Biochimica et Biophysica Acta 1529 (2000) 310^320

www.elsevier.com/locate/bba

Review
Dietary cholesterol and atherosclerosis
Donald J. McNamara *
Egg Nutrition Center, 1050 17th St. NW, Suite 560, Washington, DC 20036, USA

Received 16 May 2000; received in revised form 14 August 2000; accepted 15 August 2000

Abstract

The perceived relationship between dietary cholesterol, plasma cholesterol and atherosclerosis is based on three lines of
evidence : animal feeding studies, epidemiological surveys, and clinical trials. Over the past quarter century studies
investigating the relationship between dietary cholesterol and atherosclerosis have raised questions regarding the
contribution of dietary cholesterol to heart disease risk and the validity of dietary cholesterol restrictions based on these
lines of evidence. Animal feeding studies have shown that for most species large doses of cholesterol are necessary to induce
hypercholesterolemia and atherosclerosis, while for other species even small cholesterol intakes induce hypercholesterolemia.
The species-to-species variability in the plasma cholesterol response to dietary cholesterol, and the distinctly different plasma
lipoprotein profiles of most animal models make extrapolation of the data from animal feeding studies to human health
extremely complicated and difficult to interpret. Epidemiological surveys often report positive relationships between
cholesterol intakes and cardiovascular disease based on simple regression analyses; however, when multiple regression
analyses account for the colinearity of dietary cholesterol and saturated fat calories, there is a null relationship between
dietary cholesterol and coronary heart disease morbidity and mortality. An additional complication of epidemiological
survey data is that dietary patterns high in animal products are often low in grains, fruits and vegetables which can contribute
to increased risk of atherosclerosis. Clinical feeding studies show that a 100 mg/day change in dietary cholesterol will on
average change the plasma total cholesterol level by 2.2^2.5 mg/dl, with a 1.9 mg/dl change in low density lipoprotein (LDL)
cholesterol and a 0.4 mg/dl change in high density lipoprotein (HDL) cholesterol. Data indicate that dietary cholesterol has
little effect on the plasma LDL:HDL ratio. Analysis of the available epidemiological and clinical data indicates that for the
general population, dietary cholesterol makes no significant contribution to atherosclerosis and risk of cardiovascular
disease. ß 2000 Elsevier Science B.V. All rights reserved.

Keywords: Dietary cholesterol; Saturated fat; Atherosclerosis; Epidemiology ; Plasma cholesterol; Low density lipoprotein;
High density lipoprotein

1. Introduction intense research, and considerable debate, for a

good part of the 20th century. From the 1913 animal
The relationship between dietary cholesterol and studies of Anitschkow and Chalatow [1] showing
coronary heart disease (CHD) has been a topic of that feeding cholesterol to rabbits induced atherogen-
esis to the 1999 epidemiological surveys reported by
Hu et al. [2] and Ascherio et al. [3] indicating that
dietary cholesterol is unrelated to CHD risk, the diet-
* Fax: +1-202-463-0102; E-mail: [email protected] ary cholesterol-CHD risk relationship has gone from

1388-1981 / 00 / $ ^ see front matter ß 2000 Elsevier Science B.V. All rights reserved.
PII: S 1 3 8 8 - 1 9 8 1 ( 0 0 ) 0 0 1 5 6 - 6

BBAMCB 55729 28-11-00

 D.J. McNamara / Biochimica et Biophysica Acta 1529 (2000) 310^320 311

a hypothesis to a widely accepted belief to what sclerosis can be induced. Another complication of
many today consider to be an ine¡ective dietary re- animal studies is that most animal species have a
striction. signi¢cantly di¡erent plasma lipoprotein pro¢le com-
The concept that dietary cholesterol contributes to pared to humans. Whereas humans have low density
hypercholesterolemia and CHD risk has been a fun- lipoprotein (LDL) cholesterol as the predominant
damental part of public health policy and dietary plasma lipoprotein, most animal models have high
recommendations in the United States for over 30 density lipoprotein (HDL) cholesterol as the major
years. In the 1970s the recommendation that choles- fraction. The species di¡erences in the response to
terol be restricted in the diets of the general popula- dietary cholesterol, the use of pharmacological doses
tion, and severely restricted in the diets of those with of dietary cholesterol in many studies, and di¡eren-
hypercholesterolemia, was based on three lines of ces in the plasma lipoprotein pro¢le make extrapo-
evidence: (1) animal studies showing that dietary lations from the results of animal feeding studies to
cholesterol induces hypercholesterolemia and athero- human health recommendations di¤cult, if not im-
sclerosis in some species; (2) epidemiological surveys possible.
reporting a positive relationship between dietary cho-
lesterol and CHD incidence; and (3) clinical obser-
vations that feeding cholesterol increases plasma to- 3. Epidemiological surveys
tal cholesterol levels. Based on this evidence, a
number of organizations recommended restricting di- Historically, cross-cultural epidemiological surveys
etary cholesterol levels for the population in an e¡ort have been some of the strongest evidence that dietary
to reduced plasma cholesterol levels and CHD risk cholesterol is associated with CHD incidence. Many
[4^8]. epidemiological studies have reported a signi¢cant
Over the last quarter century an extensive body of positive relationship between dietary cholesterol
research on the relationship between dietary choles- and both plasma total cholesterol levels and CHD
terol and both blood cholesterol levels and CHD incidence using simple regression analyses. For ex-
incidence has been published. These studies include ample, as shown in Fig. 1, data from the Seven
large epidemiological surveys in populations fol- Countries Study indicate a signi¢cant correlation be-
lowed for extended time periods as well as numerous tween the population average cholesterol intake (mg/
feeding trials which investigated the e¡ects of dietary 1000 kcal) and the 25 year mortality rate from CHD
cholesterol on plasma total and lipoprotein choles- (R2 = 0.298, P = 0.029) [10]. However, with today's
terol levels and on whole body cholesterol and lipo- understandings of the relationships between dietary
protein metabolism. This review summarizes the cur- factors and CHD risk, it is clear that two confound-
rent state of research on the association between
dietary cholesterol and atherosclerosis.

