. Nevertheless, it seems likely that any lower-ing of HDL that might result from cholesterol-loweringdiets would be offset by a similar reduction of LDL con-centrations; this is in accord with the observations dis-cussed above that populations with low LDL concentra-tions and decreased rates of CHD frequently have lowlevels of HDL as well.

AHA Diet for Children

The current AHA statement is consistent with the con-cept that dietary modifications should begin in childhood,but it does not elaborate. For this reason, a more specificstatement is under consideration by the Nutrition Com-mittee. More attention is now being focused on identify-ing risk factors for coronary heart disease in children sothat preventive measures can be introduced at a stagewhen they are most likely to influence the underlyingpathologic process. Although it is not known whether di-etary modification in children will alter the incidence ofCHD in later life, evidence supporting the benefit oflowering plasma cholesterol levels is strong. A majorassociated benefit is the possibility that dietary habits


Dow

nloaded from



learned in childhood may persist into adult life. The re-duction of total dietary energy in the obese childhoodpopulation is also an advantage. The recommended AHAdiet does not pose any danger of poor nutrition for chil-dren. It is not a radical diet and is consistent with a goodnutritional status.

AHA Diet and Sodium

The current diet statement also contains the recom-mendation that dietary sodium should be reduced. Thereason for decreasing sodium intake obviously relates tothe problem of hypertension, a known risk factor forCHD. A reduction in dietary sodium might accomplisheither or both of two aims: (1) prevention of hypertensionin susceptible individuals or (2) reduction in the averagelevel of blood pressure in the general population.Obviously, the relatively high consumption of salt by theU.S. public does not cause hypertension in the majorityof people, and it can be asked whether a universal andmarked change in sodium intake is needed for the pre-vention of overt hypertension in the 10-15% of thepopulation who are salt-sensitive. On the other hand, therelation between hypertension and CHD risk seems to becontinuous over a wide spread of blood pressures, evendown into the "normal" ranges.7 If the average level ofblood pressure in the general population could be re-duced by a universal decrease in salt intake, the overallrisk for CHD in Americans should be reduced. How muchdecrease in dietary sodium is needed to produce a signif-icant lowering of mean blood pressure, unfortunately, isnot known; thus, a precise limit of sodium intake has notbeen recommended. The Nutrition Committee recentlyhas been considering a specific proposal about sodium,but it recognizes that the question of whether an effortshould be made to reduce the intake of sodium for thegeneral public is an unsettled matter.

IV. Strategy of AHA Position onPlasma Lipid Modification

Based on the above considerations, a reasonableapproach to modification of high plasma lipids would betwofold: to recommend a generalized diet change forhealthy Americans and to seek out and treat individualpatients with unusually high levels of lipids or other riskfactors. For the first time, the AHA has taken the positionthat a diet recommendation for the healthy U.S. popula-tion is warranted. It should be emphasized that the rec-ommended diet is not radical, and there is little likelihoodit would be nutritionally inadequate or harmful. There isalmost uniform agreement that caloric restriction toachieve desirable weight is beneficial. There are no goodreasons related to health why Americans should not con-sume a diet containing less total fat, saturated fats andcholesterol. Since no harm can be visualized, and be-cause it may lessen the risk for CHD, a change to such adiet would seem prudent.

The second strategy is to find and treat patients withknown risk factors and to recommend modification ofthese factors. This strategy has followed two lines: (1) toeducate the public about risk factors and to urge them tobe checked by their physicians and (2) to educate physi-cians on the importance of screening for risk factors. Thechecking for risk factors has two advantages. First, it en-ables the physician to discover patients at high risk whocan be treated. For patients with genetic hyperlipidemia,

the AHA diet for healthy Americans may not be sufficientto bring plasma lipids into the desirable range. Such pa-tients may have to be treated with drugs or by more strin-gent dietary therapy. The AHA will soon publish a revisedprogram for dietary therapy of hyperiipidemia. This pro-gram will make use of the diet recommended for thegeneral public as the first phase of this therapy, but inpatients with significant hyperlipidemia, second and thirdphases will progress to more and more restrictions of fatand cholesterol intakes.

A second reason for measuring plasma lipids is thepossible identification of individuals with low lipid levels. Ifthese persons are of normal weight and have no otherrisk factors, they can be reassured that a significantchange in life-style probably is not necessary. Neverthe-less, before giving this reassurance, the physician shouldcaution the patient that cholesterol levels tend to rise withage, and future checks are advisable. There is also thepossibility that the link between diet and CHD may not bereflected entirely by plasma lipid concentrations, particu-larly at a single point in time. The data shown in table 1suggest that total emphasis should not be put on mea-sured plasma cholesterol.

To date, the AHA has not recommended or supportedwidespread screening of the general population forhyperlipidemia. This would be both costly and inefficient,particularly because of a lack of a uniform mechanism fortreating the multitude of patients who would be detected.Perhaps at a future time when the relationships betweenplasma lipoproteins and atherosclerosis are better under-stood, consideration may be given to how best to seekindividual cases for specific therapy. In the meantime,however, the AHA has taken the position that educationof the general public about hygienic measures to reduceCHD risk, and of physicians about detection and mod-ification of risk factors, is the most appropriate approachto the prevention of CHD.

