The China Study (10 page)

55
TURNING OFF CANCER
CHART 3.4: DIETARY PROTEIN AND FOCI FORMATION
3.5
3.0
2.5
OJ
2.0
VI
c
0
0..
VI
1.5
OJ
0:::
'0
0
1.0
l.L
.5
0
20%
5%
Dietary Protein Level
CHART 3.S: CARCINOGEN DOSE VERSUS PROTEIN INTAKE
100 /
./
90
80
70
OJ
VI
c
60
0
0..
VI
50
OJ
0:::
'0
40
0
l.L
30
20
.....
10
...
./
~
1
0 1.1
LowAF
High AF
High Protein
Low Protein

THE (HINA STUDY
56
still have an effect? We investigated this question by giving two groups
of rats either a high-aflatoxin dose or a low-aflatoxin dose, along with a
standard baseline diet. Because of this the two groups of rats were starting
the cancer process with different amounts of initiated, cancerous "seeds."
Then, during the promotion phase, we fed a low-protein diet to the high-
aflatoxin dose groups and a high-protein diet to the low-aflatoxin dose
group. We wondered whether the animals that start with lots of cancerous
seeds are able to overcome their predicament by eating a low-protein diet.
Again, the results were remarkable (Chart 3.5). Animals starting with
the most cancer initiation (high-aflatoxin dose) developed substantially
less foci when fed the 5% protein diet. In contrast, animals initiated with
a low-aflatoxin dose actually produced substantially more foci when sub-
s e q u e n t l y fed the 20% protein diet.
A principle was being established. Foci development, initially deter-
m i n e d by the amount of the carcinogen exposure, is actually controlled
far more by dietary protein consumed during promotion. Protein dur-
ing promotion trumps the carcinogen, regardless of initial exposure.
With this background information we deSigned a much more sub-
stantial experiment. Here is a step-by-step sequence of experiments,
carried out by my graduate student Linda Youngman.35 All animals were
dosed with the same amount of carcinogen, then alternately fed either
5% or 20% dietary protein during the twelve-week promotion stage.
We divided this twelve-week promotion stage into four periods of three
weeks each. Period 1 represents weeks one to three, period 2 represents
weeks four to six, and so on.
When animals were fed the 20% protein diet during periods 1 and 2
(20-20), foci continued to enlarge, as expected. But when animals were
switched to the low-protein diet at the beginning of period 3 (20-20-
5) , there was a sharp decrease in foci development. And, when animals
were subsequently switched back to the 20% protein diet during period
4 (20-20-5-20), foci development was turned on once again.
In another experiment, in animals fed 20% dietary protein during
period 1 but switched to 5% dietary protein during period 2 (20-5), foci
development was sharply decreased. But when these animals were re-
t u r n e d to 20% dietary protein during period 3 (20-5-20), we again saw
the dramatic power of dietary protein to promote foci development.
These several experiments, taken together, were quite profound. Foci
growth could be reversed, up and down, by switching the amount of
protein being consumed, and at all stages of foci development.

57
TURNING OFF CANCER
These experiments also demonstrated that the body could "remem-
b e r " early carcinogen insults,35, 36 even though they might then lie
dormant with low protein intake. That is, exposure to aflatoxin left a
genetic "imprint" that remained dormant with 5% dietary protein until
nine weeks later when this imprint reawakened to form foci with 20%
dietary protein. In simple terms, the body holds a grudge. It suggests
that if we are exposed in the past to a carcinogen that initiates a bit of
cancer that remains dormant, this cancer can still be "reawakened" by
bad nutrition some time later.
These studies showed that cancer development is modified by relative-
ly modest changes in protein consumption. But how much protein is too
much or too little? Using rats, we investigated a range of 4-24% dietary
protein (Chart 3.637 ). Foci did not develop with up to about 10% dietary
protein. Beyond 10%, foci development increased dramatically with in-
creases in dietary protein. The results were later repeated a second time in
my laboratory by a visiting professor from Japan, Fumiyiki Horio.38
CHART 3.6: FOCI PROMOTION BY DIETARY PROTEIN
90
80
70
60
+-'
Adequate Protein
C
ClJ
E                  for Body Growth
50
0.
o
Qj
>
ClJ
40
o
'u
o
LL
30
20
10
4     6     8    10    12   14 20
% Dietary Protein

