The Curious Case of Campbell’s Rats — Does Protein Deficiency Prevent Cancer?

By now, we’ve all heard of The China Study. First T. Colin Campbell, a lifetime expert researcher and policy maker at the highest levels, made it a best seller.  Vegetarians the world over and many others hailed the book as proof that animal foods are harmful and that purging them from our diets and replacing them with whole plant foods is the key to vibrant, lasting health.  There were a few dissenting critical reviews, including Anthony Colpo’s and my own, which for years were the go-to articles for anyone looking for an alternative view of the China Study.

Then, like earth-shattering thunder falling from the sky, just a couple months ago Denise Minger produced a massive critique of the China Study that turned many of its claims upside down, sending a shock wave through entire blogosphere and drawing the attention even of Dr. Campbell himself.  Minger’s analysis tore apart many of the most important statistical claims of the China Study using data from the original monograph of Campbell’s massive epidemiological study bearing that name, and brought to light a critical piece of information refuting once and for all Campbell’s claims that plant proteins act differently than animal proteins.

Now that the blogosphere is abuzz with China Study debate more than ever, it’s time to revisit the curious case of Campbell’s rats.  Does animal protein cause cancer?  Dr. Campbell conducted two decades of rigorous animal research addressing this question, funded by the National Institutes of Health, one of the most reputable sources of public funding in the land.  He would have us believe that it does.

Herein, however, we will take a wild ride through these decades of animal studies, discovering many glaring omissions and arriving at many new, unanswered questions.  Campbell’s animal research has, in fact, raised critically important questions about the ability of dietary protein to promote the growth of cancers once they are formed.  His failure to tell us that high levels of dietary protein offer equally dramatic protection against the initiation of cancer and that rats fed low levels of protein have many health problems of their own, however, unfortunately obscures the true importance of his work.

Buckle up everyone!  The wild ride is about to begin…

In This Blog Post

An Obscure Study From India — Low-Protein Diets Save Rats From Cancer But Kill Them Instead

Campbell tells the story like this.  In 1965, he took a faculty position at Virginia Tech, then still an advocate of animal protein as good, nourishing American fare.  In 1967, he accepted an invitation from a department head at that university to travel to the Phillipines with the task of alleviating childhood malnutrition and making sure peanuts could provide good protein without the potential harms of aflatoxin, a carcinogenic mold toxin with which peanuts are often contaminated.

A shocking revelation then came in two-fold form: first an epidemiological study suggested that liver cancer was rampant among Filipino children and that the “best-fed” rather than the malnourished children were the ones most ravaged by the disease; then, in 1968, “a research paper from India surfaced in an obscure medical journal” showing that aflatoxin only produced liver cancer in rats when they were fed high levels of casein, a milk protein.  Campbell was surprised and skeptical, but he attempted to replicate these findings, and thus was born his two-decade research program showing that animal protein, but not plant protein, was the single most important trigger that turns cancer “on” like a light switch.

Campbell never tells us, however, that these Indian researchers actually published this paper as part of a two-paper set, one showing that low-casein diets make aflatoxin much more acutely toxic to rats (1), and the other showing that these same diets make aflatoxin much less carcinogenic (2).

In the very paper (2) that Campbell cites as “a revelation to die for,” showing that a high-protein diet turns the cancer switch to the “on” position, the low-protein diet proved lethal to the animals.  The investigators gave rats a small dose of aflatoxin every day for six months and fed them either a 5 percent casein or 20 percent casein diet.  The experiment carried on for two years, in fact, but they stopped adminstering aflatoxin at six months for the simple reason that half the animals on the low-protein diet had died.  They had typical symptoms of aflatoxin toxicity including liver necrosis (cell death), proliferation of bile duct tissue, and fatty liver.

All the animals receiving 20 percent casein, on the other hand, were still alive at that point.  For the remainder of the two years, the rats receiving 20 percent casein continued to live longer, but many of them developed liver cancer or pre-cancerous changes, while none of the rats fed 5 percent casein developed liver cancer.

What a trade-off!  Somehow, I doubt many people would read this study and shout “sign me up!” for a low-protein, plant-based diet if it is going to save them from cancer at the expense of killing them in their youth.