References

1. Page IH, Stare FJ, Corcoran AC, Pollack H, WilkinsonCF. Atherosclerosis and the fat content of the diet. Circula-tion 16:163, 1957

2. Central Committee for Medical and Community Programof the American Heart Association. Dietary fat and its re-lation to heart attacks and strokes. Circulation 23:133, 1961

3. American Heart Association. Diet and Heart Disease.Dallas, Amercan Heart Association, 1965

4. American Heart Association. Diet and Heart Disease. Dal-las, American Heart Association, 1968

5. American Heart Association. Diet and Coronary Heart Dis-ease. Dallas, American Heart Association, 1973

6. American Heart Association. Diet and Coronary Heart Dis-ease. Dallas, American Heart Association, 1978

7. Pooling Project Research Group. Relationship of bloodpressure, serum cholesterol, smoking habit, relative weightand ECG abnormalities to incidence of major coronaryevents: final report of the pooling project. J Chronic Dis31:201, 1978

8. Qordo T, Castelll WP, HJortland MC, Kannel WB, DawberTR. High density lipoprotein as a protective factor againstcoronary heart disease. The Framlngham Study. Am J Med62:707, 1977

9. Wilson PW, Garrison RJ, Castelll WP, Felnlelb M, McNa-mara PM, Kannel WB. Prevalence of coronary heart dis-ease In the Framlngham Offspring Study: study of lipopro-tein cholesterols. Am J Cardiol 46:649, 1980

10. Miller GJ, Miller NE. Plasma-high-density-lipoproteln con-centration and development of ischaemic heart-disease.Lancet 1:16, 1975


Dow

nloaded from



11. Ahrens EH Jr, Insull W Jr, Blomstrand R, Hlrsch J, Tsal-tas TT, Peterson ML. The influence of dietary fats onserum-lipid levels in man. Lancet 1:943, 1957

12. Keys A, Anderson JT, Grande F. Prediction of serum-cholesterol responses of man to changes in fats in the diet.Lancet 2:959, 1957

13. Keys A, Anderson JT, Grande F. Serum cholesterol re-sponse to changes in the diet. II. The effect of cholesterol inthe diet. Metabolism 14:759, 1965

14. Keys A, Anderson JT, Grande F. Serum cholesterol re-sponse to changes in the diet. IV. Particular saturated fattyacids In the diet. Metabolism 14:776, 1965

15. Hegsted DM, McGandy RB, Myers ML, Stare FJ. Quan-titative effects of dietary fat on serum cholesterol in man. AmJCIinNutr, 17:281, 1965

16. Stamler J: Population studies. In Nutrition, Lipids, andCoronary Heart Disease, vol 1, edited by Levy Rl, RifkindBM, Dennis BH, Ernst N. New York, Raven Press, Pubs.,1979, p 72

17. Keys A (ed). Coronary heart disease in seven countries.Circulation 41: (suppl I): 1-1, 1970

18. Keys A. Mortality and coronary heart disease in theMediterranean area. Proceedings of the II International Con-gress on the Biological Value of Olive Oil (Torremolinos,Spain), p 281, 1976

19. Watklns DM, Froeb HF, Hatch FT, Gutman AB. Effects ofdiet in essential hypertension. II. Results with unmodifiedKempner rice diet in 50 fully hospitalized patients. Am J Med9:441, 1950

20. Ahrens EH Jr, Hirsch J, Oette K, Farquhar JW, Stein Y.Carbohydrate-induced and fat-induced lipemia. Trans AssocAm Physicians 74:134, 1961

21.Antonls A, Bersohn I. The Influence of diet on serum-tri-glycerldes in South African white and Bantu prisoners. Lan-cet 1:3, 1961

22. Little JA, McGuIre V, Derksen A. Available carbohydrates.In Nutrition, Lipids, and Coronary Heart Disease, vol 1,edited by Levy Rl, Rifkind BM, Dennis BH, Ernst N. NewYork, Raven Press, Pubs., 1979, p 119

23. Beverldge JMR, Connell WF, Haust HL, Mayer GA. Di-etary cholesterol and plasma cholesterol levels in man. CanJ Biochem Physiol 37:575, 1959

24. Beverldge JMR, Connell WF, Mayer GA, Haust HL Theresponse of man to dietary cholesterol. J Nutr 71:61, 1960

25. Brown HB, Page IH. Effect of polyunsaturated eggs onserum cholesterol. J Am Diet Assoc 46:189, 1965

26. Connor WE, Hodges RE, Bleller RE. Effect of dietarycholesterol upon serum lipids in man. J Lab Clin Med57:331, 1961

27. Connor WE, Hodges RE, Bleller RE. The serum lipids inmen receiving high cholesterol and cholesterol-free diets. JClln Invest 40:894, 1961

28. Connor WE, Lin DS. The intestinal absorption of dietarycholesterol by hypercholesterolemic (type II) and normo-cholesterolemic humans. J Clin Invest 53:1062, 1974

29. Keys A, Mlckelsen O, Miller EVO, Chapman CB. The rela-tion in man between cholesterol levels in the diet and inblood. Science 112:79, 1950

