REVIEW ARTICLE

Dietary cholesterol and the risk of cardiovascular disease

in patients: a review of the Harvard Egg Study and other
data
P. J. H. Jones

Canada Research Chair in

SUMMARY
What’s known Nutrition and Functional Foods,
For many years, both the medical community and the general public have incor- Richardson Centre for
Dietary factors influence CHD risk.
rectly associated eggs with high serum cholesterol and being deleterious to health, Functional Foods and
even though cholesterol is an essential component of cells and organisms. It is What’s new Nutraceuticals, University of
Dietary cholesterol level plays a minor role in Manitoba, Winnipeg, MB,
now acknowledged that the original studies purporting to show a linear relation
Canada
between cholesterol intake and coronary heart disease (CHD) may have contained altering CHD risk.
fundamental study design flaws, including conflated cholesterol and saturated fat Correspondence to:
consumption rates and inaccurately assessed actual dietary intake of fats by study P. J. H. Jones,
Canada Research Chair in
subjects. Newer and more accurate trials, such as that conducted by Frank B. Hu
Nutrition and Functional Foods,
of the Harvard School of Public Health (1999), have shown that consumption of Richardson Centre for
up to seven eggs per week is harmonious with a healthful diet, except in male Functional Foods and
patients with diabetes for whom an association in higher egg intake and CHD was Nutraceuticals, University of
Manitoba, Winnipeg, MB,
shown. The degree to which serum cholesterol is increased by dietary cholesterol
Canada
depends upon whether the individual’s cholesterol synthesis is stimulated or down- Tel.: 204 474 8883
regulated by such increased intake, and the extent to which each of these phe- Fax: 204 474 2676
nomena occurs varies from person to person. Several recent studies have shed Email: peter_jones@
umanitoba.ca
additional light on the specific interplay between dietary cholesterol and cardiovas-
cular health risk. It is evident that the dynamics of cholesterol homeostasis, and of Disclosure
development of CHD, are extremely complex and multifactorial. In summary, the None.
earlier purported adverse relationship between dietary cholesterol and heart dis-
ease risk was likely largely over-exaggerated.

that eggs are associated with high serum cholesterol

Dispersing the fog
In recent years, the prevalence of cancers has declined
significantly (1), signalling an overall improvement in
health and medical knowledge. By contrast, cardio-
vascular disease remains the leading cause of morbid-
ity and mortality in North America (2). The World
Health Organization continues to rate coronary heart
disease (CHD) and stroke as the two leading causes
of death in developed countries. These two disorders
also remain as two of the top five causes of death in
underdeveloped countries (3). Thus it is particularly
important to understand the factors that contribute
to CHD and to tailor all prevention and management
strategies accordingly, encompassing all areas of med-
ication, diet and exercise.
Early research focusing on dietary cholesterol and
cardiovascular disease often used eggs as a test
product because of their naturally high cholesterol
content. As a result, a misperception has endured
and are deleterious to health.
However, cholesterol is an essential molecule for
life. It is an integral part of cell membranes, is an
important regulator for many hormones and aids in
digestion. In addition, there is very little evidence
supporting the association of dietary cholesterol with
serum cholesterol and development of cardiovascular
disease. A recent assessment of the early studies pur-
porting to show a link between dietary cholesterol
and cardiovascular disease highlights the fundamen-
tal flaws in their designs.
Marion Volk, a naturopath and faculty member at
Charles Sturt University in Bathurst, New South
Wales, Australia, has examined early studies that
concluded a linear correlation existed between
consumption of saturated fat and cholesterol and
death from heart disease (4). She conducted a litera-
ture search of studies published in English before
1994 – because after 1994 most trials included a sta-