2. Animal studies

The plasma cholesterol response to dietary choles-

terol is highly variable across and within animal spe-
cies. While rabbits are highly susceptible to dietary
cholesterol, rats and dogs exhibit little change in
plasma total cholesterol even with high doses of diet-
ary cholesterol. Non-human primates are highly var-
iable in their responses to dietary cholesterol [9] and
in many species it is only with extremely high doses Fig. 1. Correlation of mean dietary cholesterol (mg/1000 kcal)
of dietary cholesterol (0.5^2 mg/kcal or 1250^5000 and 25 year CHD mortality (%) in 12 763 men from the 16 co-
mg/2500 kcal) that hypercholesterolemia and athero- horts of the Seven Countries Study [10].

BBAMCB 55729 28-11-00

312 D.J. McNamara / Biochimica et Biophysica Acta 1529 (2000) 310^320

Fig. 2. Correlation of mean saturated fatty acid intake (% kcal) Fig. 3. Correlation of mean saturated fatty acid intake (% kcal)
and 25 year CHD mortality (%) in 12 763 men from the 16 co- and mean dietary cholesterol (mg/1000 kcal) in 12 763 men
horts of the Seven Countries Study [10]. from the 16 cohorts of the Seven Countries Study [10].

ing variables signi¢cantly impact on the interpreta- ported in 1988 by Hegsted and Ausman in an anal-
tion of these epidemiological data. Cholesterol intake ysis of dietary data from the Twenty Countries Study
serves as a surrogate marker for two other dietary [11]. These authors found that dietary cholesterol
patterns associated with increased CHD risk; a high was signi¢cantly related to CHD incidence with sim-
intake of saturated fat resulting in elevated plasma ple correlation analysis but no longer statistically sig-
cholesterol levels, and a dietary pattern low in fruits, ni¢cant when multivariate analysis including satu-
grains and vegetables resulting in lower intakes of B rated fat calories was used to analyze the data
vitamins, antioxidants and dietary ¢ber. (Table 1).
A consistent ¢nding of epidemiological studies of In many of the more recent epidemiological sur-
the relationships between dietary factors and CHD veys, investigators have noted that with simple cor-
incidence is that the percent saturated fat calories in relations dietary cholesterol is sometimes correlated
the diet are positively correlated with CHD incidence with CHD incidence but, upon inclusion of saturated
as shown in Fig. 2 for the Seven Countries Study fat and dietary ¢ber in the analyses, dietary choles-
(R2 = 0.772, P 6 0.0001) [10]. As shown in Fig. 3, terol losses its signi¢cant relationship with CHD in-
these data also demonstrate that dietary saturated cidence [2,3,12,13]. It is clear that any evaluation of
fat and cholesterol are related covariables the epidemiological evidence for a relationship be-
(R2 = 0.380, P = 0.011). As the data demonstrate, di- tween dietary cholesterol and atherosclerosis ac-
etary cholesterol in a simple regression analysis is counts for the association between dietary cholesterol
positively related to CHD incidence but, when multi- and saturated fat in the diet and that the data be
variate analysis of the data accounts for the colinear- evaluated using multivariate analysis to correct for
ity of dietary cholesterol and saturated fat, dietary confounding by the signi¢cant impact of dietary sa-
cholesterol is no longer signi¢cantly related to CHD turated fat.
mortality rates (P = 0.976). Similar ¢ndings were re-

Table 1 4. Relationship between dietary cholesterol and CHD

Diet and CHD in 20 countries: regression coe¤cientsa
Regression Cholesterol Saturated Polyunsaturated In 1995, Ravnskov [14] published a review of 13
analysis fat fat case-control studies carried out between 1968 and
Simple 3.75* 44.1* 326.7* 1985 which measured dietary cholesterol intakes
Multiple 31.90 61.1* 331.8* and CHD (Fig. 4) [15^23]. The mean cholesterol in-
*P 6 0.05. take in CHD patients was 223 þ 11 mg/1000 kcal
a
Data from Hegsted and Ausman [11]. compared to controls with an average intake of

BBAMCB 55729 28-11-00

 D.J. McNamara / Biochimica et Biophysica Acta 1529 (2000) 310^320 313

Fig. 5. Relative risk of myocardial infarction (MI) and of fatal

CHD relative to mean dietary cholesterol intake quintile for
Fig. 4. Cholesterol intakes (mg/1000 kcal) of CHD cases and
43 757 males in the Health Professionals Follow-Up Study [3].
controls in 13 case-control studies [14]. Data from Kannel and
There was no signi¢cant relationship between dietary cholester-
Gordon [15], Garcia-Palmieri et al. [16], Gordon et al. [17],
ol intakes and MI (P = 0.48) or fatal CHD (P = 0.21).
McGee et al. [18], Kromhout et al. [19], Kushi et al. [20], Bas-
sett et al. [21], Finegan et al. [22], and Yano et al. [23].

(P = 0.06) intakes. In multiple linear regression anal-

216 þ 11 mg/1000 kcal (P = 0.021). The mean di¡er- ysis, the association of progression with dietary cho-
ence between CHD cases and controls was 6.4 mg/ lesterol was no longer signi¢cant.
1000 kcal equivalent to a di¡erence in dietary choles- Esrey et al. [26] analyzed data from the Lipid Re-
terol of 16 mg/day for someone on a 2500 kcal diet. search Clinics Prevalence Follow-Up Study of 4546
Given that the CHD cases also had a higher intake men and women and reported that dietary cholester-
of saturated fat calories than controls (P = 0.009), it ol was not signi¢cantly related to CHD deaths in any
is not surprising that the cases had higher cholesterol age gender group. Ascherio et al. [3] reported a study
intakes. Whether the increased CHD risk was due to of 43 757 males in the Health Professionals Follow-
the higher intake of saturated fat or to cholesterol Up Study and found no relationship between quintile
cannot be decided from these data; however, it is of cholesterol intake and either myocardial infarction
di¤cult to propose a plausible biological explanation or coronary deaths (Fig. 5). Dietary ¢ber intake had
for an increased CHD risk with a 16 mg/day change a signi¢cant e¡ect on the relationships between diet-
in dietary cholesterol given that the daily metabolism ary variables and CHD incidence, and inclusion of
of dietary and endogenously synthesized cholesterol dietary fat and ¢ber in the multivariate analyses re-
in the body is over 1000 mg. Based on these data,
Ravnskov [14] concluded that these case-control
studies do not provide convincing evidence for a re-
lationship between dietary cholesterol and CHD risk.
Epidemiological studies published in the last de-
cade on the role of dietary factors in CHD risk
have routinely employed multivariate analyses of
the data to account for the e¡ects of associated var-
iables such as saturated fat and dietary ¢ber [24]. For
example, Watts et al. [25] reported data from the St.
Thomas Atherosclerosis Regression Study relating
nutrient intake and progression of coronary artery
disease in 50 males. In univariate linear regression
Fig. 6. Relative risk of CHD per quintile of dietary cholesterol
analysis, progression of disease over 39 months was intake for 80 082 women in the Nurses' Health Study [2]. There
related to dietary energy (P 6 0.001), total fat was no signi¢cant relationship between dietary cholesterol in-
(P 6 0.001), saturated fat (P 6 0.001) and cholesterol takes and CHD (P = 0.24).