30. Mattson FH, Erfckson BA, Kllgman AM. Effect of dietarycholesterol on serum cholesterol in man. Am J Clin Nutr25:589, 1972

31. Friend B, Page L, Marston R. Food consumption patternsin the United States: 1909-13 to 1976. In Nutrition, Lipids,and Coronary Heart Disease, vol 1, edited by Levy Rl, Rif-kind BM, Dennis BH, Ernst N. New York, Raven Press,Pubs., 1979, p 489

32. U.S. Department of Agriculture, Economic ResearchService. Food consumption, prices, expenditures. Agr EconRep (suppl): 1975

33. Olefsky J, Reaven GM, Farquhar JW. Effects of weight re-duction on obesity. Studies of lipid and carbohydrate metab-olism In normal and hyperlipoproteinemic subjects. J Clin In-vest 53:64, 1974

34. Wolf RN, Grundy SM. Effects of caloric restriction on plas-

ma lipids and lipoproteins. Clin Res 28:55A, 198035. Antlschkow N. Experimental arteriosclerosis in animals. In

Arteriosclerosis, edited by Cowdry EV. New York, MacmillanPublishing Co., Inc., 1933

36. Katz LN, Stamler J. Experimental Arteriosclerosis. Spring-field, Illinois, Charles C. Thomas, 1953

37. Wlssler RW. Development of the arteriosclerotic plaque. InThe Myocardium: Failure and Infarction, edited by Braun-wald E. New York, H P Publishing Co., Inc., 1974, p 155

38. Carlson LA, Bottlger LE. Ischaemic heart-disease in rela-tion to fasting values of plasma triglycerides and cholesterol.Stockholm Prospective Study. Lancet 1:865, 1972

39. Friedman GD, Klatsky AL, Slegelaub AB, McCarthy N.Kaiser-Permanente epidemiologic study of myocardial in-farction. Study design and results for standard risk factors.Am J Epidemiol 99:101, 1974

40. Goldbourt U, Medalle JJ, Neufeld HN. Clinical myocardialinfarction over a five-year period. III. A multivariate analysisof incidence, the Israel Ischemic Heart Disease Study. JChronic Dis 28:217, 1975

41. Klelnbaum DG, Kupper LL, Cassel JC, Tyroler HA. Multi-variate analysis of risk of coronary heart disease in EvansCounty, Georgia. Arch Intern Med 128:943, 1971

42. Kozarevlc D, Plrc B, Ratlc Z, Dawber TR, Gordon T,Zukel WJ. The Yugoslavia Cardiovascular Disease Study. II.Factors in the incidence of coronary heart disease. Am JEpidemiol 104:133, 1976

43. MeGee D, Gordon T. The results of the Framingham Studyapplied to four other U.S.-based epidemiologic studies ofcardiovascular disease. In The Framingham Study: AnEpidemiological Investigation of Cardiovascular Disease,edited by Kannel WB, Gordon T. Washington, DC, U S GovtPrinting Off, 1976, Section 31

44. Miller NE, Thelle DS, Ferde OH, M|6s OD. The TromsdHeart-Study. High-density llpoprotein and coronary heart-disease: a prospective case-control study. Lancet 1:965,1977

45. Pelkonen R, Nlkklla EA, Kosklnen S, Penttlnen K, SarnaS. Association of serum lipids and obesity with cardiovascu-lar mortality. Br Med J 2:1185, 1977

46. Robertson TL, Kato H, Rhoada GG, Kagan A, Marmot M,Syme SL, Gordon T, Worth RM, Belsky JL, Dock DS,Mlyanlshi M, Kawamoto S. Epidemiologic studies of coro-nary heart disease and stroke in Japanese men living inJapan, Hawaii and California. Incidence of myocardial infarc-tion and death from coronary heart disease. Am J Cardiol39:239, 1977

47. Rose G, Reid DD, Hamilton PJS, McCartney P, Keen H,Jarrett RJ: Myocardial ischaemia, risk factors and deathfrom coronary heart disease. Lancet 1:105, 1977

48. Rosenman RH, Brand RJ, Shortz Rl, Friedman M. Multi-variate prediction of coronary heart disease during 8.5 yearfollow-up in the Western Collaborative Group Study. Am JCardiol 37:903, 1976

49 Tlbblln G, Wllhelmsen L, Werko L. Risk factors formyocardial infarction and death due to ischemic heart dis-ease and other causes. Am J Cardiol 35:514, 1975

50. Gordon T, Castelll WP, Hjortland MC, Kannel WB, Daw-ber TR. Predicting coronary heart disease in middle-agedand older persons. The Framingham Study. JAMA 238:497,1977

51. Nalto HK, Greenstreet RL, David JA, Sheldon WL, ShlreyEK, Lewis RC, Proudfrt WL, Gerrrty RG. HDL-cholesterolconcentration and severity of coronary atherosclerosis deter-mined by cine-angiography. Artery 8:101, 1980

52. Jenkins PJ, Harper RW, Nestel PJ. Severity of coronaryatherosclerosis related to llpoprotein concentration. Br JMed 2:388, 1978

