Authors: T. Colin Campbell,Thomas M. Campbell
275
SCIENTIFIC REDUCTIONISM
CHART 14.3: FAT INTAKE AND BREAST CANCER MORTALITY
• NETHERLANDS
25 FEMALE .UK • DENMARK
IN..
CANADA. • NEW ZEALAND
ci .SWITZERLAND
0
RELA D
a. .US
g 20 BELGIUM
AUSTRALIA •• SWEDEN
0
o· AUSTRIA. • • GERMANY
0
... NORWAY. FRANCE
• ITALY
15 .CZECH
.<:l • • FINLAND
ttl
cr::: • PORTUGAL HUNGARY
.s HONG KONG • POLAND
ttl
c 10
CIJ
CHILE •
VENEZ UELA • • • BULGARIA
• • .SPAIN
"C ROMANIA.
~ PANAMA • YUGOSLAVIA GREECE
::l
PHILIPPINES COLOMBIA. PUERTO RICO
5
'0
'1 • • • MEXICO
JAPAN. TAIWAN
• CEYLON
Ol
« THAILA~ .EL SALVADOR
0
20 40 60 80 100 140 160 180
0 120
Total Dietary Fat Intake (g/day)
that the closer a population gets to consuming a plant-based diet, the
lower its risk of breast cancer.
But because the women in the Nurses' Health Study are so far from a
plant-based diet, there is no way to study the diet and breast cancer rela-
tionship originally suggested by the international studies. There are virtu-
ally no nurses that eat a diet typical of the countries at the bottom of this
graph. Make no mistake about it: virtually this entire cohort of nurses
is consuming a high-risk diet. Most people who look at the Nurses'
Health Study miss this flaw because, as Harvard researchers will point
out, there is a wide range of fat intake among the nurses.
The group of nurses who consume the least fat eat 20-25% of their
calories as fat, and the group of nurses who consume the most fat eat
50-55% of their calories as fat. 10 At a casual glance, this range appears to
indicate substantial differences in their diets, but this is just not true, as
almost all women uniformly eat a diet very rich in animal-based foods.
That begs the question, how can their fat intake vary dramatically while
they all uniformly consume large amounts of animal-based foods?
Ever since "low-fat" became synonymous with "healthy," technology
has created many of the same foods that you know and love, without
the fat. You can now have all kinds of low-fat or no-fat dairy products,
low-fat processed meats, low-fat dressings and sauces, low-fat crackers,
low-fat candies and low-fat "junk food," like chips and cookies. In other
276 THE CHINA STUDY
words, you can eat mostly the same foods as you did twenty-five years
ago , while substantially reducing your fat intake. But you still retain the
same proportion of animal- a n d plant-based food intakes.
In practical terms this means that beef, pork, lamb and veal consump-
tion is decreasing while lower-fat chicken, turkey and fish consumption
is increasing. In fact, by consuming more poultry and fish, people have
been increasing their total meat intake to record-high amounts, II while
trying (and largely failing 12 ) to reduce their fat intake. In addition,
whole milk is being consumed less, but low-fat and skim milk are be-
ing consumed more. Cheese consumption has increased by 150% in the
past thirty years. 13
Overall, we are as carnivorous now as we were thirty years ago, but
we are able to selectively lower our fat intake if we so desire, due to the
wonders of food technology.
To illustrate, we need only to look at two typical American meals. 14, 15
Meal #1 is served in a health-conscious horne, where the main grocery
shopper in the family reads the nutrition labels on every food item he or
she buys. The result: a low-fat dining experience.
Meal #2 is served in a home where the standard American fare is ev-
eryone 's favorite. When they cook at home, they make the meal "rich. "
The result: a high-fat dining experience.
CHART 14.4: LOW-FAT AND HIGH-FAT AMERICAN DINNERS
(ONE PERSON'S DINNER)
Low-Fat Meal # 1 High-Fat Meal #2
Dinner 8 oz. roasted turkey 4 .5 oz. pan-seared steak
low-fat gravy Green beans almondine
Golden roasted potatoes Herb-seasoned potato
pockets
Beverage 1 cup skim milk Water
Nonfat yogurt
Dessert Apple crisp
Reduced-fat cheesecake
Both meals provide roughly 1,000 calories, but are markedly different
in their fat content. The low-fat meal (#1) contains about twenty-five
grams of fat, and the high-fat meal (#2) contains just over sixty grams
of fat. In the low-fat meal, 22% of the total calories come from fat, and
in the high-fat meal, 54% of the calories come from fat.