BBAMCB 55729 28-11-00

314 D.J. McNamara / Biochimica et Biophysica Acta 1529 (2000) 310^320

The authors reported that, after adjustment for age

and energy intake, intake of animal fat, saturated fat,
monounsaturated fat, and cholesterol were positively
related to wall thickness, while vegetable fat and
polyunsaturated fat intakes were negatively related.
No analyses based on multiple linear regression anal-
yses were reported. Mann et al. [29] reported the
results from a study of 10 802 relatively healthy con-
scious men and women (some 50% vegetarian) and
found a signi¢cant relationship between tertiles of
Fig. 7. Incidences of CHD events and CHD deaths for quintile dietary cholesterol and ischemic heart disease mortal-
of dietary cholesterol intakes in 21 930 men in the Alpha-Toco- ity. These data were analyzed using univariate anal-
pherol, Beta-Carotene Cancer Prevention Study [12]. There was yses and the interactive e¡ects of saturated fat and
no signi¢cant relationship between dietary cholesterol intakes dietary ¢ber on the dietary cholesterol relationship
and CHD events (P = 0.81) or CHD deaths (P = 0.77).
were not tested.
Other studies have investigated the relationships
sulted in a non-signi¢cant e¡ect of dietary cholester- between dietary factors and CHD incidence, and
ol. Hu et al. [2] reported data from 80 082 women in while dietary cholesterol was included as a measured
the Nurses' Health Study showing that dietary cho- dietary variable in the study, the primary data were
lesterol was not related to CHD incidence (Fig. 6). not reported for cholesterol; however, the text con-
As in the Health Professionals Study, relative risk for tains a statement that dietary cholesterol was not a
CHD did not di¡er for women in the quintiles of signi¢cant factor (see for example [30,31]). Willett
cholesterol intakes. Noted in this report was a sig- recently summarized the results from 20 prospective
ni¢cant relationship between dietary cholesterol and cohort studies of dietary factors in relation to risk of
saturated fatty acids. Those in the lowest saturated CHD and noted that dietary cholesterol was re-
fat intake had the lowest cholesterol intakes and vice ported to be signi¢cantly associated with CHD in
versa. only two studies [24]; and in these two studies only
Pietinen et al. [12] reported an analysis of diet- simple regression analyses were reported.
CHD relationships from the Alpha-Tocopherol,
Beta-Carotene Cancer Prevention Study of 21 930
men. Consistent with ¢ndings from other epidemio- 5. Eggs and atherosclerosis
logical surveys, dietary cholesterol was not a signi¢-
cant contributor to CHD incidence, either events or One-third of the cholesterol in the American diet
deaths (Fig. 7). The intakes of cholesterol by this
Finnish population were substantially higher than
that of the various USA study groups yet the data
indicate that even at these higher intakes (upper
quintile 768 mg/day), dietary cholesterol was not re-
lated to CHD incidence. In a Greek case-control
study, Tzonou et al. [27] reported no association be-
tween dietary cholesterol and relative risk for CHD
in 329 CHD patients and 570 controls.
Two studies have reported a positive association
between dietary cholesterol and atherosclerosis using
simple regression analyses. Tell et al. [28] published
data from the Atherosclerosis Risk in Communities Fig. 8. Percent contribution of cholesterol and saturated fat
(ARIC) Study of 13 148 men and women relating from fats/oils, meats, dairy products and eggs in the US diet
dietary variables and carotid artery wall thickness. [32].

