Newsletters & Publications
15 - Number 1
BMI: body mass index (kg/m2)
CHD: coronary heart disease
CI: confidence interval
CVD: cardiovascular disease
HDL: high density lipoprotein
LDL: low density lipoprotein
Lp(a): lipoprotein (a)
MI: myocardial infarction
MUFA: monounsaturated fatty acids
NCEP: National Cholesterol Education Program
P:S: dietary polyunsaturated:saturated fat ratio
PUFA: polyunsaturated fatty acids
RR: relative risk
SFA: saturated fatty acids
VLDL: very low density lipoprotein
A criticism of previous studies of the effects
of adding eggs to the diet on plasma cholesterol levels has been
that the studies were carried out in young, normocholesterolemic
subjects. While this has not been a justifiable criticism, since
many studies were carried out in middle-aged and older men and
women, there has not been a study specifically investigating the
effects of added eggs on the plasma cholesterol levels of middle-aged
men and women with different types of hyperlipidemic.
In the past, egg feeding studies conducted in
healthy populations with normal plasma lipid levels showed that
dietary cholesterol had a minimal effect on blood cholesterol levels
in these individuals. Now, a study by Knopp et al. has investigated
the effects of egg feeding in 161 free-living adults with either
hypercholesterolemia (HC, n=79) or combined hyperlipidemia (CHL,
n=52 ). All subjects had LDL cholesterol levels between 130-190
mg/dl, but ones with TAG greater than the 75th percentile for age
and sex were classified as CHL. The subjects selected for this
study were free of diabetes, uncontrolled hypothyroidism or heart
disease. All women in this study were post-menopausal and averaged
60 years of age and the average age of men was 52 years.
Prior to initiating the experimental phase of
this double-blind study, all subjects were instructed on NCEP Step1
Diet by a registered dietitian. They followed this diet for 6 weeks.
Then 60% of subjects were randomly assigned to the egg diet and
40% were assigned to an egg-substitute placebo diet. The total
duration of the experimental phase was 12 weeks. Each egg supplement
contained either 100 grams of homogenized whole egg, the equivalent
of 2 eggs, or 100 grams of a cholesterol-free egg substitute. Subjects
were instructed to keep a food record to assess dietary compliance.
On week 4, 8, 11, and 12, subjects returned to the clinic to receive
egg products and check weights and obtain fasting blood measurements.
Following the study, researchers found that 131
subjects complied with study protocol of maintaining their baseline
weight and eating egg supplements more than 80% of the time. At
week 12, subjects in the egg group consumed 29.6% of energy from
fat, an increase of 2.7% from baseline, while total fat intake
in the placebo group remained relatively stable at 25.8% of energy.
In addition to the increase in total fat, saturated fat and dietary
cholesterol increased in the egg-fed group, 8.2% to 9.1% of calories
and 187 mg/d to 601 mg/d, respectively, in contrast, saturated
fat and dietary cholesterol in the placebo group decreased slightly
from 7.8% to 7.3% of calories and 185 mg/d to 155 mg/d dietary
Researchers found the following changes in plasma
lipoprotein concentrations. The plasma LDL cholesterol increased
in both the HC and CHL egg-fed groups, 3 mg/dl and 12 mg/dl (P<0.001),
respectively, but no significant change was noted in either placebo
groups. Only the increase in LDL cholesterol in CHL subjects were
significant. Plasma HDL cholesterol levels also increased in the
egg-fed group but no changes occurred in the placebo groups. HC
subjects on the egg diet increased their HDL cholesterol by 4 mg/dl
and CHL group s HDL cholesterol increased by 3 mg/dl. The total
cholesterol level increased in both egg fed and placebo CHL subjects
but it did not change in HC groups. TAG levels did not change in
any of the 4 groups following the diet change. Apo A and apo B
concentrations did change in certain groups. Apo A in egg fed HC
subjects increased by 11 mg/dl and apo B increased by 4 mg/dl in
egg fed CHL and 8 mg/dl in placebo CHL. Plasma glucose and immunoreactive
insulin concentrations were not altered by the intervention diet.