ª 2009 The Author

Journal compilation ª 2009 Blackwell Publishing Ltd Int J Clin Pract, October 2009, 63 (Suppl. 163), 1–8
doi: 10.1111/j.1742-1241.2009.02136.x 1
2 Risk of cardiovascular disease in patients

tin control, which was not a feature of earlier studies
– on cholesterol levels of fat in diets and heart
disease in adults. Volk whittled the resulting 2669
studies down to 16 highest-quality, cross- and
within-population studies by applying stringent crite-
ria such as validity, high quality of study design and
random selection of participants.
Two fundamental studies in the field led by Ancel
Keys are included in Volk’s analysis: a pilot, six-
country study conducted in 1949 (5), and a seven-
country study conducted between 1958 and 1964 (6).
Keys was the first to propose a relation between die-
tary and serum cholesterol levels and CHD mortality
in adults. The paper on his six-country study
included a figure showing the relation between the
national death rates for men aged 45–49 and 55–
59 years from heart disease, and the proportion of
fat calories available from the respective national
diets. The figure displayed a regular progression of
fat consumption and increased heart disease from
Japan to Italy, Sweden, England, Wales, Canada,
Australia and the United States.
However, Volk pointed out that Keys included no
information about the basis upon which he selected
only these six countries, when data were available
from a total of 22 countries. Furthermore, the esti-
mate of the proportion of fat in the overall diet was
based on the food and fat available for consumption
in each country (i.e. the estimate of fat in the diet
was based on the national production of fat plus
import, minus food not for human consumption),
rather than on food and fat actually eaten. Thirdly,
the mortality data were derived from a variety of
definitions, reporting patterns, diagnostic habits and
inter-region standards.
To validate his hypothesis, keys next examined 25-
year mortality from cardiovascular disease and serum
cholesterol in 16 local populations from seven
countries: The Netherlands, Yugoslavia, Finland,
Japan, Greece, Italy and the United States. He
reported a robust correlation between serum choles-
terol and death rates from heart disease. However,
this study also contained unexplained discrepancies,
such as the significant differences found in death
rates from one region of a country to another, and
the lack of mortality rate increases in accordance
with serum cholesterol level increases over the
25-year period of observation.
Volk concluded that ‘these results neither sup-
port nor invalidate the lipid hypothesis; they pres-
ent a long-term follow up of an ageing population.
The validity of the results is restricted to the par-
ticular population and the specific historical period
during which they were studied. The results cannot
be generalised or taken as a basis for other stud-
ies’. The Framingham Heart Study, an ongoing
study involving adults residing in Framingham,
Massachusetts, USA, was also reviewed. Volk’s criti-
cisms of this widely quoted study are listed in
Table 1.
The Framingham study did not demonstrate a lin-
ear correlation between serum cholesterol levels and
risk of CHD (4). In addition, other researchers found
that the average values for risk factors used in the
Framingham publications were about twice those of
studies from northern Europe (7). Despite the defi-
ciencies of this study, the National Cholesterol Edu-
cation Program has made extensive use of
Framingham study data to develop its strategy for
preventing heart disease through the maintenance of
low cholesterol levels (8).
Volk stated that the continued use of conclusions
from the Framingham studies is a questionable prac-
tice and ‘may in part explain why there is still such
debate about the connection between both dietary
and serum lipids and CHD. …The many problems
associated with research into fats and CHD make it
logical to conclude that the lipid hypothesis of ath-
erosclerosis is based on several false premises, includ-
ing linear causation, fallacious national mortality
statistics, biased age and subject selection and meth-
odological inaccuracies’ (4).

Table 1 Possible shortcomings of the Framingham Heart Study (4)

A lack of information on subjects who refused to participate or dropped out

A lack of consistency in the publication of results (i.e. different publications of the results contain inconsistent numbers of subjects, age
ranges and time frames)
Families were invited to join the study, thus raising the possibility of an aggregation of genetic, physical or psychological characteristics
– for example, familial hypercholesterolaemia should have been a concern – as well as lifestyle and dietary habits
The wide age distribution of the subjects resulted in the risk calculation for age groups spanning several decades – even though heart
disease has various manifestations, each with marked age dependence
The study failed to show an association of serum cholesterol and cardiovascular disease in individuals over 50 years, despite this being
the part of the population with the highest prevalence of this disease

ª 2009 The Author

Journal compilation ª 2009 Blackwell Publishing Ltd Int J Clin Pract, October 2009, 63 (Suppl. 163), 1–8