BBAMCB 55729 28-11-00

 D.J. McNamara / Biochimica et Biophysica Acta 1529 (2000) 310^320 315

comes from eggs, a high-cholesterol, low-saturated results were analyzed using a multivariate model in-
fat food while two-thirds of the cholesterol come cluding total energy, smoking, alcohol consumption,
from other animal products which contribute both hypertension, parental history of CHD, body mass
cholesterol and a high percentage of the saturated index, multivitamin use and vitamin E supplement
fat in the diet (Fig. 8) [32]. While the evidence for use. As shown in Fig. 9, egg consumption, up to
a relationship between total dietary cholesterol and 1+ eggs per day, was unrelated to CHD risk in wom-
CHD incidence are complicated by the colinearity of en (RR 0.82, 95% con¢dence interval (CI) 0.60^1.13,
saturated fat with cholesterol in the diet, studies of P for trend 0.95) and in men (RR 1.08, 95% CI 0.79^
the relationship between egg consumption and CHD 1.27, P for trend 0.75). Egg consumption was also
speci¢cally test whether dietary cholesterol is associ- unrelated to risk of either ischemic stroke (RR for 1+
ated with CHD risk. eggs/day 0.81, 95% CI 0.46^1.42, P for trend 0.81) or
In 1982 Dawber et al. [33] reported no association hemorrhagic stroke (RR for 1+ eggs/day 1.07, 95%
between egg consumption and incidence of CHD in CI 0.56^2.03, P for trend 0.81).
912 participants in the Framingham Heart Study The investigators determined that the background
even with a 10-fold range of egg consumption be- cholesterol intake did not change the ¢ndings. In the
tween the lowest and highest tertile. Two other stud- Nurses' Health Study 4.8% (3844) reported that they
ies also reported no relationships between egg con- almost never ate eggs and 1.6% (1281) reported con-
sumption and CHD incidence [34,35]. Data from the suming two or more eggs a day. Comparison of these
California Adventists Study indicated that CHD rel- two extreme groups using multivariate analysis indi-
ative risk was 1.01 with a higher egg intake (v 3 per cated that the relative risk for CHD was 0.76 (95%
week) compared to the lowest intake ( 6 1 per week) CI 0.43^1.35) in the high egg consuming group com-
[35]. Gramenzi et al. [34] reported data from a case- pared to those who never ate eggs [36].
control study indicating that the CHD relative risk In a subgroup analysis (strati¢ed risk factors in-
(RR) for women in the upper third of egg intake was cluding hypercholesterolemia, diabetes, hypertension,
0.8 compared to those in the lower third. smoking, alcohol use, body mass index, age, and in-
Hu et al. [36] reported data from the Nurses' takes of saturated fat, polyunsaturated fat and car-
Health Study (NHS, n = 80 082 followed for 14 years) bohydrates) the authors reported that there was no
and Health Professionals Follow-Up Study (HPFS, evidence of a positive association of egg consumption
n = 37 851 followed for 8 years) on the relationships and CHD in any subgroup except a suggestion of
between egg consumption and risk of CHD and elevated risk among those with diabetes. The obser-
stroke. The investigators classi¢ed egg consumption vation that egg consumption increased CHD risk in
patterns into ¢ve groups: 6 1 egg/week, 1 egg/week, male and female diabetics in the study by Hu et al.
2^4 eggs/week, 5^6 eggs/week and v 1 egg/day. The [36] is in contrast to the report by Toeller et al. [13]
from the EURODIAB IDDM Complications Study
that dietary cholesterol was not signi¢cantly related
to either plasma cholesterol levels or CHD incidence
in 2868 subjects with type I diabetes. The interac-
tions between eggs [36], dietary cholesterol and diet-
ary ¢ber [13], type of diabetes, and CHD risk clearly
needs further research to determine what factors are
involved in increasing CHD risk.
Data from the Seven Countries Study and other
epidemiological surveys reported positive correla-
tions between total cholesterol intakes and CHD
mortality rates across populations. Surprisingly, a
Fig. 9. Relative risk of CHD incidence in 37 851 males (Health similar analysis of egg consumption versus CHD
Professionals Follow-Up Study) and 80 082 females (Nurses' mortality rates indicates that there is a negative rela-
Health Study) versus weekly egg consumption [36]. tionship between per capita egg consumption and

BBAMCB 55729 28-11-00

316 D.J. McNamara / Biochimica et Biophysica Acta 1529 (2000) 310^320

is that this extreme cholesterol intake in the top quin-

tile suggests that the overall dietary pattern of these
individuals was substantially more complicated than
just high in cholesterol. This level of cholesterol in-
take suggests extreme intakes of animal products
and, correspondingly, low intakes of grains, fruits
and vegetables. Here again, dietary cholesterol serves
as a surrogate marker for high intakes of saturated
fat and animal protein as well as low intakes of
grains, vegetables, and fruits in this subset. The un-
Fig. 10. Relationship between CVD mortality rates per 100 000 resolved question is whether the higher CHD inci-
men aged 35^74 years [37] and per capita egg consumption in dence in this group was due to what was excessive
24 industrialized countries (Annual Promotion and Marketing in the diet or, perhaps, what was inadequate in the
Survey No. 30, International Egg Commission, London, 1999). diet. With today's understanding of the role of diet-
ary ¢ber, vegetable protein, antioxidants and B vita-
CHD mortality rates per 100 000 men age 35^74 in mins in CHD risk [39^41], it is clear that increased
24 industrialized countries (Fig. 10). Similar results risk occurs not only from nutrient excesses but also
are obtained when the data are analyzed using fe- from suboptimal intakes of speci¢c nutrients. The
male CHD mortality rates. The countries with the potential importance of such confounding dietary
highest per capita egg consumptions are Japan, Mex- variables, which could contribute to the higher
ico, Spain and France which have low CHD mortal- CHD incidence in this quintile, was not evaluated
ity rates [37]. by Stamler and colleagues and raises questions re-
The studies which have reported data on egg con- garding the `independent e¡ect' hypothesis.
sumption and CHD rates uniformly indicate a null In contrast, studies reported by Connor and co-
relationship and suggest that the previously observed workers [42] indicate that di¡erent populations con-
positive relationship between total dietary cholesterol suming diets with similar Cholesterol-Saturated Fat
and CHD mortality rates is in large part explained Index (CSI: 1.01Ug SFA+0.05Umg cholesterol per
by the association between dietary saturated fat calo- 1000 kcal) values have di¡erent CHD incidence rates.
ries and dietary cholesterol, and the low ¢ber intakes While France and England have similar dietary CSI
in diets high in animal products [2,3,13,36]. patterns, they di¡er almost 5-fold in CHD mortality
rates. These investigators concluded that a high CSI
value was not a contributor to CHD risk if the diet
6. Independent e¡ect of dietary cholesterol on CHD contained large amounts of fruits and vegetables and
vegetable oils. Similarly, in the Ireland-Boston Study
Shekelle and Stamler [38] reported that in the the CHD cases had a lower vegetable food score
Western Electric Study those individuals in the upper (30.44 vs. 0.06) and a higher animal food score
most quintile of cholesterol intakes had signi¢cantly (0.24 vs. 30.04) than controls [20]. These data are
increased relative risk for CHD even after adjust- consistent with the concept that the diet-heart disease
ment for plasma cholesterol levels. The level of cho- relationship is a function of both what is in the diet,
lesterol intake in the ¢fth quintile averaged 1079 mg/ as well as what is missing from the diet and only
day and was associated with an increased CHD in- through appropriate correction for confounding diet-
cidence. In contrast, the fourth quintile, with an ary factors can valid dietary associations with CHD
average cholesterol intake of 827 mg/day, was not be determined.
signi¢cantly di¡erent than the bottom quintile. There
are two aspects of this report which need to be con-
sidered. The ¢rst is the recognition that dietary fac- 7. Clinical studies
tors can in£uence CHD risk through pathways other
than changes in plasma cholesterol levels. The second It is not possible to have a direct clinical trial of