Campbell’s Protein-Deficient Rats

Campbell writes on page 51 of The China Study that he first investigated the effect of dietary casein on drug-metabolizing enzymes in order to test the hypothesis that low-protein diets might protect against cancer:

How does protein intake affect cancer initiation? Our first test was to see whether protein intake affected the enzyme principally responsible for aflatoxin metabolism, the mixed function oxidase (MFO). . . . At the time we started our research, we hypothesized that the protein we consume alters tumor growth by changing how aflatoxin is detoxified by the enzymes present in the liver. . . . Decreasing protein intake like that done in the original research in India (20% to 5%) not only greatly decreased enzyme activity but did so very quickly. What does this mean? Decreasing enzyme activity via low-protein diets implied that less aflatoxin was being transformed into the dangerous aflatoxin metabolite that had the potential to bind and to mutate the DNA.

Strangely, however, Campbell’s first paper on this topic, published in 1972 (3), doesn’t even contain the word “cancer.”  Instead, it starts off by discussing aflatoxin toxicity.  Here is the first sentence:

A deficiency of dietary protein was shown to increase the toxicity of aflatoxin for rats (1,2).

Campbell and his graduate student referred to their model as “protein deficiency” throughout the paper.  As another example, this is the first sentence of their abstract:

The effect of protein deficiency in male weanling rats on the activity of the hepatic microsomal enzyme system was studied.

Perhaps Campbell continued to refer to his model as “protein deficiency” during the 1970s because he was trying to slip the provocative nature of his research under the radars of reviewers who would otherwise consider him a “heretic,” a concern he describes repeatedly in his book.  Or perhaps as a skeptical scientist he was still yet to be convinced of the virtues of a low-protein diet.  As we will see below, however, this study actually confirmed the then-conventional view that 5% casein diets were deficient in protein.

Let’s first consider these rats’ complete and total failure to grow.  The rats on the 5% casein diet ate much less food than the rats on the 20% casein diet.  Expecting this, Campbell and his student divided the rats into three groups rather than two.  They fed the third  group 20% casein but restricted their total food intake to the measly amount of food the 5% casein group was eating spontaneously.  Here is a graph of their food intake:

T. Colin Campbell's low-protein Sprague Dawley rats ate less than their high-protein counterparts.
Campbell’s Sprague-Dawley rats, having just been weaned, were three weeks old at the beginning of the study.  Rats only live about two years if they are lucky, but even for rats three weeks is still just a baby.  Here is a graph produced by Harlan, a company that sells Sprague Dawley rats, showing their expected growth during the first thirteen weeks of their lives:

Campbell's rats should have grown from 50 to 100 grams during the two weeks of the study.
According to this graph, rats fed an amount of protein that Harlan and the scientific community in general consider adequate (18%) grow from 50 grams to over 100 grams during the course of time corresponding to the duration of Campbell’s study, indicated by the large red bracket.  Among Campbell’s rats, however, only the rats eating the 20% casein diet achieved this body weight:

Campbell's low-protein rats achieved a weight only half of what should be expected.
The 5% rats achieved only half the body weight expected for their age.  They gained much less weight than the pair-fed high-protein group, even though the two groups were eating the same amount of calories.  The differences are even more dramatic if we consider the growth of these rats once beginning the low-protein diet at three weeks of age:

Campbell's three-week-old low-protein rats didn't grow at all over the course of the study.
The 5% rats hardly grew at all!  Certainly, Dr. Campbell makes an important point repeatedly throughout The China Study: the amount of protein that maximizes growth may not be the amount of protein that maximizes health.  How many of us, however, would deliberately feed a two-year old a diet that would cause them to stop growing altogether?

The signs of deficiency didn’t stop at failure to grow.  The animals also developed fatty livers.  Even the decrease in the level of drug-detoxifying enzymes could be seen as a symptom of deficiency.  Indeed, Campbell suggested it was due either to a disruption of cell proliferation that stunted the growth of the liver or to a disruption of protein synthesis.  Here is a quote from the discussion where Campbell describes the fatty liver and likens the decrease in cell proliferation to the retardation of brain growth that occurs in malnourished animals:

First, the reduced DNA content could be indicative of a lower cell number per gram of liver and would accordingly imply larger cells in the protein-deprived group. These cells could be larger in response to lipid infiltration since the livers of the low protein group were observed to be very fatty. Consequently, the normal rate of cell proliferation would have been decreased during protein deprivation, which is similar to the retardation of brain cell growth of young malnourished animals described by Winick and Rosso (18).