Harvard egg study
A more definitive study was published on the topic
of dietary cholesterol and cardiovascular disease in
men and women. This analysis, led by Frank Hu and
other researchers at the Harvard School of Public
Health examined egg consumption and its relation
to the risk of cardiovascular disease in men and
women. The study was published in 1999 in the
Journal of the American Medical Association (9) and
is colloquially referred to as the Harvard Egg Study.
In this study, investigators examined data from
the Health Professionals Follow-up Study and the
Nurses’ Health Study. Their analysis was funded by
the National Institutes of Health. The Health Profes-
sionals Follow-up Study was initiated in 1986 with
the completion of a detailed questionnaire by 51,529
male American dentists, optometrists, pharmacists,
podiatrists and veterinarians aged 40–75 years. Ques-
tions within this comprehensive diet survey included
how often the men ate meals at home vs. a restau-
rant, and what kind of fats were used at home for
frying and baking; items on lifestyle practice such as
whether they had smoked 20 or more packs of ciga-
rettes in their lifetime and, if yes, whether they cur-
rently smoked; as well as medical history. Follow-up
questionnaires were sent to participants in 1988,
1990, 1992 and 1994. Hu and his co-investigators
analysed responses from 37,851 of these men. These
were the respondents who did not have diagnoses at
baseline of cardiovascular disease, cancer, diabetes or
hypercholesterolaemia, reported daily energy intakes
of between 3360 and 17,640 kJ (between 800 and
4200 calories), and answered ‡ 60 of the 131 food-
related questions on the questionnaires.
The Nurses’ Health Study was established in
1976 when 121,700 female registered nurses (aged
30–55 years) in the United States provided detailed
information about their medical history and lifestyle
characteristics, including whether they had a history
of diabetes, elevated cholesterol or a myocardial
Table 2 Harvard Egg Study: coronary heart disease and stroke risk according to category of egg consumption (8)
Number of Men – multivariate
eggs consumed relative risk for
per week CHD (95% CI)
<1 1.0
1 1.06 (0.88–1.27)
2–4 1.12 (0.95–1.33)
5–6 0.90 (0.63–1.27)
‡7 1.08 (0.79–1.48)
CHD, coronary heart disease.
ª 2009 The Author
Journal compilation ª 2009 Blackwell Publishing Ltd Int J Clin Pract, October 2009, 63 (Suppl. 163), 1–8
Risk of cardiovascular disease in patients 3
infarction, or if they had ever used oral contracep-
tives. Every 2 years they received follow-up ques-
tionnaires, with the 1980 questionnaire including 61
items on food frequency. The list was expanded to
116 items in 1984, with similar questions being
used in 1986 and 1990. Hu et al. analysed data
from 80,082 women. These were respondents who
did not leave ‡ 10 items blank; did not have
implausibly high or low scores for total food or
energy intake; and did not have previously diag-
nosed cancer, cardiovascular disease, hypercholeste-
rolaemia or diabetes.
The investigators pooled the survey and outcome
data and performed univariate and multivariate
logistic regression analyses. Adjustments in the multi-
variate analyses were made for age, total energy
intake, smoking, alcohol consumption, history of
hypertension, parental history of myocardial infarc-
tion, body mass index, current multivitamin use and
vitamin E supplement use. In addition, among
women an adjustment was made for menopausal
status, postmenopausal hormone use, and regular
vigorous exercise; among men an adjustment was
made for physical activity.
The results revealed no overall association between
egg consumption and risk of CHD or stroke. Even at
the highest level of egg consumption – at least one
per day – the multivariate relative risk for CHD
among both men and women was 1.08 [95% confi-
dence interval (CI) 0.63–1.27]. The multivariate
relative risk for stroke among men was 1.07 (95% CI:
0.66–1.75) and among women it was 0.89 (95% CI:
0.60–1.31); the risk remained insignificant when
divided into ischaemic or haemorrhagic stroke.
Strokes caused by infection or neoplasia were not
included. These results are shown in Table 2.
One subgroup demonstrated an increased risk of
CHD with high egg consumption: diabetic men who
consumed at least one egg a day. These subjects had
a multivariate relative risk of CHD of 2.02 (95% CI:
1.05–3.87; p = 0.04). The investigators hypothesised
Women – multivariate Men – multivariate Women – multivariate
relative risk for CHD relative risk for relative risk for stroke
(95% CI) stroke (95% CI) (95% CI)
1.0 1.0 1.0
0.82 (0.67–1.00 1.06 (0.76–1.49) 0.89 (0.70–1.13)
0.99 (0.82–1.18) 0.95 (0.69–1.31) 0.83 (0.66–1.05)
0.95 (0.70–1.29) 1.43 (0.85–2.43) 0.89 (0.60–1.32)
0.82 (0.60–1.13) 1.07 (0.66–1.75) 0.89 (0.60–1.31)
4 Risk of cardiovascular disease in patients
that this heightened risk may have been related to