BBAMCB 55729 28-11-00

 D.J. McNamara / Biochimica et Biophysica Acta 1529 (2000) 310^320 317

the e¡ects of dietary cholesterol on atherosclerosis, increases the plasma LDL cholesterol by 1.9 mg/dl
and accordingly studies have investigated the e¡ects and HDL cholesterol by 0.4 mg/dl. Numerous cho-
of dietary cholesterol on plasma lipids and lipopro- lesterol feeding studies have reported that dietary
teins, and endogenous cholesterol and lipoprotein cholesterol has no e¡ect on the LDL:HDL choles-
metabolism. The early metabolic ward studies con- terol ratio [57^62]. Thus, it is possible to increase
sistently documented an increase in plasma total cho- total plasma cholesterol levels without a signi¢cant
lesterol levels with an increase in cholesterol intake change in CHD risk provided that the LDL:HDL
and predictive equations were reported in 1965 by ratio remains constant. Addition of 100 mg/day cho-
Keys [43] and Hegsted [44]. There have been a large lesterol to the diet of a subject with a plasma choles-
number of cholesterol feeding studies carried out terol pro¢le of 220 mg/dl total cholesterol, 150 mg/dl
over the past 40 years and these studies have been LDL cholesterol and 50 mg/dl HDL cholesterol
analyzed in detail by numerous investigators [45^55]. (LDL:HDL ratio = 3.00) would be predicted [52] to
The overall consensus is that dietary cholesterol does increase plasma LDL by 1.9 mg/dl and HDL by 0.4
have a statistically signi¢cant, small e¡ect on plasma mg/dl, yet these increases would result in no signi¢-
cholesterol levels, and that the plasma cholesterol cant change in the LDL:HDL ratio (3.01) and, the-
response to dietary cholesterol is highly variable, oretically, no change in CHD risk.
with about 75^80% of the population classi¢ed as Three studies have found that dietary cholesterol
hypo-responders and 15^20% as hyper-responders. has no e¡ect on postprandial lipoproteins or on the
Most analyses of the data indicate that the average e¤cacy of plasma to facilitate cholesterol e¥ux from
plasma cholesterol response to a 100 mg/day change cells. Reports by Ginsberg et al. [57,58], Clifton and
in dietary cholesterol is between 2.2 and 2.5 mg/dl Nestel [63] and Knopp et al. [59] indicate that dietary
[48,52,53]. Meta-analyses of the data indicate that cholesterol has no negative e¡ects on the pattern of
the plasma cholesterol response to dietary cholesterol postprandial lipoproteins and that there are no sig-
is independent of dietary fat type and amount, and ni¢cant increases in any candidate atherogenic par-
unrelated to the baseline plasma cholesterol level ticles with either acute or long term dietary choles-
[48,52,53]. There also is no evidence for gender di¡er- terol feeding. Blanco-Molina et al. [64] reported that
ences in the plasma cholesterol response to a dietary cholesterol feeding to humans increased plasma-in-
cholesterol challenge [56]. duced cholesterol e¥ux from cells in culture. These
The clinical data justifying dietary cholesterol re- data do not provide any evidence that dietary cho-
strictions are based on the early observations that lesterol induces production of atherogenic lipopro-
dietary cholesterol raises plasma cholesterol which tein particles or inhibits reverse cholesterol transport
was the endpoint for determining dietary e¡ects on
CHD risk. The clear evidence that dietary cholesterol
does in fact increase plasma cholesterol to a small 8. Summary and conclusions
degree (0.024 mg/dl per mg/day) would seem to con-
£ict with the epidemiological evidence that dietary Over the past two decades there have been numer-
cholesterol is not associated with CHD risk. How- ous reports from clinical trials of dietary cholesterol
ever, the e¡ects of dietary cholesterol on plasma total feedings, epidemiological surveys and prospective
cholesterol cannot provide a true estimate of its ef- studies, and meta-analyses of various collections of
fects on CHD risk since changes can occur in both dietary lipid feeding trials showing that dietary cho-
the atherogenic LDL cholesterol fraction as well as lesterol has a small, but signi¢cant e¡ect on plasma
in the anti-atherogenic HDL cholesterol fraction. cholesterol levels (0.022^0.027 mg/dl per mg dietary
The meta-analysis of metabolic ward cholesterol cholesterol) which has little meaning relative to CHD
feeding studies reported by Clarke et al. [52] indicates risk. The epidemiological data from large popula-
that dietary cholesterol increases total cholesterol tions consistently show that dietary cholesterol has
levels by increasing both LDL and HDL cholesterol little e¡ect on CHD incidence. There have been no
levels. A 100 mg/day increase in dietary cholesterol studies validating the `independent e¡ect' of dietary