Egg feeding did not have an effect on LDL subclass phenotype expression
or on the composition of postprandial lipoprotein. Knopp and colleagues
concluded that adding 2 eggs to a NCEP Step 1 Diet did not result
in significant increase in LDL cholesterol in HC subjects, however,
it did raise LDL cholesterol in CHL subjects.
- Adding 2 eggs per day did not increase LDL cholesterol in middle-aged
hypercholesterolemic men and women.
- Adding 2 eggs per day did increase LDL cholesterol in middle-aged
combined hyperlipidemic men and women.
- 2 eggs per day increased HDL cholesterol by 4 mg/dl in combined
hyperlipidemic subjects and 3 mg/dl in hypercholesterolemics.
The report by Knopp et al. represents an important step in defining
the metabolic factors involved in the heterogeneity of individual
responses to dietary cholesterol. Previous studies have documented
the existence of hyper-responders and hypo-responders to dietary
cholesterol but without characterization of individual subjects.
Studies by Ginsberg and colleagues showed that young, normo-cholesterolemic
males and females were mostly resistant to the plasma cholesterol
raising effects of dietary cholesterol. Knopp et al. have now
extended their observations to include middle-aged, mildly
hypercholesterolemic men and women as resistant to effects
of dietary cholesterol in eggs on plasma cholesterol levels.
In contrast, these data clearly show that individuals with
combined hyperlipidemia are dietary cholesterol sensitive.
The results also provide an estimate of the effects of dietary
cholesterol in a population of hyper-responders. Based on the
results of a comprehensive meta-analysis of fat and cholesterol
feeding studies, Howell et al. found a mean plasma cholesterol
response of 2.2 mg/dl per 100 mg/d dietary cholesterol. From
the results of the Knopp et al. study we can estimate that
hypo-respondents have an average plasma cholesterol response
of 1.5 mg/dl per 100 mg/d cholesterol whereas hyper-respondents
have a value of 2.9 mg/dl, or almost double the response in
hypo-respondents. Studies such as the one by Knopp et al. will
eventually provide the data necessary for targeted dietary
restrictions. Until that time, it is clear that restrictions
of dietary cholesterol and egg consumption are most appropriate
for subjects with elevated levels of plasma TAG and cholesterol.
For the rest of the population such restrictions would seem
to have very limited values
Knopp, R.H., Retzlaff, B.M., Walden, C.E., et
al. A double-blind, randomized, controlled trial of the effect
so two eggs per day in moderately hypercholesterolemic and combined
hyperlipidemic subjects taught the NCEP Step 1 Diet. J Am Coll
Nutr 1997; 16:551-561.
Table of Contents
In spite of the conflicting findings from studies
of the effects of very low-fat diet on plasma lipids, thereare
still many proponents for this restrictive diet. To answer questions
regarding the effects of a moderate fat-restriction diet versus
an aggressive fat-restriction diet in free-living hyper-lipidemics,
Knopp and colleagues conducted a randomized, parallel, comparison
trial in men from the Seattle-Puget Sound region. The study addressed
two questions: one, does greater fat restrictions result in greater
LDL reductions, and two, does the plasma lipoprotein response differ
between patients with various hyperlipidemias. Of the 444 men who
completed this year long study, all subjects had LDL cholesterol
levels greater than the 75th age-specific percentile (160 mg/dl).
Those with plasma levels of TAG at or above the 75th age-specific
percentile were classified as combined hyperlipidemic (CHL) and
those with TAG less than 75th age-specific percentile were defined
as hypercholesterolemic (HC).
Baseline clinical characteristics were measured
and HC subjects were randomly assigned to study diets 1, 2, 3,
and 4 and CHL were assigned to study diets 1, 2, and 3. Study diet
4 was eliminated from the CHL group due to the smaller number of
CHL subjects (174) compared to HC subjects (270). Study diet 1,
2, 3, and 4 provided 30%, 26%, 22%, and 18% calories from fat and
300 mg, 200 mg, 100 mg, and 100 mg/d of cholesterol, respectively.