58                           THE CHINA STUDY
The most significant finding of this experiment was this: foci devel-
o p e d only when the animals met or exceeded the amount of dietary
protein (12%) needed to satisfy their body growth rate. 39 That is, when
the animals met and surpassed their requirement for protein, disease
onset began.
This finding may have considerable relevance for humans even
though these were rat studies. I say this because the protein required
for growth in young rats and humans as well as the protein required to
maintain health for adult rats and humans is remarkably similar.4o,41
According to the recommended daily allowance (RDA) for protein
consumption, we humans should be getting about 10% of our energy
from protein. This is considerably more than the actual amount required.
But because requirements may vary from individual to individual, 10%
dietary protein is recommended to insure adequate intake for virtually
all people. What do most of us routinely consume? Remarkably, it is
considerably more than the recommended 10%, The average American
consumes 15-16% protein. Does this place us at risk for getting cancer?
These animal studies hint that it does.
Ten percent dietary protein is equivalent to eating about 50-60 grams
of protein per day; depending on body weight and total calorie intake. The
national average of 15-16% is about 70-100 grams of protein per day; with
men at the upper part of the range and women at the lower end, In food
terms, there are about twelve grams of protein in 100 calories of spinach
(fifteen ounces) and five grams of protein in 100 calories of raw chick peas
Gust over two tablespoons). There are about thirteen grams of protein in
100 calories of porterhouse steak Gust over one and a half ounces).
Yet another question was whether protein intake could modify the
all-important relationship between aflatoxin dose and foci formation.
A chemical is usually not considered a carcinogen unless higher doses
yield higher incidences of cancer. For example, as the aflatoxin dose
becomes greater, foci and tumor growth should be correspondingly
greater. If an increasing response is not observed for a suspect chemical
carcinogen, serious doubt arises whether it really is carcinogenic.
To investigate this dose-response question, ten groups of rats were
administered increasing doses of aflatoxin, then fed either regular levels
(20%) or low levels (5-10%) of protein during the promotion period
(Chart 3.7 34 ) .
In the animals fed the 20% level of protein, foci increased in number and
size, as expected, as the aflatoxin dose was increased. The dose-response

59
TURNING OFF CANCER
CHART 3.7: AFLATOXIN DOSE-FOCI RESPONSE
(])
Vl
C
o
a.
_   20% Protein
Vl
(])
a::
_     5% Protein
·0
o
u..
235      275               350
200                      300
Aflatoxin Dose (mcglkg body weight/day)
relationship was strong and clear. However, in the animals fed 5% protein,
the dose-response curve completely disappeared. There was no foci response,
even when animals were given the maximum tolerated aflatoxin dose. This
was yet another result demonstrating that a low-protein diet could over-
ride the cancer-causing effect of a very powerful carcinogen, aflatoxin.
Is it possible that chemical carcinogens, in general, do not cause
cancer unless the nutritional conditions are "right"? Is it possible that,
for much of our lives, we are being exposed to small amounts of cancer-
causing chemicals, but cancer does not occur unless we consume foods
that promote and nurture tumor development? Can we control cancer
through nutrition?
NOT ALL PROTEINS ARE ALIKE
If you have followed the story so far, you have seen how provocative
these findings are. Controlling cancer through nutrition was, and still
is, a radical idea. But as if this weren't enough, one more issue would
yield explOSive information: did it make any difference what type of
protein was used in these experiments? For all of these experiments,
we were using casein, which makes up 87% of cow's milk protein. So
the next logical question was whether plant protein, tested in the same
way, has the same effect on cancer promotion as casein. The answer is
an astonishing "NO." In these experiments, plant protein did not promote
cancer growth, even at the higher levels of intake. An undergraduate pre-
medical student doing an honors degree with me, David Schulsinger,
did the study (Chart 3.842 ) . Gluten, the protein of wheat, did not produce
the same result as casein, even when fed at the same 20% level.