abnormal cholesterol transport as result of the
decreased levels of apolipoprotein E and increased
levels of apolipoprotein III.
Overall, the team concluded it is possible that ‘the
small adverse effect of cholesterol in an egg on plasma
low-density lipoprotein (LDL) levels is
counterbalanced by potential beneficial effects on
high-density lipoprotein and triglycerides, and of
other nutrients including antioxidants, folate other B
vitamins and unsaturated fats’. The investigators’
conclusions support those from several other
well-controlled trials, including the California Sev-
enth-Day Adventist Study (N = 26,473), which dem-
onstrated that consuming a diet containing at least
three eggs a week or cheese at least three times a week
did not increase the risk of coronary events (10).
Do dietary cholesterol levels affect
serum cholesterol levels?
Is the lack of correlation between dietary cholesterol
and heart disease due to dietary cholesterol not mod-
ulating serum cholesterol, or to serum cholesterol
having no effect on an individual’s risk for heart dis-
ease? There are various ways to attempt to answer
this question. Herron et al. (11) at the University of
Connecticut studied the influence of diets high or
low in cholesterol on the atherogenicity of LDL
particles. The study was supported by the American
Egg Board and the University of Connecticut
Research Foundation. The participants were men
aged 20–50 (n = 40) and premenopausal women
(n = 51) who did not have hypercholesterolaemia,
hypertriglyceridaemia, hypertension or diabetes; and
who were not receiving lipid-lowering drugs. Subjects
were randomly assigned to either consumption of
the liquid equivalent of three whole eggs per day –
adding approximately 640 mg cholesterol ⁄ day to
their diets – or an identical weight of cholesterol-
and fat-free egg substitute. Subjects were also
instructed to adhere to the National Cholesterol Edu-
cation Programme Step 1 Diet, which stipulates that
no more than 30% of total energy should be derived
from fat, and that saturated fat should comprise just
10% of the fat. Participants were also instructed not
to consume more than 300 mg ⁄ day of dietary choles-
terol aside from their egg ⁄ placebo intake. Plasma
samples were analysed from 25 of the men and 27 of
the women.
In this study, 14 men and 14 women had serum
cholesterol levels that were increased by more than
the average in response to the high-cholesterol diet.
Another 13 men and 13 women had serum choles-
terol levels that were not influenced at all by the high
Journal compilation ª 2009 Blackwell Publishing Ltd Int J Clin Pract, October 2009, 63 (Suppl. 163), 1–8
cholesterol intake (Figure 1). These are known as
hyper- and hypo-responders respectively. Results of
this study demonstrated that increases of serum cho-
lesterol induced by changing dietary cholesterol may
depend on the prevalence of hyper- or hypo-
responders in the population being examined (11).
Another study was conducted to ascertain whether
dietary cholesterol consumption can suppress choles-
terol synthesis, thus helping to explain the phenome-
non of hypo-responders. This study also explored
whether a ceiling effect exists if cholesterol synthesis
suppression occurs (12). This second study recruited
30 men and women who were not taking medica-
tions that affected lipid metabolism. Seven subjects
had low serum total cholesterol levels (< 190 mg ⁄ dl),
12 had normal total cholesterol levels (190–
250 mg ⁄ dl) and 11 were hypercholesterolaemic
(> 250 mg ⁄ dl). Subjects consumed, in random order,
diets containing 50, 350 or 650 mg choles-
terol ⁄ 2800 kcal. Cholesterol was provided as egg
yolks and whole eggs incorporated into foods such as
custards, omelettes and cookies. In the 50 mg choles-
terol ⁄ kcal diet, the egg yolk was partly or completely
replaced with egg whites and a mixture of fats simi-
lar in fatty-acid composition to that in egg yolk.
Subjects remained on each diet for 4 weeks before
entering a 4-week washout period and then consum-
ing another diet. The investigators also measured
whole-body cholesterol biosynthesis by giving sub-
jects water incorporating deuterium in the last week
of each diet and subsequently measuring deuterium
incorporation into the serum cholesterol and by
measuring urinary mevalonate levels (an index of
cholesterol-synthesis rate). The study was funded by
the Heart and Stroke Foundation of British
Columbia and the Yukon, and by the National Insti-
tutes of Health (12).
LDL-Egg LDL-placebo
4 HDL-Egg HDL-placebo
** **
NS
Cholesterol (mmol/L)
NS
3
** NS
2
NS
**
1
0
Men Women Men Women
Hyper-Responders Hypo-Responders
Figure 1 Effect of high-cholesterol and placebo diets on
LDL- and HDL-cholesterol levels among male and female
hyper- and hypo-responders to dietary cholesterol (11)
ª 2009 The Author
 Risk of cardiovascular disease in patients 5