The target protein intake of the 4 diets were between 16% and 18%
of energy. However, the 4 day food records indicated that the majority
of subjects did not achieve their target fat intake. For example,
subjects in the HC groups consumed 27%, 26%, 25%, and 22% of energy
from fat for the four test diets and subjects in the CHL groups
consumed 28%, 26%, and 25% fat. Target cholesterol intakes were
also not achieved. The subjects in diet groups 1 and 2 consumed
less than their goals of 300 mg and 200 mg/d, while subjects in
groups 3 and 4 consumed greater than 100 mg/d. The table below
presents the results of different fat restriction diet on plasma
lipids and lipoprotein levels in HC and CHL subjects. Study diets
2 and 4 resulted in lower total cholesterol, LDL, and apo B levels
in HC. More restrictive fat diets did not yield further benefits
on body weight, glucose, insulin, or blood pressure. On the contrary,
the very low-fat diet actually resulted in an unwanted lipoprotein
profile by increasing TAG and VLDL levels and decreasing HDL cholesterol
in HC subjects and decreasing apo B in HC and CHL groups.
The authors concluded that implementation of NCEP
Step 1 and Step 2 diets in HC and CHL individuals can be effective
in lowering plasma levels of atherogenic lipoproteins. However,
aggressive fat restrictions did not show evidence of further benefit.
In fact, the observed decrease in HDL levels in HC subjects on
the 25% and 22% fat diets suggest potentially adverse effects.
The results of this study suggest that more aggressive dietary
fat lowering recommendations are probably not appropriated for
free-living population at this time.
Percent changes in plasma lipids, lipoproteins
and apo B from baseline
Knopp, R. H., Walden, C.E., Retzlaff, B.M., et
al. Long-term cholesterol-lowering effects of 4 fat-restricted
diets in hypercholesterolemic and combined hyperlipidemic men.
The dietary alternative study. JAMA 1997;278:1509-1515.
Table of Contents
Using data from the Cancer Prevention Study I,
Stevens et al. investigated the effect of age on the association
between BMI and mortality. From the original study cohort, 62,116
men and 263,019 women were selected for this study. These subjects
were Caucasian in heritage, free of chronic disease, and nonsmokers.
The follow-up period lasted from 1960 to 1972. For the purpose
of this study, subjects were separated into the following BMI and
age categories: BMI of <19.0, 19.0-21.9, 22.0-24.9, 25.0-26.9,
27.0-28.9, 29.0-31.9, and >32.0, and 30-44 year, 45-54 year,
55-64 year, 65-74 year, 75-84 year, and >85 year old,
As expected, the incidence of death from all causes
and death from CVD were higher in the older population. The mean
BMI was relatively constant in males of all age groups except the
oldest group. The mean BMI in women increased from 23.8 to 25.2
from the 30-44 year old group to 55-64 year old group, then steadily
declined in the older age category. Results indicated a higher
incidences of CVD mortality among heavier people than thin people
within the same age category. For example, CVD mortality was 22
deaths per 100,000 person-year in 30-44 year old with BMI of 19.0-21.9
compared to 51 deaths per 100,000 person-year in the same age group
of women with BMI of 29.0-31.9. However, the incidence of CVD mortality
was much higher in older women. Women in the 65-74 year old group
with BMI of 19.0-21.9 had CVD mortality incidence of 1854 per 100,000
person-year and ones with BMI 29.0-31.9 was 1854 per 100,000 person-year.
This trend was not observed in people over 75 years old. CVD mortality
was relatively constant in this older group; 5259 per 100,000 person-year
and 5227 per 100,000 person-year in the 19.0 to 21.9 BMI and 29.0
to 31.9 BMI group, respectively. A higher BMI was associated with
a higher relative risk for all causes of death and CVD death in
younger subjects than older people.
The authors concluded that excess body weight
increases the risk of death from both CVD and all causes in adults
between the ages of 30 and 74 years. The data are consistent with
the hypothesis that the relative risk of death associated with
excess adiposity is lower for older than for younger adults. Overall,
the results support the healthy weight ranges in the 1995 Dietary
Guidelines for Americans which omitted the age-specific weight
recommendations of earlier version of the Guidelines.