In another study published eight years later in 1980 (4), Campbell replicated the initial Indian report that had showed low-protein diets to dramatically increase the acute toxicity of aflatoxin.  Campbell and K. D. Mainigi fed rats high- and low-protein diets with or without aflatoxin.  They spiked the 20% casein diet with five parts per million (5 ppm) aflatoxin, but spiked the 5% casein diet with only 2.5 ppm aflatoxin because “5 ppm was found to be lethal for this dietary group.”

Deficiency?  Certainly sounds like it.

But the increased susceptibility of rats fed low-protein diets to environmental toxins doesn’t stop at aflatoxin.  In the introduction to a 1978 paper further investigating the effects of low-protein diets on detoxification enzymes (5), Campbell and his colleagues described susceptibility to other environmental toxins on low-protein diets as well:

The toxicities of several pesticides have been shown to be markedly increased (2), such as that of captan which is increased 2,100 times by protein deficiency (3).

Campbell co-authored a review in 1978 entitled “The Effect of Quantity and Quality of Dietary Protein on Drug Metabolism” that described conflicting effects of low-protein diets on the suceptibility to different pesticides and other environmental toxins (6).  The authors compiled a table summarizing these findings.  The column showing compounds whose toxicity decreased on low-protein diets contained only three toxins.  The column showing compounds whose toxicity increased on low-protein diets, by contrast, contained a whopping, six-fold greater eighteen toxins.

Which would you place your odds with?

The increase in aflatoxin toxicity seen on the low-protein diets wasn’t just a matter of drug-metabolizing enzymes, however.  If only it were so simple.  Rather, the ways in which these low-protein diets compromise health are myriad.  Dr. Campbell and one of his undergraduate students co-authored a paper in 1989 in which all the rats were dosed with aflatoxin early on and were all fed 20% casein diets while aflatoxin was still in their systems (7).  Then, Campbell and company switched half of them to 5% casein diets.  The rats fed 5% casein once the aflatoxin was gone from their systems still showed greater symptoms of toxicity!

Campbell and his co-authors concluded in their final sentence:

This observation suggests that the low protein intake was not sufficient to allow for tissue recovery from the acute toxic effects.

Alas, we find that these low-protein diets made the rats eat less food, fail to grow, and unable to efficiently detoxify aflatoxin and a multitude of other toxins.  They destroyed their ability to repair damaged tissue, gave them fatty liver, stopped their internal organs from developing, and if the rats encountered toxic substances, the diets dug them an early grave.

Certainly, Campbell and his colleauges were justified in the 1970s in calling their 5% casein model “protein deficiency.”

Protein Deficiency Disappears Down the Memory Hole

While Campbell’s earlier scientific papers present a clear picture of protein deficiency in rats fed 5% casein, we get no sense from reading The China Study that these rats had anything other than perfect health.  This is how Campbell describes the health of animals on the low-protein diets in the appendix on page 352:

Many researchers have long assumed that animals fed diets this low in protein would not be healthy. However, the low-protein animals were healthier by every indication. They lived longer, were more physically active, were slimmer and had healthy hair coats at 100 weeks while the high-protein counterpart rats were all dead. Also, animals consuming less dietary casein not only ate more calories, but they also burned off more calories. Low-protein animals consumed more oxygen, which is required for the burning of these calories, and had higher levels of a special tissue called brown adipose tissue (5,6), which is especially effective in burning off calories. This occurs through a process of “thermogenesis,” i.e., the expenditure of calories as body heat. This phenomenon had already been demonstrated many years before (7-11). Low-protein diets enhance the burning off of calories, thus leaving less calories for body weight gain and perhaps also less for tumor growth as well.

The claim that rats on the low-protein diets ate more but weighed less, based on papers published in the 1990s, conflicts with Campbell’s earlier studies showing that rats actually ate less food on low-protein diets.  In 1980, three years before the launch of the massive epidemiological study in China bearing the same name as Campbell’s book, Campbell’s research group switched from using Sprague Dawley rats to Fisher 344 rats (4).  Unlike the Sprague Dawley rats used in the group’s earlier experiments, the Fisher 344 rats did not develop fatty liver when fed the 5% casein diets (8).  Campbell and his colleagues did not report food intakes in most of the papers they published between 1980 and 1989 using these rats (4, 7, 8, 911, 12, 13), but they reported in one of them that the level of dietary protein had no effect on food intake (10) and reported the same thing in a single 1985 study using Wistar rats (15).

