The China Study (14 page)

85
LESSONS FROM CHINA
implied that diet and lifestyle were the principle causes of these dis-
eases. It also suggested that genes are not necessarily that important.
As noted earlier, a very prominent report by Sir Richard Doll and Sir
Richard Peto of the University of Oxford (U.K.) submitted to the u.s.
Congress summarized many of these studies and concluded that only
2-3% of all cancers could be attributed to genes. 4
Do the data from these international and migrant studies mean
that we can lower our rate of breast cancer to almost zero if we make
perfect lifestyle choices? The information certainly suggests that this
could be the case. Concerning the evidence in Chart 4.7, the solution
seems obvious: if we eat less fat, then we'll lower our breast cancer risk.
Most scientists made this conclusion and some surmised that dietary
fat caused breast cancer. But that interpretation was too simple. Other
charts prepared by Professor Carroll were largely, almost totally, ignored
(Charts 4.8 and 4.9). They show that breast cancer was associated with
animal fat intake but not with plant fat.
In rural China, dietary fat intake (at the time of the survey in 1983)
was very different from the United States in two ways. First, fat was only
14.5% of calorie intake in China, compared with about 36% in the U.s.
Second, the amount of fat in the diets of rural China depended almost
entirely on the amount of animal-based food in the diet, just like the
CHART 4.8: ANIMAL FAT INTAKE AND BREAST CANCER
25         FEMALE                               ,NETHERLANDS
,UK ,DENMARK
CANADA'        'NEW ZEALAND
SWITZERLAND,
,IRELAND
BELGIUM" US
GERMANY. ,SWEDEN 'AUSTRALIA
,    '     ,NORWAY
AUSTRIA 'FRANCE
'ITALY
,
, CZECH
,FINLAND
'PORTUGAL
HUNGARY
HONG KONG
BULGARIA C~ILE            ' POLAND
VENEZUELA\"'ROMANIA
PANAM!GREECE~p!NYUGOSLAVIA
PHILIPPINES. COLOMBIA      , PUERTO RICO
JAPAN',' MEXICO
EYlON    'TAIWAN
• THAILAND
, 'El SALVADOR
o
o                                                                          180
140     160
20       40      60        80       100      120
Animal Fat Intake (g/day)

86                                    THE (HINA STUDY
CHART 4.9: PLANT FAT INTAKE AND BREAST CANCER
FEMALE
e NETHERLANDS
25
e e DENMARK
UKe
e
NEW ZEALAND    CANADA
eSWITZERLAND
e IRELAND  BELGIUM e
0..
e
o
USA
~ 20
e AUSTRALIA
o                                                    eSWEDEN
o
6
NORWAye         AUSTRIA e e GERMANY
o
e FRANCE elTALY
......
2        15                                 e CZECHOSLOVAKIA
ro
cr:                FINLANDe eHUNGARY
.r:.                                                             e PORTUGAL
+-'
ro
(l)
o
10              POLAND e       HONG KONGeeBULGARIA
""0
~                               CHILE e       e                                      eSPAIN
ROMANIAe VENEZUELA
:J
~
e        epANAMA                      eGREECE
e PUERTO RICO     YUGOSLAVIA
(l)
PHILIPPINEse eCOLOMBIA
::t       5
e MEXICO
TAIWANe e JAPAN
eCEYLON
e EL SALVADOR
THAILAND e
o
o                                                                                70
60
10        20            30          40            50
Vegetable Fat Intake (g/day)
findings in Chart 4.7. The correlation between dietary fat and animal
protein in rural China was very high, at 70-84%,33 similar to the 93%
seen when comparing different countries. 25
This is important because in China and the international studies, fat
consumption was only an indication of animal-based food consumption.
Thus, the association between fat and breast cancer might really be tell-
ing us that as consumption of animal-based foods goes up, so does breast
cancer. This is not the case in the U.s., where we selectively add or re-
move fat from our foods and our diets. We get as much or more fat from
plant-based food (potato chips, French fries) as we get from processed
animal-based foods (skim milk, lean cuts of meat) . China does not tin-
k e r with fat in their food supply as we do here.
At this very low range of dietary fat in China, from 6%-24%, I ini-
tially thought that dietary fat would not be linked with diseases like
heart disease or the various cancers, as it is in the West. Some people
in the U.S.-like many of my colleagues in science and medicine-call
a 30% fat diet a "low-fat" diet. Therefore, a low-fat diet containing only