Stevens, J., Cai, J., Pamuk, E.R., et al. The
effect of age on the association between body-mass index and mortality.
N Engl J Med 1998;338:1-7.
Table of Contents
A meta-analysis of dietary intervention trial
by Brunner and colleagues evaluated the effectiveness of dietary
recommendations on CVD risk factors. Based on study criteria, 17
published studies were selected for the meta-analysis dietary recommendations
on CVD risk factors. Based on study criteria, 17 published studies
were selected for the meta-analysis. All 6893 subjects in the studies
were free-living adults randomly assigned to either a dietary intervention
or control group for greater than 3 month. The outcomes Brunner
et al. compared were percent calories from fat, serum total cholesterol,
urinary sodium, and diastolic blood pressure levels divided into
either 3-6 months or 9-18 months duration.
Subjects in the intervention groups consumed an
average 6% less dietary fat than controls, except for women in
a breast cancer prevention study. Subjects in this study were highly
motivated and reduced their fat intake by 40%. The median serum
total cholesterol level in intervention groups decrease by 10.8
mg/dl at 3 to 6 months and 8.5 mg/dl at 9 to 18 months. However,
researchers noted a significant heterogeneity in total cholesterol
levels among different studies. An increase in frequency of interaction
between subjects and researchers resulted in a greater decline
in serum cholesterol level. The mean urinary sodium excretion was
32 mmol/24 hr less in the intervention group. This level equals
19 grams of sodium chloride or a 20% reduction in salt intake.
Lastly, except in 2 studies, dietary intervention also resulted
in a decline in diastolic blood pressure. At 3 to 6 months and
9 to 18 months, diastolic blood pressure of intervention groups
was 0.7 mm Hg and 1.2 mm Hg less, respectively, than control subjects.
In spite of the differences in the study design,
data collection, and possible bias due to missing follow-up data,
Brunner and associates were able to observe an improvement in subjects
CVD risk as a result of dietary intervention. Using data from a
primary prevention trial and cohort studies which estimated that
a 10% reduction (23 mg/dl) in serum cholesterol reduces CHD by
25%, and a 5 mm Hg reduction in diastolic blood pressure results
in a 21% reduction in CHD and a 34% reduction in stroke, researchers
concluded that dietary intervention could reduce incidence of CHD
and stroke by 14% and 8%, respectively, decline in CHD risk due
to a decrease in serum cholesterol was 9%. Brunner et al. also
noticed following trends; subjects were more likely to change and
adhere to a low fat diet if they thought it would reduce future
disease, programs with greater intervention were associated with
greater net effect, and dietary compliance was low beyond 18 months.
Researchers concluded that individual dietary interventions can
achieve moderate reduction in CVD risk.
Brunner, E., White, I., Thorogood, M., et al.
Can dietary interventions change diet and cardiovascular risk factors?
A meta-analysis of randomized controlled trials. Am J Public
Table of Contents
There are limited data on the effects of a Step
1 Diet on plasma lipids and lipoproteins of obese, postmenopausal
women. It is also unclear what effects the Step 1 Diet plus weight
loss will have on plasma lipoprotein profiles in this population.
In this study, Nicklas and colleagues investigated the changes
in lipoprotein lipids of obese postmenopausal women as a result
of consuming a Step1 Diet alone and then with weight loss. The
baseline mean age and BMI for the 48 study subjects were 61 years
and 32.7 kg/m2, respectively. Following baseline measurements
of height, weight, dietary intake, and plasma lipoprotein lipid
concentrations, all subjects underwent a 2 month eucaloric Step
1 Diet followed by 6 months of dietary intervention with weight
loss. During both dietary interventions, volunteers were enrolled
in a weekly nutrition class taught by a registered dietitian. The
attendance rate was 90% during the Step1 Diet phase and 79% during
the diet-weight loss phase. The weight loss diet was designed to
provide a gradual weight loss of 0.25-0.5 kg/week by decreasing
energy by 1.0-1.5 mJ/d. Physical activity level was constant throughout
the study period. Lipoprotein lipid concentrations were measured
following each intervention period.