Some of the main results are shown in Figure 2, (LDL-C) concentrations when on the higher-choles-
which is a modified and updated version of the data terol diets, but there were no significant differences
presented in the original paper (12). The cholesterol between the groups. The other main results are
hyper-responders appear to have experienced an shown in Figure 3. These results show that increased
increase in total cholesterol and LDL-cholesterol dietary cholesterol consumption depressed choles-
terol synthesis. There was a significant reduction in
deuterium uptake and hence cholesterol synthesis
400 between low- and medium-cholesterol diets, and
Total Chol (mg/dL)

Means + SEM
350
300
between low- and high-cholesterol diets (p < 0.05).
250 Moreover, there also was a significant decrease in
200 urinary mevalonate excretion between low- and
150
high-cholesterol diets, and between medium- and
100
50 high-cholesterol diets (p < 0.01).
0 It therefore can be concluded that there is a trend
for serum cholesterol levels to increase with increases
70
60
in dietary cholesterol, and for mechanisms that regu-
HDL-C (mg/dL)

50 late the serum levels to become activated at higher

40 dietary cholesterol levels, but the extent to which
30 each of these phenomena takes place is relatively
20 modest and appears to vary from individual to
10 individual.
0

250
What factors modulate the effect of
LDL-C (mg/dL)

200 serum cholesterol on cardiovascular

150 disease risk?
100 As with dietary cholesterol levels modulating serum
50 cholesterol levels, individualised effects may operate
0 at the nexus between serum cholesterol and cardio-
Low Medium High vascular disease. Half of patients with heart attacks
Dietary Cholesterol Level have low-cholesterol levels, which is indicative that
Hypoc Normoc Hyperc many factors may be involved (13). It is well-estab-
Figure 2 Effect of dietary cholesterol level on total lished that obesity, other metabolic factors and
cholesterol, HDL- and LDL-cholesterol among female and smoking are among the most important influences
male hypo- and hyper-responders to dietary cholesterol on the incidence of CHD (14). Several recent studies
(12) are continuing to shed light on the specific interplay

Deuterium Incorporation Urinary mevalonate

Urinary mevalonate (micromoles/day)

0.10 3.00
Synthesis (% per day)

0.08 2.40

0.06 1.80

0.04 1.20

0.02 0.60

0.00 0.00
Hypochol Normochol Hyperchol Hypochol Normochol Hyperchol
Subject Type
50 mg/d 350 mg/d 650 mg/d

Figure 3 Effect of dietary cholesterol level on cholesterol biosynthesis indicators over 24 h in hypo- and hyper-responders
to dietary cholesterol (12)

ª 2009 The Author

Journal compilation ª 2009 Blackwell Publishing Ltd Int J Clin Pract, October 2009, 63 (Suppl. 163), 1–8
6 Risk of cardiovascular disease in patients