In 1991, however, Campbell’s group published a study using Fisher 344 rats showing that rats fed 4% casein ate more and weighed less than rats fed greater amounts of casein ranging from 8% to 20% (16).  There were no differences between rats fed 8%, 12%, 16%, or 20% casein.  Consider this graph of food efficiency, which is the ratio of body weight gain to food intake:

Diets with only 4% casein (3.5% protein) decrease food efficiency, which Campbell asserts is a

Casein is only 87% protein, so the rats fed 4% casein were actually only consuming 3.5% protein.  This extremely low level of protein seemed to turn on “thermogenesis” like a light switch.  Since many fruits and most vegetables have more protein than this, it is difficult to see how anyone could possibly eat such a small amount of protein on a diet containing anything resembling food.

The “light switch” effect at such a low level of protein rather conspicuously suggests an effect of deficiency rather than some kind of benefit resulting from curbing an excess, as if 7% protein (8% casein) could truly be “excessive.”  One of the classical symptoms of essential fatty acid deficiency, for example, is that animals consume a massive amount of food but fail to gain weight (17).  Much more modest levels of protein restriction decrease all of the enzymes involved in producing arachidonic acid and DHA, the two physiologically essential fatty acids, from their dietary precursors (18).  One certainly has to wonder whether Campbell’s 4% casein rats were suffering from a mild essential fatty acid deficiency.

Apart from the virtual impossibility of consuming a food-containing diet with less less than 4% protein, except perhaps a well-crafted feast of fruitarian fare, it is difficult to see how we can extrapolate a “thermogenic” effect from rats to humans when we cannot even extrapolate the effect from one strain of rat to another.

But back to how “protein deficiency” disappeared down the memory hole.

While Fisher 344 rats failed to develop fatty liver on Campbell’s 5% low-protein diets, this seems to reflect a general immunity of this strain to fatty liver.  One recent study showed that Fisher 344 rats are also immune to fatty liver when fed 37% of their calories as alcohol (19).  The study showed that ethanol-fed Fisher rats had similar levels of liver fat as control rats of all strains, whereas Sprague Dawley rats and Long Evans rats quickly developed a liver stuffed with more fat than an Eskimo’s yummy dinner plate.

Campbell’s Fisher 344 rats still showed much greater vulnerability to aflatoxin toxicity on a low-protein diet.  Aflatoxin proved lethal to Fisher 344 rats fed low-protein diets when rats fed high-protein diets were immune to the same doses (4).  Low-protein diets prevented tissue repair in Fisher 344 rats (7).  Fisher 344 rats fed 5% casein developed the following symptoms when dosed with aflatoxin (11):

Some degree of bile duct proliferation was observed in all animals dosed with AFB1. However, the groups fed the 5% casein diet during the dosing period had relatively severe bile duct proliferation and cholangiofibrosis [fibrosis of the bile duct]. In these groups, the architecture of the liver was often distorted by fibrous septa. Groups fed the 20% casein diet during the dosing period had mild bile duct proliferation and no cholangiofibrosis.

Nevertheless, by 1991 Campbell was claiming in such prestigious journals as The Journal of Nutrition that the health of 5% casein rats was in every way superior to the health of rats fed higher levels of protein (20):

Although a 5% casein diet is not generally considered nutritionally adequate (i.e., it does not support maximal growth), for every health index we have thus far measured, the 5% casein diet supports better health in rats than does the 20% casein diet.

And thus disappeared all the protein deficiency symptoms Campbell had uncovered during his career, down the memory hole and locked away for decades.  By the time The China Study hit shelves, these findings had so many years of practice making the perfect disappearing act that they spent 417 whole pages disappearing into the oblivion of the forgotten past with exquisite mastery.

If protein had such a profound ability to protect against the toxic effect of aflatoxin, however, is it possible that it could also protect against its carcinogenesis?  This brings us to what is perhaps Dr. Campbell’s most glaring omission: that while high-protein diets promoted the growth of pre-cancerous lesions once they were formed, they protected against the initation of those lesions with just as much power.