53. Rhoads GG, Gulbrandsen CL, Kagan A. Serum lipopro-teins and coronary heart disease in a population study ofHawaii Japanese men. N Engl J Med 294:293, 1976

54. Tyroler HA, Hames CG, Krishan I, et al. Black-white differ-ences in serum lipids and lipoproteins in Evans County. PrevMed 4:541, 1975


Dow

nloaded from



55. Connor WE, Cerquelra MT, Connor RW, Wallace FIB,Mallnow MR, Casdorph HR. The plasma lipids, lipopro-teins, and the diet of the Tarahumara Indians of Mexico AmJ Clin Nutr 31:1131, 1978

56. Hulley SB, Rosenman RH, Bawol RD, Brand RJ. Epi-demiology as a guide to clinical decisions. The associationbetween triglyceride and coronary heart disease. N Engl JMed 302:1383, 1980

57. Myers LH, Phillips NR, Havel RJ. Mathematical evaluationof methods for estimation of the concentration of the majorlipid components of human serum lipoproteins. J Lab ClinMed 88:491, 1976

58. Phillips NR, Havel RJ, Kane JP. Levels and interrela-tionships of serum and lipoprotein cholesterol and triglyc-erides. Association with adiposity and the consumption ofethanol, tobacco, and beverages containing caffeine. Arte-riosclerosis 1:13, 1981

59. Elsenberg S. Type III hyperlipoproteinaemia. Clin Endocri-nolMetabol 2:111, 1973

60. Stamler J. Public health aspects of optimal serum lipid-lipoprotein levels. Prev Med 8:733, 1979

61. Goldstein JL, Albers JJ, Schrott HG, Hazzard WR, Bler-man EL, Motulsky AG. Plasma lipid levels and coronaryheart disease in adult relatives of newborns with normal andelevated cord blood lipids. Am J Hum Genet 26:727, 1974

62. Goldstein JL, Schrott HG, Hazzard WR, Blerman EL,Motulsky AG. Hyperlipidemia in coronary heart disease. II.Genetic analysis of lipid levels in 176 families and delinea-tion of a new inherited disorder, combined hyperlipidemia. JClin Invest 52:1544, 1973

63. Greten H, Wagner M, Schettler G. Early diagnosis and inci-dence of familial type II hyperiipoproteinemia. Analysis ofumbilical cord blood from 1323 newborns. Dtsch MedWochenschr 99:2553, 1974

64. Tsang RC, Fallat RW, Glueck CJ. Cholesterol at birth andage 1: comparison of normal and hypercholesterolemicneonates. Pediatrics 53:458, 1974

65. Tsang RC, Glueck CJ, Fallat RW, Mellles M. Neonatalfamilial hypercholesterolemia. Am J Dis Child 129:83, 1975

66. Langer T, Strober W, Levy Rl. The metabolism of low den-sity lipoprotein in familial type II hyperiipoproteinemia. J ClinInvest 51:1528, 1972

67. Relchl D, Simons LA, Myant NB. The metabolism of low-density lipoprotein in a patient with familial hyper beta hpo-proteinaemia. Clin Sci Mol Med 47:635, 1974

68. Bllhelmer DW, Stone NJ, Grundy SM. Metabolic studies infamilial hypercholesterolemia. Evidence for a gene-dosageeffect in vivo. J Clin Invest 64:524, 1979

69. Brown MS, Faust JR, Goldstein JL. Role of the low densi-ty lipoprotein receptor in regulating the content of free andesterified cholesterol in human fibroblasts. J Clin Invest55:783, 1975

70. Brown MS, Goldstein JL. Familial hypercholesterolemia: agenetic defect in the low-density lipoprotein receptor. N EnglJ Med 294:1386, 1976

71Brunzell JD, Schrott HG, Motulsky AG, Blerman EL.Myocardial infarction In the familial forms of hypertriglyc-erldemia. Metabolism 25:313, 1976

72. Fredrlckson DS, Goldstein JL, Brown MS. The familialhyperllpoproteinemias. In The Metabolic Basis of InhentedDisease, ed 4, edited by Stanbury JB, Wyngaarden JB,Fredrickson DS. New York, McGraw-Hill Book Co., 1978,Chap 31, pp 604-655

73. Armstrong ML, Warner ED, Connor WE. Regression ofcoronary atheromatosis in rhesus monkeys. Circ Res 27:59,1970

74. Mahley RW, Holcombe KS. Alterations of the plasmalipoproteins and apoproteins following cholesterol feeding inthe rat. J Upld Res 18:314, 1977

75. Dletschy JM, Gotto AM Jr, Ontko JA. Alterations in plas-ma lipoproteins induced by cholesterol feeding in animalsincluding man. In Disturbances in Lipid and LipoproteinMetabolism, edited by Dietschy JM. Baltimore, Williams &Wilkins Co., 1978, chap 2, pp 181-197

76. Rudel LL, Pitts LL II, Nelson CA. Characterization of plas-ma low density lipoproteins on nonhuman primates fed di-etary cholesterol. J Lipid Res 18:211, 1977