between dietary cholesterol and cardiovascular health significant increase in risk for myocardial infarction
risk. (Table 3). Qureshi et al. posit that the lack of rela-
A team from the University of Medicine and Den- tion between egg consumption and cardiovascular
tistry of New Jersey led by Adnan Qureshi et al. (15) diseases may be attributable to a lack of association
examined the association between egg consumption between serum cholesterol and egg consumption.
and the risk of cardiovascular diseases in a nationally Another recent study has shown that using eggs as
representative cohort of men and women aged 25– a significant part of the diet does not modify the
74 years. The study was funded by departments salubrious effects of a carbohydrate-reduced diet on
within the university. Qureshi et al. used data from metabolic factors. Thirty-one overweight or obese
the National Health Epidemiologic Follow-up Study, men aged 40–70 years who did not have hyperthy-
which followed subjects in the First National Health roidism, documented heart disease, type 1 diabetes,
and Nutrition Examination Survey between 1982 and gout or egg allergies were recruited and counselled to
1992. Of the 13,586 subjects who participated in the consume a carbohydrate-reduced diet (16). These
National Health Epidemiologic Follow-up Study, subjects each were randomly assigned to consume
10,717 answered questions regarding egg intake. either three eggs per day or the equivalent of zero-
Qureshi et al. excluded 983 subjects because of his- cholesterol egg substitute. Three individuals dropped
tory of stroke or heart attack; unknown daily egg out because of compliance issues.
consumption; or missing data on body mass index, The energy intake of subjects in both groups
serum cholesterol or systolic blood pressure. decreased and they also experienced decreased
A multivariate analysis of the data from the weight and waist circumference. Their plasma tri-
remaining 9734 subjects indicated a lack of relation glyceride concentration and LDL-C: high-density
between consumption of more than six eggs per lipoprotein cholesterol (HDL-C) ratios did not
week and stroke, coronary artery disease and mortal- change the study. However, plasma HDL-C concen-
ity, after adjustment for age, gender, race ⁄ ethnicity, tration increased in the group consuming three eggs
systolic blood pressure, diabetes, serum cholesterol, per day (Figure 4). These researchers suggested that
smoking, educational status and body mass index eggs may be a preferred food item in carbohydrate-
(Table 3). Only among people with diabetes who reduced diets.
consumed more than six eggs per week, there was a New data are also pointing to the ability of plant
sterols to reduce serum cholesterol levels, regardless
of levels of dietary cholesterol. A study published in
2003 in the Journal of the American Medical Associa-
Table 3 Weekly egg consumption and risk for stroke, tion set the baseline for these studies by demonstrat-
coronary artery disease and mortality in overall cohort ing that a diet high in plant sterols, soy protein,
and subjects with diabetes (12) viscous fibres and almonds has a similar effect on
Multivariate hyperlipidaemic adults’ lipid and C-reactive protein
Categories adjusted relative Multivariate adjusted levels as lovastatin (17). Subjects experienced a
according to risk (95% CI) – relative risk (95% CI) – decrease in LDL-C and C-reactive protein levels
weekly egg entire cohort subjects with diabetes compared with individuals on a diet low in saturated
consumption (N = 9734) (N = 349) fat.
Similar findings were reported in a study by Kassis
All stroke
et al. In this work reduced total cholesterol and
£ 1 egg 1.0 1.0
LDL-C levels were identified in individuals with low
1–6 eggs 0.9 (0.7–1.0) 1.1 (0.5–2.5)
baseline serum cholesterol levels who consumed rela-
> 6 eggs 0.9 (0.7–1.1) 0.6 (0.2–1.5)
Ischaemic stroke tively high cholesterol levels in the form of approxi-
£ 1 egg 1.0 1.0 mately one egg per day (18).
1–6 eggs 0.8 (0.7–1.0) 1.1 (0.4–2.1)
> 6 eggs 0.9 (0.7–1.1) 0.5 (0.2–1.4)
Conclusion
Coronary artery disease
£ 1 egg 1.0 1.0 Trials such as the Harvard Egg Study help clarify the
1–6 eggs 1.0 (0.9–1.1) 1.2 (0.7–2.3) relation between dietary cholesterol intake, serum
> 6 eggs 1.1 (0.9–1.3) 1.9 (1.0–3.5) cholesterol levels and incidence of CHD. Although
Mortality
this large and well-conducted study had its weak-
£ 1 egg 1.0 1.0
nesses and a large, prospective study in the field
1–6 eggs 0.9 (0.8–1.0) 1.0 (0.7–1.5)
remains to be undertaken, the evidence strongly indi-
> 6 eggs 1.0 (0.9–1.1) 1.3 (0.9–1.9)
cates that consumption of up to seven eggs per week

ª 2009 The Author

Journal compilation ª 2009 Blackwell Publishing Ltd Int J Clin Pract, October 2009, 63 (Suppl. 163), 1–8