BBAMCB 55729 28-11-00

318 D.J. McNamara / Biochimica et Biophysica Acta 1529 (2000) 310^320

cholesterol on CHD risk and this observation can But often the recommendations are repeated so
readily be accounted for due to confounding dietary many times that people forget they were rough
covariables. guesses in the ¢rst place and come to think they
And yet the argument is made that there is a pos- are hard facts.' Public health recommendations can
itive relationship between dietary cholesterol and become dogma without the necessary scienti¢c evi-
plasma cholesterol, and that while the increase in dence, and eventually become impervious to argu-
CHD risk may be too small to determine in epide- ment and re-evaluation. The dietary cholesterol rec-
miological surveys, any reduction in plasma choles- ommendation is so widely accepted that it is now in
terol is desirable and adds to overall CHD risk the situation of `reverse onus' where it is no longer
reduction. The ¢nding that dietary cholesterol is necessary for those making the recommendation to
positively related to both LDL cholesterol and prove its validity but rather it is up to those who
HDL cholesterol, with little change in the question the restriction to prove it is not scienti¢cally
LDL:HDL ratio, provides a di¡erent interpretation justi¢ed.
of the data and suggests that the reason epidemio- In a discussion regarding community interventions
logical surveys fail to detect a relationship between to postpone CHD, Shaper and Marr [66] noted over
dietary cholesterol and CHD incidence is because 20 years ago that there was considerable confusion
there is no measurable change in risk. There is regarding dietary intervention strategies and that
good evidence to indicate that not all increases in `This confusion is aggravated by the production of
plasma total cholesterol levels are related to in- diet sheets and cookery books designed to provide
creased CHD risk and that changes in atherogenic low-cholesterol rather than cholesterol-lowering diets.
and anti-atherogenic lipoprotein particles are the The emphasis on the role of dietary cholesterol tends
major determinant of changes in CHD risk. to diminish the impact of more important recom-
The original rationale for a dietary cholesterol re- mendations about the types of fat in the diet.'
striction was based on three observations: simple
regression analysis of cross-cultural epidemiological
data showing a positive relationship between choles-
References
terol intake and CHD incidence; animal studies
showing that, in some species, dietary cholesterol in-
[1] N. Anitschkow, S. Chalatow, Ueber Experimentelle Choles-
duced hypercholesterolemia and atherosclerotic le- terinsteatose und ihre Bedeutung fu«r die Entstehung einiger
sions; and metabolic ward experiments demonstrat- pathologischer Prozesse, Zentralbl. Allg. Pathol. Anat. 24
ing that high intakes of cholesterol increased plasma (1913) 1^9.
cholesterol levels. Thirty years ago these observations [2] F.B. Hu, M.J. Stampfer, J.E. Manson, E. Rimm, G.A. Col-
were the basis for a recommendation that dietary ditz, B.A. Rosner, C.H. Hennekens, W.C. Willett, Dietary
fat intake and the risk of coronary heart disease in women,
cholesterol be limited to less than 300 mg/day. In
New Engl. J. Med. 337 (1997) 1491^1499.
large part this recommendation was based on the [3] A. Ascherio, E.B. Rimm, E.L. Giovannucci, D. Spiegelman,
`precautionary principle' which suggests that when M. Stampfer, W.C. Willett, Dietary fat and risk of coronary
information about risk is uncertain, it is prudent to heart disease in men: cohort follow up study in the United
assume the worst. Today the dietary cholesterol re- States, Br. Med. J. 313 (1996) 84^90.
striction is widely accepted even though there are in [4] US Department of Agriculture/US Department of Health
and Human Services, Nutrition and Your Health: Dietary
fact limited data to support it, and considerable data Guidelines for Americans, US Government Printing O¤ce,
accumulated over the last quarter century which con- Washington, DC, 1995.
tradict it. The recommendation persists in large part [5] National Research Council, Food and Nutrition Board,
because it is an established part of public health pol- Commission on Life Sciences, Diet and Health: Implications
icy. As noted by Dr. Walter Willett [65] regarding for Reducing Chronic Disease Risk, National Academy
Press, Washington, DC, 1989.
public confusion with ever changing diet and health
[6] National Cholesterol Education Program, Report of the Ex-
issues: `One of the problems is that strong recom- pert Panel on Population Strategies for Blood Cholesterol
mendations have often been made on very weak Reduction: executive summary, Arch. Intern. Med. 151
data. It may have been the best guess at the moment. (1191) 1071^1084.