Baseline data showed that the diet consisted of
33% calories from fat, 10% saturated fat, 6% polyunsaturated fat,
215 mg/d dietary cholesterol, and 2842 mg Na per day. Following
initiation of the Step 1 Diet, subjects consumed less total fat
calories (23%), saturated fat (6.7%), cholesterol (170 mg/d), and
sodium (2,456 mg/d) and more calories from carbohydrate (59%).
While on this diet, subjects lost a small, but significant amount
of weight (2.0+0.3 kg) and during the weight loss phase
of the study, the average weight loss was 5.6 +0.7 kg. Also,
following the dietary intervention, all lipoprotein lipid concentrations,
except HDL cholesterol, improved. For example, after the Step 1
Diet, total cholesterol, LDL cholesterol, and HDL cholesterol levels
decreased by 7%, 6%, and 14%, respectively. Volunteers with undesirable
TAG levels decreased from 10% to 2%, LDL cholesterol from 54% to
44%, and total cholesterol from 67% to 50%. However, the number
of subjects with undesirable HDL (<35 mg/dl) increased from
8% to 13% of the study population. Following the weight loss phase,
HDL cholesterol and HDL2 cholesterol increased by 8%
and 11%, respectively, thus reducing the number of women with undesirable
HDL cholesterol levels from 40% to 21%. While weight loss resulted
in a 9% decrease in plasma TAG levels, but total and LDL cholesterol
concentrations did not change with weight loss. Lastly, the data
indicated that hypercholesterolemic subjects responded better to
the Step 1 Diet and the Step1 with weight loss than normocholesterolemic
and mildly hypercholesterolemic individuals. For example, following
both diet interventions, TAG, total cholesterol, and LDL cholesterol
concentrations decreased by 19%, 13%, and 14%, respectively, in
hypercholesterolemic women but there were no significant plasma
lipid changes in the other 2 groups.
In conclusion, even though the Step 1 Diet plus
weight loss did improve the plasma lipoprotein lipid profile, and
in theory the risk for CVD, in obese, postmenopausal women with
hypercholesterolemia, it had minimal benefit for obese, postmenopausal
women with normal or mildly elevated plasma cholesterol. In that
the Step 1 Diet decreased HDL levels when no weight loss was achieved
raises serious question about the efficacy of a low-fat diet in
improving the overall plasma lipoprotein profile.
Nicklas, B.J., Katzel, L.I., Bunyard, L.B., et
al. Effects of an American Heart Association diet and weight loss
on lipoprotein lipids in obese, postmenopausal women. Am J
Clin Nutr 1997;66:853-859.
Table of Contents
A recently published study reported on differences
in HDL cholesterol concentrations in Japanese, American, and Australian
children. The data show that concentrations of HDL cholesterol
were higher in Japanese than U.S. or Australian children. More
importantly, the plasma total cholesterol:HDL ratio remained stable
between ages 8 to 10 years and 12 to 15 years in Japanese boys
and girls whereas it increased in U.S. and Australian boys. Comparisons
of Japanese and Australian children showed that BMI values were
similar but the Japanese children were considerably more physically
active. Japanese children consumed less fat (27% ene) than the
Australian children (37% ene) yet they consumed more total calories.
Japanese children consumed more soy bean products, fish, and eggs
than Australian children. For tofu and related products, there
was virtually no consumption by Australian children while 61% of
Japanese children included this food items in their 24-hour dietary
recall. Sixty-eight percent of Japanese children consumed eggs
compared to only 18% of Australian children. [Editor's Note:
Japan has the highest per capita egg consumption in the world.]
These observed differences in HDL levels and in the age-related
changes in plasma total:HDL cholesterol ratio may help to explain
why CHD mortality rates in Japan are low compared to other developed
Dwyer, T., Iwane, H., Dean, K., et al. Differences
in HDL Cholesterol Concentrations in Japanese, American, and Australian
Children. Circulation 1997;96:2830-2836.