60                            THE CHINA STUDY
CHART 3.8: PROTEIN TYPE AND FOCI RESPONSE
100
OJ
.20% Casein
Vl
C
o
0..
.20% Gluten
lD       50
0:::
0 5 % Casein
'0
o
L.L
o   -i"---
Protein Type
We also examined whether soy protein had the same effect as casein
on foci development. Rats fed 20% soy protein diets did not form. early foci,
just like the 20% wheat protein diets. Suddenly protein, milk protein in this
case, wasn't looking so good. We had discovered that low protein intake
reduces cancer initiation and works in multiple synchronous ways. As if
that weren't enough, we were finding that high protein intake, in excess
of the amount needed for growth, promotes cancer after initiation. Like
flipping a light switch on and off, we could control cancer promotion
merely by changing levels of protein, regardless of initial carcinogen
exposure. But the cancer-promoting factor in this case was cow's milk
protein. It was difficult enough for my colleagues to accept the idea that
protein might help cancer grow, but cow's milk protein? Was I crazy?
ADDITIONAL QUESTIONS
For those readers who want to know somewhat more, I've includ-
e d a few questions in AppendiX A.
THE GRAND FINALE
Thus far we had relied on experiments where we measured only the ear-
ly indicators of tumor development, the early cancer-like foci. Now, it
was time to do the big study, the one where we would measure complete
tumor formation. We organized a very large study of several hundred
rats and examined tumor formation over their lifetimes using several
different approaches. 36, 43

TURNING OFF CANCER                            61
The effects of protein feeding on tumor development were nothing
less than spectacular. Rats generally live for about two years, thus the
study was 100 weeks in length. All animals that were administered afla-
t o x i n and fed the regular 20% levels of casein either were dead or near
death from liver tumors at 100 weeks.36 . 43 All animals administered the
same level of aflatoxin but fed the low 5% protein diet were alive, active
and thrifty, with sleek hair coats at 100 weeks . This was a virtual 100 to
o score, something almost never seen in research and almost identical
to the original research in India. 1   6
In this same experiment,36 we switched the diets of some rats at either
forty or sixty weeks, to again investigate the reversibility of cancer pro-
motion. Animals switched from a high-protein to a low-protein diet had
significantly less tumor growth (35°tb-40% less!) than animals fed a high-
p r o t e i n diet. Animals switched from a low-protein diet to a high-protein
CHART 3.9A: TUMOR DEVELOPMENT AT 100 WEEKS
3330
QJ
VI
c
o
0..
VI
QJ
': 2350
o
E
:J
f-
:J
LL
240
6%                     22%
14%
CHART 3.98: EARLY FOCI, "LlFETIME"
40
6
QJ
VI
C
0
0..                                                             •      14
VI
.... 20
QJ
'u
o
0
LL
22
0
% Dietary Casein

THE CHINA STUDY
62
diet halfway through their lifetime started growing tumors again. These
findings on full-blown tumors confirmed our earlier findings using foci.
Namely, nutritional manipulation can turn cancer "on" and "off."
We also measured early foci in these "lifetime" studies to see if their
response to dietary protein was similar to that for tumor response. The
correspondence between foci growth and tumor growth could not have
been greater (Chart 3. 9a). 36,43
How much more did we need to find out? I would never have dreamed
that our results up to this point would be so incredibly consistent, bio-
logically plausible and statistically significant. We had fully confirmed the
original work from India and had done it in exceptional depth.
Let there be no doubt: cow's milk protein is an exceptionally potent
cancer promoter in rats dosed with aflatoxin. The fact that this promotion
effect occurs at dietary protein levels 00-20%) commonly used both in
rodents and humans makes it especially tantalizing-and provocative.
OTHER CANCERS, OTHER CARCINOGENS
Okay, so here's the central question: how does this research apply to hu-
m a n health and human liver cancer in particular? One way to investigate
this question is to research other species, other carcinogens and other
organs. If casein's effect on cancer is consistent across these categories,
it becomes more likely that humans better take note. So our research be-
came broader in scope, to see whether our discoveries would hold up.
While our rat studies were underway, studies were published44 , 4 5
claiming that chronic infection with hepatitis B virus (HBV) was the
major risk factor for human liver cancer. It was thought that people who
remained chronically infected with HBV had twenty to forty times the
risk of getting liver cancer.
Over the years, considerable research had been done on how this
virus causes liver cancer. 46 In effect, a piece of the virus gene inserts
itself into the genetic material of the mouse liver where it initiates liver
cancer. When this is done experimentally the animals are considered
transgenic.
Virtually all of the research done in other laboratories on HBV trans-
genic mice-and there was a lot of it-was done primarily to understand
the molecular mechanism by which HBV worked. No attention was
given to nutrition and its effect on tumor development. I watched with
some amusement for several years how one community of researchers
argued for aflatoxin as the key cause of human liver cancer and another