The China Study’s Best-Kept Secret — Protein Protects Against Cancer Initiation

Dr. Campbell’s research on protein and cancer is fascinating.  He deserves extraordinary credit for his rigorous experiments and his provocative, even revolutionary findings.  Campbell’s research showed that nutritional factors such as protein intake exert dramatic effects on the initiation and growth of cancer, showing that genes and exposure to environmental toxins are only two small parts of the cancer story.  Nevertheless, Campbell seems to have become so enamored with the cancer-promoting effect of protein  and the dichotomy between plant and animal foods that his research left many questions unanswered and his claims that animal protein is the root of all disease jumped the proverbial gun in the extreme.

Campbell mostly studied the development of pre-cancerous lesions in the liver.  He also studied the development of true liver cancers over the course of 100 weeks, however, which is roughly the full lifetime of a rat (21).  This tremendous study suggested that pre-cancerous lesions can be used to predict the development of true tumors with 90-98% accuracy.

Campbell first showed that high-protein diets promote the the development of cancer in rats dosed with aflatoxin in 1982 (8).  Rats fed 20% casein developed four times as much pre-cancerous tissue as rats fed 5% casein.  A more extensive dosing study showed that changes in casein intake below 10% or above 20% had negligible effects on the development of these lesions, but as casein increased from 10% to 20% the cancer-promoting effect increased continuously (13).

Despite Campbell’s repeated suggestions throughout The China Study that nutritional effects are much more powerful than exposure to carcinogens, he published one study suggesting that the two factors were equally powerful (12).  When rats were all fed 20% casein, the dose that provided the maximal cancer-promoting effect, those dosed with 0.4 milligrams per kilogram body weight (0.4 mg/kg) or 1.0 mg/kg of aflatoxin failed to develop any pre-cancerous lesions at all.  Those given 1.5 mg/kg developed “only a barely detectable, but significant, response.”

This is rather ironic considering 1.5 mg/kg is 30 percent of the dose required to kill 50 percent of the animals even on the protective, high-casein diet.  On page 45 of  The China Study, Campbell mocks the high doses of carcinogens used in animal studies to show that carcinogen exposure, rather than nutritional factors such as protein intake, produce cancer:

Let’s look at one nitrosamine, NSAR. . . . How much NSAR did the rats get? Both groups of rats were given an incredible amount. Let me translate the “low” dose by giving you a little scenario. Let’s say you go over to your friend’s house to eat every meal. This friend is sick of you and wants to give you throat cancer by exposing you to NSAR. So he gives you the equivalent of the “low” level given to the rats. You go to his house, and your friend offers you a bologna sandwich that has a whole pound of bologna on it! You eat it. He offers you another, and another, and another . . . . You’ll have to eat 270,000 bologna sandwiches before your friend lets you leave. You better like bologna, because your friend is going to have to feed you this way every day for over thirty years! If he does this, you will have had about as much exposure to NSAR (per body weight) as the rats in the “low” dose group.

Campbell’s experiments are, of course, much more realistic than this scenario.  If your friend offered you peanut butter sandwiches with 100 grams worth of peanut butter contaminated with the maximum amount of aflatoxin allowed by the FDA, you’d only have to eat 270,000 peanut butter sandwiches for four days to obtain the dose of aflatoxin that produced a “barely detectable response” in Campbell’s study.  Still, 1,125,000 peanut butter sandwiches is an awful lot of peanut butter sandwiches and you’d better have one heck of a toothbrush.  Clearly, exposure to carcinogens is important.

In the first paper that Campbell published on the protein-cancer connection (8), he suggested in the introduction that high protein diets should promote the initiation of pre-cancerous lesions as well as their promotion of larger lesions and transformation into true cancers.  This suggestion, however, was highly speculative:

Dietary protein has been shown to modify the enzymatic activation of aflatoxin B1 (AFB1) in the formation of DNA adducts in rat liver [6]. Protein status should, therefore, influence at least the initiation of AFB1-induced hepatocarcinogenesis, if the extent of adduct formation is related to initiation. On the other hand, the effects of dietary protein in the promotional phase have not been very well characterized.