77. Clarkson TG, Hamm TE, Bullock BC, Lehner NDM.Atherosclerosis in old world monkeys. In Atherosclerosis inPrimates (Primates in Medicine, vol 9), edited by Strong JP.New York, S Karger, 1975, pp 66-89

78. Clarkson TB, Lofland HB, Bullock BC, Lehner NDM, St.Clalr R, Prltchard RW. Atherosclerosis in some species ofNew World monkeys. Ann N Y Acad Sci 162:103, 1969

79. Strong JP, McGIII HC Jr. Diet and experimental atheroscle-rosis in baboons. Am J Pathol 50:669, 1967

80. Bullock BC, Clarkson TB, Lehner NDM, Lofland HB Jr,St. Clalr RW. Atherosclerosis in Cebus albifrons monkeys.III. Clinical and pathologic studies. Exp Mol Pathol 10:39,1969

81. McGIII HC Jr, Mott GE, Bramblett CA. Experimentalatherosclerosis in the baboon. In Atherosclerosis in Primates(Primates in Medicine, vol 9), edited by Strong JP. New York,S Karger, 1975, pp 41-65

82. Krltchevsky D, Davidson LM, Shapiro IL, Kim HK, Klta-gawa M, Malhotra S, Nalr PO, Clarkson TB, Winter PAD.Lipid metabolism and experimental artherosclerosis in ba-boons: influence of cholesterol-free, semisynthetlc diets. AmJ Clin Nutr 27:29, 1974

83. Strong JP, Eggen DA, Jlrge SK. Atherosclerotic lesionsproduced in baboons by feeding an atherogenic diet for fouryears. Exp Mol Pathol 24:320, 1976

84. The Geographic Pathology of Atherosclerosis, edited byMcGIII HC. Baltimore, Williams & Wilkins Co., 1968, p 41

85. Kagan A, Harris BR, Wlnkelsteln W Jr, Johnson KG,Kato H, Syme SL, Rhoads GG, Gay ML, Nichaman MZ,Hamilton HB, Ullotson J. Epidemiologic studies of coro-nary heart disease and stroke in Japanese men living inJapan, Hawaii and California: demographic, physical, dietaryand biochemical characteristics. J Chronic Dis 27:345, 1974

86. Kato H, Tlllotson J, Nichaman MZ, Rhoads GG, HamiltonHB. Epidemiologic studies of coronary heart disease andstroke in Japanese men living in Japan, Hawaii and Califor-nia. Am J Epidemiol 97:372, 1973

87. Marmot MG, Syme SL, Kagan H, Kato H, Cohen JB,Belsky J. Epidemiologic studies of coronary heart diseaseand stroke in Japanese men living in Japan, Hawaii andCalifornia: prevalence of coronary and hypertensive heartdisease and associated risk factors. Am J Epidemiol102:514, 1975

88. Robertson TL, Kato H, Gordon T, Kagan A, Rhoads GG,Land CE, Worth RM, Belsky JL, Dock DS, Myanlshl M,Kawamoto S. Epidemiologic studies of coronary heart dis-ease and stroke in Japanese men living in Japan, Hawaiiand California. Coronary heart disease risk factors in Japanand Hawaii. Am J Cardiol 39:244, 1977

89. Robertson TL, Kato H, Rhoads GG, Kagan A, Marmot M,Syme SL, Gordon T, Worth RM, Belsky JL, Dock DS,Mlyanlshl M, Kawamoto S. Epidemiologic studies of coro-nary heart disease and stroke in Japanese men living inJapan, Hawaii and California. Incidence of myocardial infarc-tion and death from coronary heart disease. Am J Cardiol39:239, 1977

90. Tlllotson JL, Kato H, Nichaman MZ, Miller DC, Gay ML,Johnson KG, Rhoads GG. Epidemiology of coronary heartdisease and stroke In Japanese men living in Japan, Hawaii,and California: methodology for comparison of diet. Am JClin Nutr 26:177, 1973

91. Worth RM, Kato H, Rhoads GG, Kagan A, Syme SL. Epi-demiologic studies of coronary heart disease and stroke inJapanese men living in Japan, Hawaii and California: mortal-ity. Am J Epidemiol 102:481, 1975

92. Taylor CB, Allen ES, Mlkkelson B, Kang-Jey H. Serumcholesterol levels of Seventh-Day Adventists. Paroi Arterlelle3:175, 1976

93. Walden RT, Schaefer LE, Lemon FR, Aunshine A, Wy-nder EL. Effect of environment on the serum cholesterol-triglyceride distribution among Seventh-Day Adventists. Am


Dow

nloaded from



J Med 36:269, 196494. West RO, Hayes OB. Diet and serum cholesterol levels —

a comparison between vegetarians and nonvegetarians ina Seventh-Day Adventist group. Am J Clin Nutr 21:853,1964

95.Wynder EL, Lemon FR. Cancer, coronary artery diseaseand smoking: a preliminary report on differences In inci-dences between Seventh-Day Adventists and others. CalifMed 89:267, 1958

96Wynder EL, Lemon FR, Bross IJ. Cancer and coronaryartery disease among Seventh-Day Adventists. Cancer12:1016, 1959