Table of Contents
The multicenter, randomized, controlled Dietary
Intervention Study in Children (DISC) investigated the safety and
efficacy of dietary fat and cholesterol reductions on plasma LDL
cholesterol in 663 pubescent children stigated the safety and efficacy
of dietary fat and cholesterol reductions on plasma LDL cholesterol
in 663 pubescent children wi th elevated baseline LDL cholesterol.
Investigators also examined the association between baseline BMI
and sexual maturation on LDL cholesterol concentrations.
Children were divided into intervention and usual
care study groups. Subjects in the intervention group were provided
with individual eating plans as well as continuous nutritional
education classes throughout the follow-up period. Parents of the
usual care children were told that their children had elevated
blood cholesterol and provided with general publications on heart-health
at the onset of the study. Following the 3 year intervention period,
children in the intervention group consumed less total fat (28.6%
vs 33%), saturated fat (10.2% vs 12.3%), and cholesterol (95.0
vs 112.9 mg/ 4.2 MJ) than the usual care group. During the study
period, subjects weight, height, and lipoprotein concentrations
were measured at baseline, and first and third year of follow-up.
Subjects sexual maturation were assessed annually using Tanner
Sexual maturation and BMI were two biological
factors which had strong influences on LDL levels. A 1 unit increase
in BMI was associated with a 0.6 mg/dl increase in LDL cholesterol
in boys and 1.1 mg/dl in girls. Sexual maturation was inversely
associated with plasma LDL cholesterol concentrations. The LDL
cholesterol levels in boys in Tanner stage 4+ were 23.3 mg/dl lower
than boys in stage 1 and girls in Tanner stage 4 were 10.6 mg/dl
lower than girls in Tanner stage 1.
Dietary factors also had significant effects on
LDL cholesterol concentrations. The children in the intervention
group had lower LDL cholesterol than children in the usual group.
The LDL cholesterol level was 5.0 2.1 mg/dl lower in intervention
girls and 2.7 2.4 mg/dl in intervention boys. The difference in
LDL cholesterol was significant in girls only. However, dietary
cholesterol was directly associated with LDL cholesterol level
in boys. For each 10 mg/1000 kcal decrease in dietary cholesterol,
plasma LDL cholesterol decreased by 0.7 mg/dl. [2.8 mg/dl per 100
In boys, higher dietary cholesterol and BMI were
associated with higher plasma LDL cholesterol. However, in girls,
only high BMI was associated with higher LDL cholesterol. There
was no relationship between dietary fat intake and LDL cholesterol
in girls. But "Tanner stage 4+ was significantly related to
lower LDL cholesterol in both boys and girls."
Kwiterovich, P.O., Barton, B.A., McMahon, R.P.,
et al. Effects of diet and sexual maturation on low-density lipoprotein
cholesterol during puberty. The dietary intervention study in children
(DISC) Circulation 1997;96:2526-2533.
Table of Contents
In our weight conscious society, many people,
both scientists and lay persons, have arguedabout the role genetics
plays in body weight. Using medical records of 854 subjects, Whitaker
et al. analyzed the effects of childhood and parental obesity on
a child s risk of becoming obese in their 20s. All the cohorts
in this retrospective study were members of a Group Health Cooperative
and their longterm weight and height records were available for
analysis. Parents weight and height were also available for the
majority of subjects. Researchers classified children with BMI
at or above the 85th percentile for age and sex as obese and BMI
at or above 95th of age and sex as very obese. In adults, men with
BMI of >27.8 were considered obese as were women with
The data indicated a direct relationship between
the age of initial obesity and adult obesity. For example, the
odds ratio for obese child developing adult obesity was 1.3 between
ages 1-2 years, 4.1 between 3-5 years, 10.3 between 6-9 years,
28.3 between 10-14 years, and 20.3 between 15-17 years. Also, very
obese children at all ages were always more likely to become an
obese young adult. At every age, children with obese parents had
a higher risk of adult obesity, and this was especially significant
in children under 10 years of age. Parents obesity status was a
primary predictor of adult obesity in children younger than 3 years
old, while a child s obesity status was more important in predicting
adult obesity in children over 10 years old. Lastly, children with
two obese parents had a higher risk of obesity than children with
only one obese parent.