Campbell had conducted research showing that low-protein diets suppress the enzymatic detoxification of aflatoxin.  In doing so, they suppress the formation of an unstable intermediate that is capable of binding to DNA.  I added the bold and italics above to emphasize the point that this does not in and of itself show that low-protein diets protect against the initiation of pre-cancerous lesions.  In fact, Campbell later conducted a study to test this hypothesis, and its results promptly disappeared down the memory hole, just like so many other critical findings.

In this amazing experiment (11), Campbell’s group fed rats either 5% or 20% casein during the aflatoxin dosing period, when pre-cancerous lesions should be initiated, and either 5% or 20% casein during the 12 weeks after, when pre-cancerous lesions already formed should be promoted.  There were thus four groups of rats: one fed 20% the whole time, one fed 5% the whole time, one fed 5% during the initiation period and 20% during the promotion period, and one fed 20% during the initation period and 5% during the promotion period.  This was the first study where Campbell provided the low-protein diet rather than the high-protein “control” diet during the dosing period.  Let’s take a look at the results:

High-protein diets during the initiation period protected against the development of pre-cancerous lesions. Thereafter, Campbell used 20% casein during initiation and 5% casein during the promotion period and attributed the protection simply to the low-protein diet.
The rats fed 20% casein through the entire experimental period had somewhat more pre-cancerous lesions than the rats fed 5% during the whole period, but the difference is not very dramatic.  The dramatic difference we can see from this graph is between the second and third groups.  These results clearly show that while 20% casein provided during the promotion period promoted the growth of pre-cancerous lesions, 20% casein provided during the initiation period proved dramatically and powerfully protective.

In Campbell’s first protein-aflatoxin-cancer study published in 1982 (8) and in virtually every such study thereafter (10, 12, 13, 16, 20, 21, 22, 23, 24), Campbell and his research group used 20 percent casein for several weeks during the initation period for all the animals.  The fact that the dramatic reduction of pre-cancerous lesions in the rats fed 5% casein owed in part to the high-protein diet they were fed during the initation period was forever lost into the memory hole.

These findings should provoke a number of important questions.  Is there a level of dietary protein somewhere between 5% and 20% that provides maximal protection during both the initiation and promotion periods?  Is the effect of the high-protein diet during the promotion period a result of the protein itself, or is the protein raising the need for other nutrients needed to protect against cancer?  If so, can protein and those other nutrients be provided together to provide maximal protection during all phases of cancer development?

Rather than investigating these questions, Campbell seized on the adverse effect of protein when fed during the promotion period, and these questions persist unanswered.

Plant Vs. Animal Protein — Campbell Proved There’s No Difference

Campbell tells us on page 59 of The China Study that plant proteins act fundamentally differently than animal proteins.  Gluten, the protein of wheat, did not promote cancer, while casein, the protein of milk, promoted it powerfully.  This study (7), however, showed that gluten was just as powerful as casein when lysine, its limiting amino acid, was provided.  Campbell never tells us this in The China Study. Nor does he tell us that casein is just as much an incomplete protein as gluten and that the reason it proved so effective in promoting cancer in his models was because he supplemented all of the diets with methionine.  Casein’s limiting amino acids are methionine and cysteine, which can be converted into one another.  Thus methionine or cysteine make casein complete in the same way that lysine makes wheat protein complete.

In this paper, Campbell acknowledged that this was a general effect of protein, not something specific to specific proteins or to animal proteins:

[I]n 1945 Larsen and Heston found that the incidence of spontaneous pulmonary tumors was doubled in strain A mice fed low-casein diets supplemented with cystine (the most limiting amino acid).  Silverstone and Tannenbaum (14) showed that the development of spontaneous hepatomas was enhanced in C3H mice fed a gelatin-containing diet when methionine and cystine were added.  A review of the somewhat limited data from these and earlier studies (1) indicated that inhibition of tumor development as a result of marginal intakes of various proteins could be abolished by supplementation with the respective limiting amino acid for each protein. . . . [O]ur results suggest that the enhancement of focus development by lysine supplementation of gluten is due to a general improvement in dietary protein quality and not to any particular metabolic effect peculiar to lysine. This conclusion is supported by previous work (1, 12-14) showing that various low-quality proteins are better able to enhance tumor development when they are supplemented with the amino acid in greatest deficit.