97Shekelle RB, Shryock AM, Paul 0, Leppar M, Stamler J,Liu S, Raynor WJ Jr. Diet, serum cholesterol, and deathfrom coronary heart disease. The Western Electric study. NEngl J Med 304:65, 1981

98. Armstrong BK, Mann Jl, Adelsteln AM, Eskln F. Com-modity consumption and ischemic heart disease mortality,with special reference to dietary practices. J Chronic Dis28:455, 1975

99. Jollltfe N, Archer M. Statistical associations between inter-national coronary heart disease death rates and certain en-vironmental factors. J Chronic DIs 9:636, 1959

100. Lopez-S A, Krehl WA, Hodges RE, Good El. Relationshipbetween food consumption and mortality from atheroscle-rotic heart disease in Europe. Am J Clin Nutr 19:361, 1966

101.Maslronl R. Dietary factors and coronary heart disease.Bull WHO 42:103, 1970

102. Klmura N. Analysis of 10,000 postmortem examinations inJapan. In World Trends in Cardiology: I. CardiovascularEpidemiology, edited by Keys A, White PD. New York, Hoe-ber-Harper, 1956, p 159

103. Mann GV, Munoz JA, Scrimshaw NS. Serum lipoproteinand cholesterol concentrations of Central and North Amer-icans with different dietary habits. Am J Med 19:25, 1955

104. Scrimshaw NS, Trulson M, Te|ada C, Hegsted DM, StareFJ. Serum lipoprotein and cholesterol concentrations: com-parisons of rural Costa Rlcan, Guatemalan, and UnitedStates populations. Circulation 15:805, 1957

1O5.Te|ada C, Gore I. Comparison of atherosclerosis In Guate-mala City and New Orleans. Am J Pathol 33:887, 1957

106. Katz LN, Stamler J, Pick R. Nutrition and Atherosclerosis.Philadelphia, Lea & Febiger, 1958

107. Brown J, Bourke GJ, Gearty GF, Flnnegan A, Hill M,Heffeman-Fox FC, Fitzgerald DE, Kennedy J, ChlldersRW, Jessop WJE, Trulson MF, Latham MC, Cronln S,McCann MB, Clancy RE, Gore I, Stoudt HW, HegstedWM, Stare FJ. Nutritional and epidemiologic factors relatedto heart disease. In World Review of Nutrition and Dietetics,vol 12, edited by Bourne GH. New York, S Karger, 1970,P1

108. Eskln F. The role of the egg as a factor in the aetiology ofcoronary heart disease. Community Health 2:179, 1971

109.Malmros H. Relation of nutrition to health — a statisticalstudy of the effect of war-time on arteriosclerosis, cardio-sclerosis, tuberculosis and diabetes. Acta Med Scand246(suppl):137, 1950

110. Wen CP, Gershoff SN. Changes in serum cholesterol andcoronary heart disease mortality associated with changesin the postwar Japanese diet. Am J Clin Nutr 26:616, 1973

m.Joossens JV, Brems-Heyns E, Claes JH, DeBacker G,Grafter M, Kesteloot H, Kornltzer M, Lequlme J, PannierR, Raes A, VanHoute O, Vastesaeger M, Verdonk G. Thepattern of food and mortality in Belgium. Lancet 1:1069,1977

112. Kozarevlc D, Plrc B, Dawber TR, Gordon T, Zukel WJ,Vojvodlc N. The Yugoslavia Cardiovascular Disease Study.I. The incidence of coronary heart disease by area. J Chro-nic Dis 29:405, 1976

113. Kozarevlc D, Plrc B, Raclc Z, Dawber TR, Gordon T,Zukel WJ. The Yugoslavia Cardiovascular Disease Study.II. Factors in the incidence of coronary heart disease. Am JEpidemiol 104:133, 1976

114. The Framlngton Diet Study, diet and the regulation of

serum cholesterol. In The Framingham Study: An Epidemi-ologlcal Investigation of Cardiovascular Disease, edited byKannel WB, Gordon T Washington DC, U S Govt PrintingOff, 1970, Section 24

115. Nichols AB, Ravenscroft C, Lamphlear DE, OstranderLD. Daily nutritional intake and serum lipid levels. The Te-cumseh Study. Am J Clin Nutr 29:1384, 1976

116. Rlnzler SH. Primary prevention of coronary heart diseaseby diet. Bull N Y Acad Med 44:936, 1968

117. Stamler J. Acute myocardial infarction — progress in pri-mary prevention. Br Heart J 33 (suppl): 145, 1971

118 Hall Y, Stamler J, Cohen DB, Mojonnler L, Epstein MB,Berkson DM, Whlpple IT, Catchlngs S. Effectiveness of alow saturated fat, low cholesterol, weight-reducing diet forthe control of hypertriglyceridemia. Atherosclerosis 16:389,1972

119. Mlettlnen M, Turpelnen O, Karvonen MJ, Elosuo R,Paavllalnen E. Effect of cholesterol-lowering diet on mor-tality from coronary heart-disease and other causes. Atwelve-year clinical trial in men and women. Lancet 2:835,1972