BBAMCB 55729 28-11-00

 D.J. McNamara / Biochimica et Biophysica Acta 1529 (2000) 310^320 319

[7] R.M. Krauss, R.J. Deckelbaum, N. Ernst, E. Fisher, B.V. Kevaney, Diet and 20-year mortality from coronary heart
Howard, R.H. Knopp, T. Kotchen, A.H. Lichtenstein, H.C. disease, New Engl. J. Med. 312 (1985) 811^818.
McGill, T.A. Pearson, T.E. Prewitt, N.J. Stone, L. Van [21] D.R. Bassett, M. Abel, R.C. Moellering Jr., G. Rosenblatt,
Horn, R. Weinberg, Dietary guidelines for healthy American J.D. Stokes, Coronary heart disease in Hawaii: dietary in-
adults ^ A statement for health professionals from the Nu- take, depot fat, `stress', smoking, and energy balance in Ha-
trition Committee, American Heart Association, Circulation waiian and Japanese men, Am. J. Clin. Nutr. 22 (1969)
94 (1996) 1795^1800. 1483^1503.
[8] The Surgeon General's Report on Nutrition and Health, [22] A. Finegan, N. Hickey, B. Maurer, R. Mulcahy, Diet and
DHHS, Washington, DC, 1988. coronary heart disease: dietary analysis on 100 male pa-
[9] L.L. Rudel, Genetic factors in£uence the atherogenic re- tients, Am. J. Clin. Nutr. 21 (1968) 143^148.
sponse of lipoproteins to dietary fat and cholesterol in non- [23] K. Yano, G.G. Rhoads, A. Kagan, J. Tillotson, Dietary
human primates, J. Am. Coll. Nutr. 16 (1997) 306^312. intake and the risk of coronary heart disease in Japanese
[10] D. Kromhout, A. Menotti, B. Bloemberg, C. Aravanis, H. men living in Hawaii, Am. J. Clin. Nutr. 31 (1978) 1270^
Blackburn, R. Buzina, A.S. Dontas, F. Fidanza, S. Giam- 1279.
paoli, A. Jansen et al., Dietary saturated and trans fatty [24] W. Willett, Nutritional Epidemiology, 2nd edn., Oxford
acids and cholesterol and 25-year mortality from coronary University Press, New York, 1998.
heart disease: the Seven Countries Study, Prev. Med. 24 [25] G.F. Watts, P. Jackson, S. Mandalia, J.N. Brunt, E.S.
(1995) 308^315. Lewis, D.J. Coltart, B. Lewis, Nutrient intake and progres-
[11] D.M. Hegsted, L.M. Ausman, Diet, alcohol and coronary sion of coronary artery disease, Am. J. Cardiol. 73 (1994)
heart disease in men, J. Nutr. 118 (1988) 1184^1189. 328^332.
[12] P. Pietinen, A. Ascherio, P. Korhonen, A.M. Hartman, [26] K.L. Esrey, L. Joseph, S.A. Grover, Relationship between
W.C. Willett, D. Albanes, J. Virtamo, Intake of fatty acids dietary intake and coronary heart disease mortality: lipid
and risk of coronary heart disease in a cohort of Finnish research clinics prevalence follow-up study, J. Clin. Epide-
men ^ The alpha-tocopherol, beta-carotene cancer preven- miol. 49 (1996) 211^216.
tion study, Am. J. Epidemiol. 145 (1997) 876^887. [27] A. Tzonou, P. Lagiou, A. Trichopoulou, V. Tsoutsos, D.
[13] M. Toeller, A.E. Buyken, G. Heitkamp, W.A. Scherbaum, Trichopoulos, Dietary iron and coronary heart disease
H.M.J. Krans, J.H. Fuller, EURODIAB IDDM Complica- risk: a study from Greece, Am. J. Epidemiol. 147 (1998)
tions Group. Associations of fat and cholesterol intake with 161^166.
serum lipid levels and cardiovascular disease: the EURO- [28] G.S. Tell, G.W. Evans, A.R. Folsom, T. Shimakawa, M.A.
DIAB IDDM Complications Study, Exp. Clin. Endocrinol. Carpenter, G. Heiss, Dietary fat intake and carotid artery
Diabetes 107 (1999) 512^521. wall thickness: the Atherosclerosis Risk in Communities
[14] U. Ravnskov, Quotation bias in reviews of the diet-heart (ARIC) Study, Am. J. Epidemiol. 139 (1994) 979^989.
idea, J. Clin. Epidemiol. 48 (1995) 713^719. [29] J.I. Mann, P.N. Appleby, T.J. Key, M. Thoorogood, Diet-
[15] W.B. Kannel, T. Gordon, The Framingham diet study: diet ary determinants of ischaemic heart disease in health con-
and the regulation of serum cholesterol, in: The Framing- scious individuals, Heart 8 (1997) 450^455.
ham Study. An Epidemiological Investigation of Cardiovas- [30] C.M. Burch¢el, D.M. Reed, J.P. Strong, D.S. Sharp, P.H.
cular Disease, Section 241970, Washington, DC, 1970. Chyou, B.L. Rodriguez, Predictors of myocardial lesions in
[16] M.R. Garcia-Palmieri, P. Sorlie, J. Tillotson, R. Costas Jr., men with minimal coronary atherosclerosis at autopsy. The
E. Cordero, M. Rodriguez, Relationship of dietary intake to Honolulu heart program, Ann. Epidemiol. 6 (1996) 137^
subsequent coronary heart disease incidence: the Puerto 146.
Rico Heart Health Program, Am. J. Clin. Nutr. 33 (1980) [31] L.C. Lyu, M.J. Shieh, B.M. Posner, J.M. Ordovas, J.T.
1818^1827. Dwyer, A.H. Lichtenstein, L.A. Cupples, G.E. Dallal,
[17] T. Gordon, A. Kagan, M. Garcia-Palmieri, W.B. Kannel, P.W. Wilson, E.J. Schaefer, Relationship between dietary
W.J. Zukel, J. Tillotson, P. Sorlie, M. Hjortland, Diet and intake, lipoproteins, and apolipoproteins in Taipei and Fra-
its relation to coronary heart disease and death in three mingham, Am. J. Clin. Nutr. 60 (1994) 765^774.
populations, Circulation 63 (1981) 500^515. [32] Federation of American Societies for Experimental Biology,
[18] D.L. McGee, D.M. Reed, K. Yano, A. Kagan, J. Tillotson, Report on Nutrition Monitoring in the United States, vol. 1,
Ten-year incidence of coronary heart disease in the Honolulu US Government Printing O¤ce, Washington DC, 1995.
Heart Program. Relationship to nutrient intake, Am. J. Epi- [33] T.R. Dawber, R.J. Nickerson, F.N. Brand, J. Pool, Eggs,
demiol. 119 (1984) 667^676. serum cholesterol, and coronary heart disease, Am. J. Clin.
[19] D. Kromhout, C. de Lezenne Coulander, Diet, prevalence Nutr. 36 (1982) 617^625.
and 10-year mortality from coronary heart disease in 871 [34] A. Gramenzi, A. Gentile, M. Fasoli, E. Negri, F. Parazzini,
middle-aged men. The Zutphen Study, Am. J. Epidemiol. C. La Vecchia, Association between certain foods and risk of
119 (1984) 733^741. acute myocardial infarction in women, Br. Med. J. 300
[20] L.H. Kushi, R.A. Lew, F.J. Stare, C.R. Ellison, M. el Lozy, (1990) 771^773.
G. Bourke, L. Daly, I. Graham, N. Hickey, R. Mulcahy, J. [35] G.E. Fraser, Diet and coronary heart disease: beyond diet-