LESSONS FROM CHINA                            87
25-30% fat was thought to be low enough to obtain the maximum
amount of health benefits. This implied that going lower provided no
further benefit. Surprise!
Findings from rural China showed that reducing dietary fat from
24% to 6% was associated with lower breast cancer risk. However, lower
dietary fat in rural China meant less consumption not only of fat but,
more importantly, of animal-based food.
This connection of breast cancer with dietary fat, thus with animal-
based foods, brought into consideration other factors that also place a
woman at risk for breast cancer:
•   Early age of menarche (age of first menstruation)
•   High blood cholesterol
•   Late menopause
•   High exposure to female hormones
What does the China Study show regarding these risk factors? Higher
dietary fat is associated with higher blood cholesteroF and both of these
factors, along with higher female hormone levels, are associated, in
turn, with more breast cancer! and earlier age of menarche.!
The much later age of menarche in rural China is remarkable. Twen-
ty-five women in each of the l30 villages in the survey were asked when
they had their first menstrual period. The range of village averages was
fifteen to nineteen years, with an average of seventeen years. The U.S.
average is roughly eleven years!
Many studies have shown that earlier menarche leads to higher risk for
breast cancer.34 Menarche is triggered by the growth rate of the girl; the
faster the growth, the earlier the age of onset. It also is well established
that rapid growth of young girls often leads to greater adult body height
and more body weight and body fatness, each of which is associated with
higher breast cancer risk. Early age of menarche, both in Chinese and in
Western women, also leads to higher levels of blood hormones such as
estrogen. These hormone levels remain high throughout the reproductive
years if consumption of a diet rich in animal-based food is maintained.
Under these conditions, age of menopause is deferred by three to four
years,! thus extending the reproductive life from beginning to end by
about nine to ten years and greatly increasing lifetime exposure to female
hormones. Other studies have shown that an increase in years of repro-
ductive life is associated with increased breast cancer risk. 35,36
This network of relationships becomes still more impressive. Higher

88                            THE CHINA STUDY
fat consumption is associated with higher blood levels of estrogen dur-
i n g the critical years of thirty-five to forty-four years Ill and higher blood
levels of the female hormone prolactin during the later years of fifty-five
to sixty-four years.Ill These hormones are highly correlated with animal
protein intakeIll and milkIll and meatY Unfortunately, we could not dem-
o n s t r a t e whether these hormone levels were directly related to breast
cancer risk in China because the rate of disease is so low. 3?
Nonetheless, when hormone levels among Chinese women were com-
p a r e d with those of British women,38 Chinese estrogen levels were only
about one-half those of the British women, who have an eqUivalent hor-
m o n e profile to that of American women. Because the length of the repro-
ductive life of a Chinese woman is only about 75% of that of the British
(or American) woman, this means that with lower estrogen levels, the
Chinese woman only experiences about 35-40% of the lifetime estrogen
exposure of British (and American) women. This corresponds to Chinese
breast cancer rates that are only one-fifth of those of Western women.
The strong association of a high-animal protein, high-fat diet with
reproductive hormones and early age of menarche, both of which raise
the risk of breast cancer, is an important observation. It makes clear that
we should not have our children consume diets high in animal-based
foods. If you are a woman, would you ever have imagined that eating
diets higher in animal-based foods would expand your reproductive life
by about nine to ten years? As an aside, an interesting implication of
this observation, as noted by Ms. magazine founder Gloria Steinem, is
that eating the right foods could reduce teenage pregnancy by delaying
the age of menarche.
Beyond the hormone findings, is there a way to show that animal-
b a s e d food intake relates to overall cancer rates? This is somewhat dif-
ficult, but one of the factors we measured was how much cancer there
was in each family. Animal protein intake was convincingly associated
in the China Study with the prevalence of cancer in families. III This as-
s o c i a t i o n is an impressive and significant observation, considering the
unusually low intake of animal protein.
Diet and disease factors such as animal protein consumption or
breast cancer incidence lead to changes in the concentrations of cer-
tain chemicals in our blood. These chemicals are called biomarkers.
As an example, blood cholesterol is a biomarker for heart disease. We
measured six blood biomarkers that are associated with animal protein
intake. 39 Do they confirm the finding that animal protein intake is asso-

89
LESSONS FROM CHINA
cia ted with cancer in families? Absolutely. Every single animal protein-
related blood biomarker is significantly associated with the amount of
cancer in a family.lI-m
In this case, multiple observations, tightly networked into a web,
show that animal-based foods are strongly linked to breast cancer. What
makes this conclusion especially compelling are two kinds of evidence.
First, the individual parts of this web were consistently correlated and,
in most cases, were statistically significant. Second, this effect occurred
at unusually low intakes of animal-based foods.
Our investigation of breast cancer (detailed further in chapter seven)
is a perfect example of what makes the China Study so convincing.
Rather than a single, simple association of fat and breast cancer,! we were
able to construct a much more expansive web of information about how
diet affects breast cancer risk. We were able to examine in multiple ways
the role of diet and cholesterol, age of menarche and female hormone
levels, all of which are known risk factors for breast cancer. When each
new finding pointed in the same direction, we were able to see a picture
that was convincing, consistent and biologically plausible.
THE IMPORTANCE OF FIBER
The late Professor Denis Burkitt, of Trinity College, Dublin, was un-
usually articulate. His common sense, scientific credibility and sense
of humor made quite an impression on me when 1 first met him at a
Cornell seminar. The subject of his work was dietary fiber. He had trav-
eled 10,000 miles in ajeep over rugged countryside to study the dietary
habits of Africans.
He asserted that even though fiber was not digested, it was vital for
good health. Fiber was able to pull water from the body into the intestines
to keep things moving along. These undigested fibers, like stick-urn pa-
per, also gather up nasty chemicals that find their way into our intestines
and that might be carcinogenic. If we don't consume enough fiber, we are
susceptible to constipation-based diseases. According to Burkitt, these in-
clude large bowel cancer, diverticulosis, hemorrhoids and varicose veins.
In 1993, Dr. Burkitt was awarded the prestigious Bower Award, the
richest award in the world next to the Nobel Prize. He invited me to
speak at his award ceremony at the Franklin Institute in Philadelphia,
only two months before his unfortunate passing. He offered his opinion
that our China Study was the most significant work on diet and health
in the world at that time.