12O.Frantz ID Jr, Dawson EA, Kuba K, Brewer ER, Gate-wood LC, Bartsch GE. The Minnesota coronary survey:effect of diet on cardiovascular events and death, (abstr)Circulation 52 (suppl II): II-4, 1975

121. Dayton S, Pearce ML, Hashimoto S, Dlxon WJ, Toml-yasu U. A controlled clinical trial of a diet high in unsat-urated fat in preventing complications of atherosclerosis.Circulation 40 (suppl II): 11-1, 1969

122. Leren P. The effect of plasma cholesterol lowering diet inmale survivors of myocardial infarction. A controlled clinicaltrial. Acta Med Scand 466 (suppl): 1-92, 1966

123. Hjermann I, Velve Byre K, Holme I, Leren P. Effect of dietand smoking intervention on the incidence of coronaryheart disease. Lancet 2:1303, 1981

124. National Diet-Heart Study Research Group. NationalDiet-Heart Study: final report. Circulation 37 (suppl I): 1-1,1968

125. Task Force on Arteriosclerosis of the National Heartand Lung Institute. Arteriosclerosis, vol 1, Washington,DC, U S Department of Health, Education, and Welfare,DHEW Pub, 1971

126. The Multiple Risk Factor Intervention Trial (MRFIT). anational study of primary prevention of coronary heart dis-ease. JAMA 235:825, 1976

127. The Multiple Risk Factor Intervention Trial Group. Sta-tistical design considerations in the NHLI Multiple Risk Fac-tor Intervention Trial (MRFIT). J Chronic Dis 30:261, 1977

128. Rtfklnd BM. Current status of clinical trials of CHD preven-tion through lipid lowering. Arteriosclerosis, 2:67, 1977

129. Levy Rl. Cholesterol, lipoproteins, apoproteins and heartdisease: present status and future prospects. Clin Chem27:653, 1981

130. Mann GV. Diet-Heart: end of an era. N Engl J Med 297:644,1977

131.McMlchael J. Dietetic factors in coronary disease. Eur JCardiol 5:447, 1977

132. McMlchael J. Letter: Prevention of coronary heart-disease.Lancet 2:569, 1976

133.Truswell AS. Diet and plasma lipids — a reappraisal. Am JClin Nutr 31:977, 1978

134. Harper AE. Dietary goals — a skeptical view. Am J ClinNutr 31:310, 1978

135. Olson RE. Is there an optimum diet for the prevention ofcoronary heart disease? In Nutrition, Upids, and CoronaryHeart Disease, vol 1, edited by Levy Rl, Rifkind BM, DennisBH, Ernst N. New York, Raven Press, Pubs., 1979, p 350

136. Ahrens EH Jr. The management of hyeriipidemia: whether,rather than how. Ann Intern Med 85:87, 1976

137. Council on Foods and Nutrition. American MedicalAssociation and Food and Nutrition Board of theNational Academy of Sciences — National ResearchCouncil. Diet and coronary heart disease: a Council state-ment. JAMA 222:1647, 1972


Dow

nloaded from



138. Food and Nutrition Board, Division of Biological Sci-ences, Assembly of Life Sciences — National ResearchCouncil. Toward healthful diets. Washington, DC, NationalAcademy of Sciences, 1980

139. Sacks FM, Caatelli WP, Donner A, Kass EH. Plasmalipids and lipoproteins in vegetarians and controls. N Engl JMed 292: 1148, 1975

14O.Burslem J, Schonfeld G, Howard MA, Weldman SW,Miller JP. Plasma apoprotein and lipoprotein lipid levels invegetarians. Metabolism 27:711, 1978

141. Savage PJ, Hamman RF, Bartha G, Dlppe S, Miller M,Bennett PH. Serum cholesterol levels in American (Pima)Indian children and adolescents. Pediatrics 58:274, 1976

142. Keys A. Seven Countries: A Multivariate Analysis of Deathand Coronary Heart Disease. Cambridge, Massachusetts,Harvard Univ. Press, 1980

143. Zllversmit DB. A proposal linking atherogenesis to the in-teraction of endothelial lipoprotein lipase with trigtyceride-rich lipoproteins. Circ Res 33:633, 1973

144. Fielding CJ. Metabolism of cholesterol-rich chylomicrons.Mechanism of binding and uptake of cholesterol esters bythe vascular bed of the perfused rat heart. J Clin Invest62:141, 1978

145. Blackburn H. Diet and mass hyperlipidemia: a publichealth view. In Nutrition, Lipids, and Coronary Heart Dis-ease, vol 1, edited by Levy Rl, Rifkind BM, Dennis BH,Ernst N. New York, Raven Press, Pubs, 1979, p 309

146Lofland HB Jr, Clarkson TB, St. Clalr RW, Lehner NDM.Studies on the regulation of plasma cholesterol levels insquirrel monkeys of two genotypes. J Llpid Res 13:39,1972

147. Eggen DA. Cholesterol metabolism in groups of rhesusmonkeys with high or low response of serum cholesterol toan atherogenic diet. J Upid Res 17:663, 1976

148. Sturdevant RAL, Pearce ML, Dayton S. Increased preva-lence of cholelithiasis in man ingesting a serum-cholester-ol-lowering diet. N Engl J Med 288:24, 1973