Clearly the effect of protein in Campbell’s experiments had nothing to do with plant protein versus animal protein and was simply a general effect of complete protein, as he acknowledged in his own papers, at least during the 1980s.  In two papers published in 1997, by contrast, Campbell and colleauges cited the gluten-casein study as showing that plant proteins protect against cancer while animal proteins promote cancer (25, 26).  Thus by the time The China Study was published the fact that complete protein, and not animal protein specifically, promoted cancer in certain contexts was likewise lost down the memory hole.

It’s All About the Mechanism

Before we can make any dietary conclusions from these studies, we need to understand the mechanism.  Is the effect of protein the intrinsic property of any complete protein, or does it depend on other nutrients?  None of us consume 20 percent of our diet as casein, wheat gluten, freeze-dried cod protein, or any of the other proteins Campbell tested.  Many of us drink milk, or eat meat, fish, bread, legumes, fruits, vegetables, and other whole foods.  What are the effects of these whole foods on cancer?  This is a question that cannot be answered from Campbell’s rat studies.  Understanding how protein exerts its effects, however, would help us form a reasonable hypothesis.

The protective effect of protein during the initiation period is easy to explain, though experimental evidence would be needed to support this explanation.  As I pointed out in my recent blog post, “The Biochemical Magic of Raw Milk and Other Raw Foods: Glutathione,” adequate protein is necessary to synthesize glutathione, the master antioxidant and detoxifier of the cell.  In humans, the requirement appears to be about one gram of protein per kilogram of body weight, which is about 70 grams per day for someone who weighs 150 pounds.  In rats, the level of dietary protein that maximizes glutathione and its related antioxidant and detoxifying enzymes is somewhere between 7.5% and 15% methionine-supplemented casein.  According to a chapter (27) of the recent textbook Adverse Drug Reactions, published earlier this year as part of the Handbook of Experimental Pharmacology series, aflatoxin is primarily detoxified by glutathione.

The most obvious reason that protein might promote the growth of cancer in certain contexts is by providing sufficient amino acids to synthesize new proteins needed by rapidly dividing cells.  However, protein is also known to interact with a number of other dietary factors, and the most obvious explanation may not be the correct one.  High protein intakes mobilize vitamin A from the liver and increase its utilization and excretion (28).  Some evidence also suggests that high-protein diets increase the requirement for vitamin B6 (29).  When dietary protein comes from meat, especially from liver, it provides these and many other nutrients.  Do these whole foods, providing protein together with its associated nutrients, promote cancer or protect against cancer?

Campbell’s research is in fact fascinating, but without answering these deeper questions, it is difficult to interpret.  One thing is certain: low-protein diets depressed normal growth, increased the suceptibility to many toxins, killed toxin-exposed animals earlier, induced fatty liver, and increased the development of pre-cancerous lesions when fed during the initiation period of chemical carcinogenesis.  The loss of these facts down the memory hole may make Dr. Campbell’s arguments much simpler, but it does nothing to promote truth or help us understand the true significance of his work, which, once expanded on with further research, may prove incredibly profound.

Read more about the author, Chris Masterjohn, PhD, here.


1.  Madhavan TV, Gopalan C.  Effect of Dietary Protein on Aflatoxin Liver Injury in Weanling Rats.  Arch Pathol. 1965;80:123-6.

2.  Madhavan TV, Gopalan C.  The Effect of Dietary Protein on Carcinogenesis of Aflaxtoxin.  Arch Pathol. 1968;85(2)133-7.

3.  Mgbodile MUK, Campbell TC.  Effect of Protein Deprivation of Male Weanling Rats on the Kinetics of Hepatic Microsomal Enzyme Activity.  J Nutr. 1972; 53-60.

4.  Mainigi KD, Campbell TC.  Subcellular Distribution and Covalent Binding of Aflatoxins as Functions of Dietary Manipulation. J Toxicol Environ Health. 1980;6(3):659-71.

5.  Nerurkar LS, Hayes JR, Campbell TC. The Reconstitution of Hepatic Microsomal Mixed Function Oxidase Activity with Fractions Derived from Weanling Rats Fed Different Levels of Protein. J Nutr. 108:678-86.

6. Campbell TC, Hayes JR.  The effect of quantity and quality of dietary protein on drug metabolism. Fed Proc. 1976;35(13):2470-4.