Childhood and parental obesity were good predictors
of adult obesity. Knowing this, researchers concluded that infant
and toddlers with obese parent and obese children should benefit
from treatments to prevent excessive weight gain.
Whitaker, R.C., Wright, J.A., Pepe, M.S., et al.
Predicting obesity in young adulthood from childhood and parental
obesity. N Engl J Med 1997;337:869-73.
Table of Contents
Recently, many studies suggest that moderate alcohol
intake has a protective effect against CHD. However, with all the
known side effects of heavy alcohol intake, one must be cautious
when promoting alcohol s beneficial effects in reducing CHD risk.
This is especially true since data on the upper limit of alcohol
s benefit for CHD risk reduction is unclear. Using data from the
National Health and Nutrition Examination Survey I (NHANES I) Epidemiologic
Follow-up Study (NHEFS), Rehm et al. analyzed the association between
alcohol consumption and CHD incidence and mortality rates. Researchers
selected 6,788 subjects (3,828 females, 2,960 males) for the follow-up
study. All subjects were European-Americans, 40-75 years old, and
free of CHD. With the aid of a 4-question questionnaire, a subject
s baseline alcohol consumption between 1971 to 1975 was determined.
Follow-up was conducted from 1982 to 1984, in 1986, and again in
Subjects in the 2-7 drinks/week category had the
lowest risk of CHD. For women in this category, the RR for CHD
incidence and mortality was 0.51 and 0.55, respectively, and in
men it was 0.62 and 0.73, respectively. However, as alcohol consumption
increased, RR for CHD also increased. The women with the highest
alcohol consumption, greater than 29 drinks per week, had a RR
of 2.6 and 4.6 for CHD incidence and mortality, respectively. Surprisingly,
high alcohol consumption did not affect CHD risk in males to the
degree it affected women. In men, RR for CHD incidence was 0.62
for greater than 42 drinks per week and RR for CHD mortality was
0.89 for greater than 29 drinks per week. There were other sex-related
differences noted in this analysis. The female light drinkers,
less than 2 drinks per week, had less benefit from alcohol consumption
than men in the same drinking category. Also, the RR for CHD incidence
was higher for female current abstainers (1.21) than males (0.87).
The data suggest that in women, the benefit of alcohol consumption
is associated with a specific level of intake. Too much or too
little alcohol was actually related to increased CHD risk in women.
This was not the case in males. Females exhibited a U-shaped relationship
between alcohol consumption and CHD whereas males had a L-shaped
In conclusion, this study showed that alcohol
intake resulted in decreased CHD risk compared to abstainers. Also,
there is a definite gender difference in the relationship between
alcohol use and CHD risk reduction. Females obtained maximum benefit
on less than 1 drink per week and the least benefit with levels
greater than 4 drinks per day. In contrast, heavy drinking, greater
than 4 drinks per day, did not increase CHD risk in men.
Rehm, J.T., Bondy, S.J., Sempos, C.T., et al.
Alcohol consumption and coronary heart disease morbidity and mortality. Am
J Epidemiol 1997;146:495-501.
Table of Contents
Once upon a time the credibility of a research
publication was based on four criteria: reputations of the investigators;
quality of the journal; a reader s evaluation of the methods, data
analysis and interpretation; and ability of other investigators
to reproduce the results. In a more cynical era it seems the determinant
of the validity of a study is who funded it. In nutritional sciences,
industry-funded studies [especially from those members of the "bad-food
group"] are automatically suspect, despite satisfying the
classic criteria for scientific veracity. Critics have no problem
asking credible scientists, published in the most prestigious journals, "That's
all very well and good, but who funded it?" If a group is
unwilling, or unable, to attack the science, then the diversionary
tactic is to attack the funding and, by inference, the credibility
of the researchers.
As a research scientist who received government,
foundation, association and industry funding over many years, and
now administers an industry research grant-in-aid program, I am
bothered by the "who funded it" assessment process. Is
it really so easily imagined that scientists with international
reputations, various sources of funding, and involvements with
government agencies and nonprofit organizations, would jeopardize
their careers for the meager funds available from the food industry,
and falsify research published in the best scientific journals?