SCIENTIFIC REDUCTIONISM 277
The health-conscious home has managed to create a meal that is
much lower in fat than the average American dinner, but they've done
it without adjusting their proportionate intakes of animal- a n d plant-
based foods. Both meals are centered on animal-based foods. In fact,
the low-fat meal actually has more animal-based foods than the high-
fat meal. In effect, this is how the nurses in the Nurses' Health Study
achieved such a wide variation in fat intake. Some nurses simply are
more diligent about choosing low-fat animal products.
Many people would consider the low-fat meal to be a triumph of
healthy meal planning, but what about the other nutrients in these
meals? What about protein and cholesterol? As it turns out, the low-fat
meal contains more than double the protein of the high-fat meal, and almost
all of it comes from animal-based foods. In addition, the low-fat meal con-
tains almost twice as much cholesterol (Chart 14.5) .14,15
CHART 14.5: NUTRIENT CONTENTS OF TWO SAMPLE MEALS
Low-Fat Meal # 1 High-Fat Meal #2
22% 54%
Fat (percent of total calories)
36% 16%
Protein
(percent of total calories)
93% 86%
Percentage of total protein
derived from animal-based
foods
307 165
Cholesterol
An overwhelming amount of scientific information suggests that
diets high in animal-based protein can have unfavorable health con-
sequences, as can diets high in cholesterol. In the low-fat meal, the
amount of both of these unhealthy nutrients is significantly higher.
FAT VERSUS ANIMAL FOOD
When women in America, such as those in the Nurses' Health Study
and the billion-dollar4 Women's Health Trial,I6-19 reduce their fat intake,
they do not do it by reducing their consumption of animal-based foods.
Instead, they use low-fat and nonfat animal products, along with less fat
during cooking and at the table. Thus, they are not adopting the diets
that were shown, in the international correlation studies and in our ru-
ral China study, to be associated with low breast cancer rates.
278 THE CHINA STUDY
This is a very important discrepancy, and is illustrated by the correla-
tion between the consumption of dietary animal protein and dietary fat
for a group of countries (Chart 14.6) .8,9, 18,20-22 The most reliable com-
p a r i s o n was published in 1975 2°; it showed a highly convincing correla-
t i o n of more than 90%. This means that as fat intake goes up in various
countries, animal protein intake increases in an almost perfectly parallel
manner. Likewise, in the China Study, the intakes of fat and animal pro-
tein also show a similar correlation of 84%.8,21
In the Nurses' Health Study, this is not the case. The correlation
between animal protein and total fat intakes is only about 16%.9 In
the Women's Health Trial, also including American women, it is even
worse, at _17%18,21,22; as fat goes down, animal protein goes up. This
practice is typical of American women who have been led to believe
that, by decreasing their fat intake, they are changing to a healthier diet.
A nurse consuming a "low-fat" diet in the Harvard study, like American
women everywhere, is likely to continue eating large amounts of animal
protein, as shown in meal #1 (Chart 14.4).
Sadly, this evidence on the effects of animal-based food on cancer and
other diseases of affluence has been ignored, even maligned, as we con-
t i n u e to focus on fat and other nutrients in isolation. Because of this, the
Nurses' Health Study and virtually every other human epidemiological
study published to date have been seriously shortchanged in their in-
CHART 14.6: PERCENT CORRELATIONS OF TOTAL FAT
AND ANIMAL PROTEIN CONSUMPTION
100
80
China
60
40
Nurses' Health Study
20 "
•
1
-
o
I
International
-20
Women's Health Study
-40
279
SCIENTIFIC REDUCTIONISM
vestigations of diet and disease associations. Virtually all the subjects
under study consume the very diet that causes diseases of affluence. If
one kind of animal-based food is substituted for another, then the ad-
verse health effects of both foods , when compared to plant-based food,
are easily missed. To make matters worse, these studies often focus on
the consumption of just one nutrient, such as fat. Because of these very
serious flaws, these studies have been a virtual disaster for discovering
the really significant effects of diet on these diseases.