BBAMCB 55729 28-11-00

320 D.J. McNamara / Biochimica et Biophysica Acta 1529 (2000) 310^320

ary fats and low-density-lipoprotein cholesterol, Am. J. Clin. [53] W.H. Howell, D.J. McNamara, M.A. Tosca, B.T. Smith,
Nutr. 59 (1994) 1117S^1123S. J.A. Gaines, Plasma lipid and lipoprotein responses to diet-
[36] F.B. Hu, M.J. Stampfer, E.B. Rimm, J.E. Manson, A. ary fat and cholesterol: a meta-analysis, Am. J. Clin. Nutr.
Ascherio, G.A. Colditz, B.A. Rosner, D. Spiegelman, F.E. 65 (1997) 1747^1764.
Speizer, F.R. Sacks, C.H. Hennekens, W.C. Willett, A pro- [54] D.J. McNamara, Eggs, dietary cholesterol and heart disease
spective study of egg consumption and risk of cardiovascular risk: an international perspective, in: J.S. Sim, S. Nakai, W.
disease in men and women, J. Am. Med. Assoc. 281 (1999) Guenter (Eds.), Egg Nutrition and Biotechnology, CABI
1387^1394. Publishing, New York, 1999, pp. 55^63.
[37] American Heart Association, Heart and Stroke Statistical [55] H.C. McGill Jr., The relationship between dietary cholester-
Update, American Heart Association, Dallas, TX, 1998. ol to serum cholesterol concentration and to atherosclerosis
[38] R.B. Shekelle, J. Stamler, Dietary cholesterol and ischaemic in man, Am. J. Clin. Nutr. 32 (1979) 2664^2702.
heart disease, Lancet i (1989) 1177^1178. [56] R.M. Weggemans, P.L. Zock, R. Urgert, M.B. Katan, Dif-
[39] D.L. Tribble, Antioxidant consumption and risk of coronary ferences between men and women in the response of serum
heart disease: emphasis on vitamin C, vitamin E, and beta- cholesterol to dietary changes, Eur. J. Clin. Invest. 29 (1999)
carotene ^ A statement for healthcare professionals from the 827^834.
American Heart Association, Circulation 99 (1999) 591^595. [57] H.N. Ginsberg, W. Karmally, M. Siddiqui, S. Holleran,
[40] M.R. Malinow, P.B. Duell, D.L. Hess, P.H. Anderson, A.R. Tall, S.C. Rumsey, R.J. Deckelbaum, W.S. Blaner,
W.D. Kruger, B.E. Phillipson, R.A. Gluckman, P.C. Block, R. Ramakrishnan, A dose-response study of the e¡ects of
B.M. Upson, Reduction of plasma homocyst(e)ine levels by dietary cholesterol on fasting and postprandial lipid and
breakfast cereal forti¢ed with folic acid in patients with cor- lipoprotein metabolism in healthy young men, Arterioscler
onary heart disease, New Engl. J. Med. 338 (1998) 1009^ Thromb. 14 (1994) 576^586.
1015. [58] H.N. Ginsberg, W. Karmally, M. Siddiqui, S. Holleran,
[41] M.R. Malinow, A.G. Bostom, R.M. Krauss, Homocys- A.R. Tall, W.S. Blaner, R. Ramakrishnan, Increases in diet-
t(e)ine, diet, and cardiovascular diseases ^ A statement for ary cholesterol are associated with modest increases in both
healthcare professionals from the Nutrition Committee, LDL and HDL cholesterol in healthy young women, Arte-
American Heart Association, Circulation 99 (1999) 178^182. rioscler Thromb. 15 (1995) 169^178.
[42] S.M. Artaud-Wild, S.L. Connor, G. Sexton, W.E. Connor, [59] R.H. Knopp, B.M. Retzla¡, C.E. Walden, A.A. Dowdy,
Di¡erences in coronary mortality can be explained by di¡er- C.H. Tsunehara, M.A. Austin, T. Nguyen, A double-blind,
ences in cholesterol and saturated fat intakes in 40 countries randomized, controlled trial of the e¡ects of two eggs per
but not in France and Finland. A paradox, Circulation 88 day in moderately hypercholesterolemic and combined hy-
(1993) 2771^2779. perlipidemic subjects taught the NCEP step I diet, J. Am.
[43] F. Grande, J.T. Anderson, C. Chlouverakis, M. Proja, A. Coll. Nutr. 16 (1997) 551^561.
Keys, E¡ect of dietary cholesterol on man's serum lipids, [60] P.J. Nestel, Dietary cholesterol and plasma lipoproteins,
J. Nutr. 87 (1965) 52^62. Ann. NY Acad. Sci. 676 (1993) 1^10.
[44] D.M. Hegsted, R.B. McGandy, M.L. Myers, F.J. Stare, [61] E.E. Zanni, V.I. Zannis, C.B. Blum, P.N. Herbert, J.L. Bres-
Quantitative e¡ects of dietary fat on serum cholesterol in low, E¡ect of egg cholesterol and dietary fats on plasma
man, Am. J. Clin. Nutr. 17 (1965) 281^295. lipids, lipoproteins, and apoproteins of normal women con-
[45] D.M. Hegsted, Serum-cholesterol response to dietary choles- suming natural diets, J. Lipid Res. 28 (1987) 518^527.
terol: a re-evaluation, Am. J. Clin. Nutr. 44 (1986) 299^305. [62] E.A. Fisher, C.B. Blum, V.I. Zannis, J.L. Breslow, Inde-
[46] P.N. Hopkins, E¡ects of dietary cholesterol on serum cho- pendent e¡ects of dietary saturated fat and cholesterol on
lesterol: a meta-analysis and review, Am. J. Clin. Nutr. 55 plasma lipids, lipoproteins, and apolipoprotein E, J. Lipid
(1992) 1060^1070. Res. 24 (1983) 1039^1048.
[47] A. Keys, Serum cholesterol response to dietary cholesterol, [63] P.M. Clifton, P.J. Nestel, E¡ect of dietary cholesterol on
Am. J. Clin. Nutr. 40 (1984) 351^359. postprandial lipoproteins in three phenotypic groups, Am.
[48] D.J. McNamara, Relationship between blood and dietary J. Clin. Nutr. 64 (1996) 361^367.
cholesterol, Adv. Meat Res. 6 (1990) 63^87. [64] A. Blanco-Molina, G. Castro, D. Martin-Escalante, D.
[49] D.J. McNamara, Dietary cholesterol: e¡ects on lipid metab- Bravo, J. Lopez-Miranda, P. Castro, F. Lopez-Segura, J.C.
olism, Curr. Opin. Lipidol. 1 (1990) 18^22. Fruchart, J.M. Ordovas, F. Perez-Jimenez, E¡ects of di¡er-
[50] D.J. McNamara, Dietary cholesterol and the optimal diet ent dietary cholesterol concentrations on lipoprotein plasma
for reducing risk of atherosclerosis, Can. J. Cardiol. 11 concentrations and on cholesterol e¥ux from Fu5AH cells,
(Suppl. G) (1995) 123G^126G. Am. J. Clin. Nutr. 68 (1998) 1028^1033.
[51] D.J. McNamara, Cholesterol intake and plasma cholesterol: [65] D.Q. Haney, Con£icting Studies Confuse Doctors, Associ-
an update, J. Am. Coll. Nutr. 16 (1997) 530^534. ated Press, Boston, MA, 21 April 2000.
[52] R. Clarke, C. Frost, R. Collins, P. Appleby, R. Peto, Dietary [66] A.G. Shaper, J.W. Marr, Dietary recommendations for the
lipids and blood cholesterol: quantitative meta-analysis of community towards the postponement of coronary heart dis-
metabolic ward studies, Br. Med. J. 314 (1997) 112^117. ease, Br. Med. J. 1 (1977) 867^871.

BBAMCB 55729 28-11-00