I would hope that we would give our colleagues credit for professionalism
and integrity, while also recognizing the expertises of journal
reviewers, editors and readers.
A recent newspaper article stated that if industry
uses research it funds for promotion, they have blurred "the
line that separates science from marketing" (The Sunday
Record, 4 January 1998, Hackensack NJ). What line? With TV
ads encouraging an overdose of cereal to lower cholesterol levels,
to taking a statin drug to lower the risk of a first heart attack,
it is clear that many blurs this unlikely line. The same article
stated that "Critics, though, accuse food producers of seeking
out scientists whose views closely match their own, and manipulating
the scientific agenda by influencing what research gets funded." It
seems improbable that the American Egg Board, with its level of
research funding, could "manipulate the scientific agenda." And
the argument that industry only funds "scientists whose views
closely match their own" is equally untenable when industry-funded
scientists continue to advocate the nutritional conventional wisdom
even when their studies suggest otherwise.
Two examples of this criticism are worth evaluating.
The American Egg Board/Egg Nutrition Center partially supported
a study by Howell et al. (Nutrition
Close-Up 14(2), 1997) which has received industry-funded
criticism even though the results were reproduced in a study by
Clarke et al. ((Nutrition
Close-Up 14(1), 1997) which was not funded by industry.
[Interestingly, scientific advisors of some advocacy groups themselves
receive industry funding for research. It must be that they get
it from 'good-food groups'!] Industry-funded studies also run the
risk of unintended consequences as shown by the study of Knopp
et al. reported in this issue of Nutrition Close-Up. Knopp
et al. found that, while adding eggs to the diets of mildly hypercholesterolemic
individuals had little effect on blood cholesterol levels, those
with combined hyperlipidemia were sensitive to the plasma cholesterol
raising effects of dietary cholesterol. The fact is, for any industry-funded
research it is a case of pay your money, take your chances.
The stigma of "industry-funded research" is
strong enough that we often are unable to get investigators to
research specific questions. We have tried to have the relationship
between egg consumption and heart disease incidence analyzed using
data from epidemiological trials. Clearly the population studies
are there, the data collected, and the CHD incidence high enough
for statistical analysis. We know because we read results from
these studies almost monthly. Nevertheless, we get turned down
when offering to fund a grant, a fellowship, or any other funding
venue. Why? Because the investigators fear that if the data do
not show a positive relationship then the findings will be interpreted
as industry-funded research. For many investigators the level of
funds available are not worth the accusations and credibility challenges.
So let's look at the position of the egg industry.
Dietary cholesterol is accused of being a contributor to high blood
cholesterol. The USDA Dietary Guidelines established that dietary
cholesterol intake should be no more than 300 mg per day. And the
American Heart Association advises that individuals limit their
egg intake to "no more than 3 to 4 per week." There are
scientists in this country, and those on the nutrition advisory
committees of most countries of the world, who believe that eggs
are not a contributor to CHD risk. How do we test the hypothesis
if we cannot fund studies of the question? And when we do have
applications from highly qualified investigators with protocols
reviewed by a Scientific Advisory Panel with impeccable credentials,
then the study gets published in a prestigious peer-reviewed journal,
the only response from the nay sayers of nutritional politics is "invalid,
flawed, unsound industry-funded study." I guess in today s
approach to science it's just a case of "damned if you do,
out of business if you don't."
Donald J. McNamara, Ph.D.
Executive Editor, Nutrition Close-Up
Table of Contents
Editor: Donald J. McNamara, Ph.D.
Writer/Editor: Linda Min, M.S., R.D.
Nutrition Close-Up is published quarterly
by the Egg Nutrition Center. Nutrition Close-Up presents
up-to-date reviews, summaries and commentaries on the latest research
investigating the role of nutrition in health promotion and disease
prevention, and the contributions of eggs to a nutritious and healthful
diet. Nutrition and health care professionals can receive a FREE
subscription for the newsletter by contacting the ENC.