THE $1 OO( +) MILLION RESULTS
SO now that you know how I interpret the Nurses' Health Study and its
flaws, we should take a look at its conclusions. After more than $100
million and decades of work, there is no shortage of results. So what are
they? The logical place to start is, of course, the question of whether fat
intake really is linked to breast cancer. Here are some of the findings ,
cited verbatim:
• "these data provide evidence against both an adverse influence of
fat intake and a protective effect of fiber consumption by middle-
a g e d women on breast cancer incidence over eight years"23
Translation: The Nurses' Health Study did not detect a relationship
between dietary fat and fiber and breast cancer risk.
• "we found no evidence that lower intake of total fat or specific major
types of fat was associated with decreased risk of breast cancer"l0
Translation: The Nurses' Health Study did not detect a relationship
between reducingfat, whether it be total fat or certain kinds offat, and
breast cancer risk.
• "existing data, however, provide little support for the hypothesis
that reduction in dietary fat composition, even to 20% of energy
during adulthood, will lead to a substantial reduction in breast
cancer in Western cultures"24
Translation: The Nurses' Health Study did not detect a breast cancer
association with fat even when women reduced their fat consumption
all the way down to 20% of calories.
r
THE CHINA STUDY
280
• "relative risks for ... monounsaturated and polyunsaturated fat ...
were close to unity"25
Translation: The Nurses' Health Study did not detect a relationship
between these "good" fats and breast cancer risk.
• "we found no significant associations between intake of meat and
dairy products and risk of breast cancer"26
Translation: The Nurses' Heath Study did not detect a relationship
between meat and dairy consumption and breast cancer risk.
• "our findings do not support a link between physical activity, in
late adolescence or in the recent past, and breast cancer risk among
young adult women"27
Translation: The Nurses' Health Study did not detect a relationship
between exercise and breast cancer risk.
• "these data are suggestive of only a weak positive association with
substitution of saturated fat for carbohydrate consumption; none
of the other types of fat examined was significantly associated with
breast cancer risk relative to an equivalent reduction in carbohy-
d r a t e consumption"28
Translation: The Nurses' Health Study detected little or no effect on
breast cancer when women substituted fat for carbohydrates.
• "selenium intake later in life is not likely to be an important factor
in the etiology of breast cancer"29
Translation: The Nurses' Health Study did not detect a protective effect
of selenium on breast cancer risk.
• "these results suggest that fruit and vegetable consumption during
adulthood is not significantly associated with reduced breast can-
cer risk"30
Translation: The Nurses' Health Study did not detect a relationship
between fruits and vegetables and breast cancer risk.
So there it is, readers. Breast cancer risk does not rise with increased
intakes of fat, meat, dairy or saturated fat. Breast cancer is not prevented
SCIENTIFIC REDUCTIONISM 281
by increased intakes of fruits and vegetables, or reduced by exercise
(either during the teenage years or during adulthood), dietary fiber,
monounsaturated fats or polyunsaturated fats. Also, the mineral sele-
n i u m , long considered to be protective of certain cancers, has no effect
on breast cancer. In other words, we might as well conclude that diet is
completely unrelated to breast cancer.
I can understand the frustration of Professor Meir Stampfer, one of
the leading researchers in this group, when he was quoted as saying,
"This has been our greatest failure and disappointment-that we have
not learned more about what people can do to lower their risk."6 He
was making his comment in response to an opinion that "the Single big-
gest challenge for the future [is] sorting out the mess of contradictory
findings and lack of information on breast cancer."6 I applaud Professor
Stampfer for his candor, but it's unfortunate that so much money has
been spent to learn so little. Perhaps the most rewarding finding, ironi-
cally, was the demonstration that tinkering with one nutrient at a time,
while maintaining the same overall dietary patterns, does not lead to
better health or to better health information.
Yet Harvard researchers have been steadily cranking out their find-
ings, despite these challenges. From their slew of studies, here are some
findings that I would consider as very troubling contradictions when
comparing disease risks for men versus women:
• Men who consume alcohol three or four times a week have a lower
heart attack risk. 31
• Men with Type 2 diabetes who consume a moderate amount of
alcohol have a lower risk of coronary heart disease.32
And yet . . .
• Alcohol consumption increases breast cancer incidence by 41%
for women consuming 30-60 glday of alcohol compared to non-
d r i n k i n g women. 33
Apparently alcohol is good for heart disease and bad for breast can-
cer. The husband can have a drink with dinner but should never share
it with his wife. Is this a difference between men and women, or is this
a difference in response between heart disease and cancer? Do you feel
more informed or more confused?
Then there are those wonderful omega-3 fatty acids. Some types of
fish contain relatively large amounts of these fats and have been getting