90                           THE CHINA STUDY
Dietary fiber is exclusively found in plant-based foods. This mate-
rial, which gives rigidity to the cell walls of plants, comes in thousands
of different chemical variations. It is mostly made of highly complex
carbohydrate molecules. We digest very little or no fiber. Nonetheless,
fiber, having few or no calories itself, helps dilute the caloric density of
our diets, creates a sense of fullness and helps to shut down appetite,
among other things. In doing so, it satisfies our hunger and minimizes
the overconsumption of calories.
Average fiber intake (Chart 4.lO) is about three times higher in China
than in the U.S. 40 These differences are exceptional, especially consider-
i n g the fact that many county averages were even much higher.
But according to some "experts" in the U.s., there is a dark side to
dietary fiber. They contend that if fiber intake is too high our bodies
are not able to absorb as much iron and related minerals, which are
essential for health. The fiber may bind with these nutrients and carry
them through our system before we are able to digest them. They say
that the maximum level of fiber intake should be around thirty to thirty-
five grams per day, which is only about the average intake of the rural
Chinese.
We studied this iron/fiber issue very carefully in the China Study. As
it turns out, fiber is not the enemy of iron absorption as so many experts
claim it to be. We measured how much iron the Chinese were consum-
i n g and how much was in their bodies. Iron was measured in six differ-
e n t ways (four blood biomarkers and two estimates of iron intake) and
CHART 4.10: AVERAGE INTAKES OF DIETARY FIBER, GM/DAY
40
30
20
10
o
China                       United States

LESSONS FROM CHINA                              91
when we compared these measurements with fiber intake, there was no
evidence showing that increasing fiber intake impaired iron absorption in
the body. In fact, we found the opposite effect. A good indicator of how
much iron is in the blood, hemoglobin, actually increased with greater
intakes of dietary fiber. I As it turns out, high-fiber foods, like wheat
and corn (but not the polished rice consumed in China) also happen to
be high in iron, meaning that the higher the consumption of fiber, the
higher the consumption of ironYI Iron intake in rural China (34 mg/
day) was surprisingly high when compared to the average American in-
take (18 mg/day) and it was far more associated with plant-based foods
than with animal-based foodsY
The China findings on dietary fiber and iron, like so many other ob-
s e r v a t i o n s in this study, did not support the common view of Western
scientists. People who consume more plant-based foods, thus more di-
etary fiber, also consume more iron,III all of which results in statistically
significant higher levels of hemoglobin. Unfortunately, a bit of confu-
s i o n has arisen over the fact that some people in rural China, including
women and children, have low iron levels. This is especially true in
areas where parasitic diseases are more common. In areas of rural China
where parasitic diseases were more common, iron status was lower. I
This has given some the opportunity to claim that these people need
more meat, but the evidence indicates that the problem would be much
better corrected by redUcing parasitism in these areas.
Much of the initial interest in dietary fiber arose with Burkitt's travels
in Africa and his claim that large bowel cancer is lower among popula-
t i o n s who consume high-fiber diets. Burkitt made this claim popular
but the story is at least 200 years old. In England during the late eigh-
t e e n t h century and early nineteenth century, it was claimed by some of
the leading phYSicians that constipation, which was associated with less
bulky diets (i.e., low-fiber diets), was associated with a higher risk of
cancer (usually breast and "intestinal" cancers).
At the beginning of the China Study, this belief that fiber might pre-
v e n t large bowel cancer was the prevailing view, although the 1982 Na-
tional Academy of Sciences Committee on Diet, Nutrition and Cancer,
"found no conclusive evidence to indicate that dietary fiber ... exerts a
protective effect against colorectal cancer in humans." The report went
on to conclude, " ... if there is such an effect, specific components of
fiber, rather than total dietary fiber, are more likely to be responsible."20
In hindSight, our discussion of this issue was inadequate. The question,