75
EGG
65
Percent Change in HDL-C
55
45
35
25
15
5
–5
Baseline
Figure 4 Per cent change in plasma HDL-C concentrations from baseline to 12 weeks in overweight men who consumed a
carbohydrate-reduced diet including three eggs (‘EGG’, n = 15) or an egg substitute (‘SUB’, N = 13) (the grey line
indicates the mean of each group) (16)
is congruent with a healthy diet. Family practitioners
need to be current on these developments so that
they may advise their patients. For instance, includ-
ing eggs in a diet can be a beneficial way for patients
to get protein and other nutrients and it would be
unfortunate to deny patients this option as a result
of misconceptions regarding eggs and CHD. How-
ever, family physicians should caution males with
diabetes to limit their egg intake because an associa-
tion may exist in this population between high egg-
intake levels and CHD, although the mechanisms for
this remain unknown. Research should now begin
focusing on the more serious threats to cardiovascu-
lar health: the metabolic complications of obesity,
diabetes and metabolic syndrome.
Discussion
Q – Dr Constance: What are the weaknesses of the
Harvard Egg Study?
A – Dr Jones: The dietary intake assessment
approaches were flawed. This is shown by the fact
that as egg consumption per day increases, the
energy intake and the saturated fat intake per day
also increase. The investigators did not examine or
control for whether this was the result of people
inaccurately reporting food intake and, instead,
either under- or over-estimating their total food
intake, including eggs, energy and fat.
A – Dr Lau: Also, the baseline lipid profiles of the
subjects were not analysed. And the results may not
be generalisable to a larger population because the
subjects were largely middle class, Caucasian and
self-selected in the sense that they were relatively
healthy, normal individuals.
ª 2009 The Author
Journal compilation ª 2009 Blackwell Publishing Ltd Int J Clin Pract, October 2009, 63 (Suppl. 163), 1–8
Risk of cardiovascular disease in patients 7
40
SUB
30
20
10
0
–10
–20
–30
–40
–50
–60
12 wk Baseline 12 wk
Q – Dr Constance: Will our ability to detect hyper-
and hypo-responders one day be useful clinically in
counselling individuals regarding whether they
should be careful in the number of eggs they con-
sume?
A – Dr Jones: This is complicated by the fact that
people who have higher cholesterol absorption, who
are hence hyper-responders, tend to have a lower
cholesterol-synthesis rate. It’s really a complex story.
But the biggest, most important measure, and one
I’d like to underscore, isn’t around cholesterol and it
may not even be around saturated fat, but it is
around body mass index and taking your tape mea-
sure out. If your tape measure shows that your waist
line is too large, that is indicative of the highest
risk for CHD that you can find, particularly among
individuals with diabetes.
Q – Dr Lau: Are we sure that any increase in HDL-C
is beneficial and any decrease is detrimental with
respect to CHD? Because, for example, vegetarians
have lower HDL-C levels than non-vegetarians, but
they have lower levels of heart disease. Isn’t the pic-
ture of influences on HDL levels and effect on heart
disease, more complicated than that?
A – Dr Jones: Yes, I think one could look at that
study that shows that in carbohydrate-restricted diets
that HDL-C levels are boosted with a grain of salt,
because other things are probably more important
modulators of HDL. Certainly, exercise is an impor-
tant factor. You do not have to exercise that much
to see a shift in the profile of HDL-C particle size or
degree of cardioprotectiveness. HDL-C also seems to
go up with moderate alcohol consumption.
Q – Dr Lau: How does one interpret the
plethora of data showing inter-population and
8 Risk of cardiovascular disease in patients
inter-individual variability in responsiveness to die-
tary cholesterol?
A – Dr Jones: There exists significant variability even
within a population. For example, we think that by
giving infants formula without cholesterol, we may be
doing them a disservice, in that by giving breast milk,