149.Grundy SM. Effects of polyunsaturated fats on lipid metab-olism in patients with hypertriglyceridemia. J Clin Invest55:269, 1975

150. Carroll KK, Gammal EB, Plunkett ER. Dietary fat andmammary cancer. Can Med Assoc J 98:590, 1968

151. Carroll KK, Khor HT. Effects of level and type of dietaryfat on incidence of mammary tumors induced in femaleSprague-Dawley rats by 7, 12 dimethylbenz () anthracene.Lipids 6:415, 1971

152. Gammal EB, Carroll KK, Plunkett E R. Effects of dietaryfat on mammary carcinogenesis by 7, 12-dimethymethy-

benz (alpha) anthracene in rats. Cancer Res 27:1737, 1967153. King ME, Spector AA. Effect of specific fatty acyl enrich-

ments on membrane physical properties detected with aspin label probe. J Biol Chem 253:6493, 1978

154. Rose G, Blackburn H, Keys A, et al. Colon cancer andblood cholesterol. Lancet 1:181, 1974

155.Beaglehole R, Foulkes MA, Prior IA, Eyles EF. Choles-terol and mortality In New Zealand Maoris. Br Med J280:285, 1980

156. Rose G, Shipley MJ. Plasma lipids and mortality: a sourceof error. Lancet 1:523, 1980

157. Camblen R, Duclmetlere P, Richard J. Total serumcholesterol and cancer mortality in a middle-aged malepopulation. Am J Epidemiol 112:388, 1980

158. Kark JD, Smith AH, Hames CG. Trie relationship of serumcholesterol to the incidence of cancer in Evans County,Georgia. J Chronic Dis 33:311, 1980

159 Williams RR, Sorlle PD, Felnleib M, McNamara PM,Kannel WB, Dawber TR. Cancer incidence by levels ofcholesterol. JAMA 245:247, 1981

160. Cholesterol and noncardlovascular mortality. JAMA244:25, 1980

161. Liu K, Stamler J, Moss D, Garslde D, Persky V, SolteroI. Dietary cholesterol, fat, and fibre, and colon-cancer mor-tality. An analysis of international data. Lancet 2:782, 1979

162. Wynder EL. The epidemiology of large bowel cancer. Can-cer Res 35:3388, 1975

163. Mower HF, Ray RM, Shoff R, Stemmerman GN, NomuraA, Glober GA, Kamlyama S, Shlmada A, Yamakawa H.Fecal bile acids in two Japanese populations with differentcolon cancer risks. Cancer Res 39:328, 1979

164. Reddy BS, Hedges AR, Laakso K, Wynder EL. Metabolicepidemiology of large bowel cancer: fecal bulk and constit-uents of high-risk North American and low-risk Finnishpopulation. Cancer 42:2832, 1978

165. Pearce ML, Dayton S. Incidence of cancer in men on adiet high in polyunsaturated fat. Lancet 1:464, 1971

166. Ederer F, Leren P, Turpelnen O, et al. Cancer among menon cholesterol-lowering diets. Lancet 2:203, 1971

167. Shepherd J, Packard CJ, Patsch JR, Gotto AM Jr, Taun-ton OD. Effects of dietary polyunsaturated and saturatedfat on the properties of high density lipoproteins and themetabolism of apolipoprotein A-l. J Clin Invest 61:1582,1978

168. Vega GL, Groszek E, Wolf R, Grundy SM. Influence ofpolyunsaturated fats on composition of plasma lipoproteinsand apoproteins. J Lipid Res, to be published


Dow

nloaded from


nutrition committee.Rationale of the diet-heart statement of the American Heart Association. Report of the AHA

Print ISSN: 1079-5642. Online ISSN: 1524-4636 Copyright © 1982 American Heart Association, Inc. All rights reserved.

Avenue, Dallas, TX 75231is published by the American Heart Association, 7272 GreenvilleArteriosclerosis, Thrombosis, and Vascular Biology

doi: 10.1161/01.ATV.2.2.1771982;2:177-191Arterioscler Thromb Vasc Biol.

http://atvb.ahajournals.org/content/2/2/177.citationWorld Wide Web at:

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

http://atvb.ahajournals.org//subscriptions/

at: is onlineArteriosclerosis, Thrombosis, and Vascular Biology Information about subscribing to Subscriptions:

http://www.lww.com/reprints

Information about reprints can be found online at: Reprints:

document.Permissions and Rights Question and AnswerFurther information about this process is available in theis being requested is located, click Request Permissions in the middle column of the Web page under Services.Clearance Center, not the Editorial Office. Once the online version of the published article for which permission

can be obtained via RightsLink, a service of the CopyrightArteriosclerosis, Thrombosis, and Vascular Biology Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:


Dow

nloaded from

http://atvb.ahajournals.org/content/2/2/177.citation

http://www.ahajournals.org/site/rights/

http://www.lww.com/reprints

http://atvb.ahajournals.org//subscriptions/


Special Report Rationale of the Diet-Heart Statement …atvb.ahajournals.org/content/atvbaha/2/2/177.full.pdfSpecial Report Rationale of the Diet-Heart Statement of the American Heart

Documents