the cholesterol in the breast milk turns down the
body’s manufacturing rate and keeps it there. There
are data that suggest this, and so it may be that chil-
dren in other countries are healthier and can eat an
egg a day because they have an entrained lower choles-
terol level because they had breast milk as babies.
Whereas North Americans have high cholesterol levels
because their synthesis was moved into high gear
because they had cholesterol-free infant formula.
References
1 Rabin RC. New cases of cancer decline in the U.S. The New York
Times. 26 November 2008. http://www.nytimes.com/2008/11/26/
health/research/26cancer.html (accessed March 2009).
2 Kavey R-EW, Daniels SR, Lauer RM et al. AHA Scientific State-
ment: American Heart Association Guidelines for primary preven-
tion of atherosclerotic cardiovascular disease beginning in
childhood. Circulation 2003; 107: 1562–6.
3 World Health Organization. The Top 10 Causes of Death. http://
www.who.int/mediacentre/factsheets/fs310/en/print.html (accessed
March 2009).
4 Volk MG. An examination of the evidence supporting the associa-
tion of dietary cholesterol and saturated fats with serum choles-
terol and development of coronary heart disease. Altern Med Rev
2007; 12: 228–45.
5 Keys A. Atherosclerosis: a problem in newer public health. J Mt
Sinai Hosp NY 1953; 20: 118–39.
6 Keys A, Menotti A, Aravanis C et al. The Seven Countries Study:
2,290 deaths in 15 years. Prev Med 1984; 13: 141–54.
7 Haq IU, Ramsay LE, Yeo WW et al. Is the Framingham risk func-
tion valid for northern European populations? A comparison of
methods for estimating absolute coronary risk in high risk men
Heart 1999; 81: 40–6.
Journal compilation ª 2009 Blackwell Publishing Ltd Int J Clin Pract, October 2009, 63 (Suppl. 163), 1–8
8 The National Cholesterol Education Program (NCEP) Expert Panel
on Detection, Evaluation, and Treatment of High Blood Choles-
terol in Adults (Adult Treatment Panel III). Third Report of the
National Cholesterol Education Program (NCEP) Expert Panel on
Detection, Evaluation, and Treatment of High Blood Cholesterol
in Adults (Adult Treatment Panel III) final report. Circulation
2002; 106: 3143–421.
9 Hu FB, Stampfer MJ, Rimm EB et al. A prospective study of egg
consumption and risk of cardiovascular disease in men and
women. JAMA 1999; 281: 1387–94.
10 Fraser GE. Diet and coronary heart disease: beyond dietary fats
and low-density-lipoprotein cholesterol. Am J Clin Nutr 1994; 59
(Suppl.): 1117S–23S.
11 Herron KL, Lofgren KE, Sharman M et al. High intake of choles-
terol results in less atherogenic low-density lipoprotein particles in
men and women independent of response classification. Metabo-
lism 2004; 53: 823–30.
12 Jones PJH, Pappu AS, Hatcher L et al. Dietary cholesterol feeding
suppresses human cholesterol synthesis measured by deuterium
incorporation and urinary mevalonic acid levels. Arterioscler
Thromb Vasc Biol 1996; 16: 1222–8.
13 Sachdeva A, Cannon CP, Deedwania PC et al. Lipid levels in
patients hospitalized with coronary artery disease: an analysis of
136,905 hospitalizations in Get With The Guidelines. Am Heart J
2009; 157: 111–7. e2, Epub 22 October 2008.
14 Yusuf S, Hawken S, Ounpuu S et al. Effect of potentially modifi-
able risk factors associated with myocardial infarction in 52 coun-
tries (the INTERHEART study): case–control study. Lancet 2004;
364: 937–52.
15 Qureshi AI, Suri MFK, Ahmed S et al. Regular egg consumption
does not increase the risk of stroke and cardiovascular diseases.
Med Sci Monit 2007; 13: CRI-8.
16 Mutungi G, Ratliff J, Puglisi M et al. Dietary cholesterol from eggs
increases plasma HDL cholesterol in overweight men consuming a
carbohydrate-restricted diet. J Nutr 2008; 138: 272–6.
17 Jenkins DJA, Kendall CWC, Marchic A et al. Effects of a dietary
portfolio of cholesterol-lowering foods vs lovastatin on serum lip-
ids and C-reactive protein. JAMA 2003; 290: 502–10.
18 Kassis A, Vanstone C, AbuMweis S, Jones P. Efficacy of plant ster-
ols is not influenced by dietary cholesterol intake in hypercholes-
terolemic individuals. Metabolism 2007; 57: 339–46.
Paper received 11 May 2009, accepted 3 June 2009
ª 2009 The Author