7.  Schulsinger DA, Root MM, Campbell TC.  Effect of Dietary Protein Quality on Development of Aflatoxin B1-Induced Hepatic Preneoplastic Lesions.  J Natl Cancer Inst. 1989;81:1241-5.

8.  Appleton BS, Campbell TC.  Inhibition of Aflatoxin-Initiated Preneoplastic Liver Lesions by Low Dietary Protein.  Nutr Cancer. 1982;3(4):200-6.

9.  Prince LO, Campbell TC.   Effects of Sex Difference and Dietary Protein Level on the Binding of Aflatoxin B1 to Rat Liver Chromatin Proteins in Vivo.  Cancer Res. 1982;42(12):5053-9.

10.  Appleton BS, Campbell TC.  Dietary Protein Intervention During the Postdosing Phase of Aflatoxin B1-Induced Hepatic Preneoplastic Lesion Development.  J Natl Cancer Inst. 1983;70(3):547-9.

11.  Appleton BS, Campbell TC.  Effect of High and Low Dietary Protein on the Dosing and Postdosing Periods of Aflatoxin B1-induced Hepatic Preneoplastic Lesion Development in the Rat.  Cancer Res. 1983;43(5):2150-4.

12.  Dunaif GE, Campbell TC.  Relative Contribution of Dietary Protein Level and Aflatoxin B1 Dose in Generation of Presumptive Preneoplastic Foci in Rat Liver. J Natl Cancer Inst. 1983;78(2):365-9

13.  Dunaif GE, Campbell TC.  Dietary Protein Level and Aflatoxin B1-Induced Preneoplastic Hepatic Lesions in the Rat.  J Nutr. 1987;117(7):1298-1302.

15.  O’connor TP, Roebuck BD, Peterson F, Campbell TC.  Effect of Dietary Intake of Fish Oil and Fish Protein on the Development of L-Azaserine-Induced Preneoplastic Lesions in the Rat Pancreas.  J Natl Cancer Inst. 1985;75(5):959-62.

16.  Horio F, Youngman LD, Bell RC, Campbell TC.  Thermogenesis, Low-Protein Diets, and Decreased Development of AFB1-Induced Preneoplastic Foci in Rat Liver.  Nutr Cancer. 1991;16(1):31-41.

17.  Burr GO, Burr MM.  A New Deficiency Disease Produced by the Rigid Exclusion of Fat from the Diet.  J Biol Chem. 1929;82(2):345-67.

18.  Torres N, Bautista CJ, Tovar AR, Ordaz G, Rodriguez-Cruz M, Ortiz V, et al.   Protein restriction during pregnancy affects maternal liver lipid metabolism and fetal brain lipid composition in the rat.  Am J Physiol Endocrinol Metab. 2010; 298(2):E270-7.

19.  Denucci SM, Tong M, Longato L, Lawton M, Setshedi M, Carlson RI, et al.  Rat strain differences in susceptibility to alcohol-induced chronic liver injury and hepatic insulin resistance.  Gastroenterol Res Pract. 2010; Epub August 16, 2010.

20.  Youngman LD, Campbell TC.  High Protein Intake Promotes the Growth of Hepatic Preneoplastic Foci in Fischer #344 Rats: Evidence that Early Remodeled Foci Retain the Potential for Future Growth. J Nutr. 1991;121(9):1454-61.

21.  Youngman LD, Campbell TC.  Inhibition of aflatoxin B1-induced gamma-glutamyltranspeptidase positive (GGT+) hepatic preneoplastic foci and tumors by low protein diets: evidence that altered GGT+ foci indicate neoplastic potential.  Carcinogenesis. 1992;13(9):1607-13.

22.  Youngman LD, Campbell TC.  The Sustained Development of Preneoplastic Lesions Depends on High Protein Intake.  Nutr Cancer. 1992;18:131-42.

23.  Youngman LD, Campbell TC.  Attenuation of preneoplastic lesion development by dietary protein intervention: apparent persistence and regression.  Cancer Letters. 1992;66:165-74.

24.  Bell RC, Golemboski KA, Dietert RR, and Campbell TC.  Long-Term Intake of a Low-Casein Diet is Associated With Higher Relative NK Cell Cytotoxic Activity in F344 Rats.  Nutr Cancer. 1994;22:151-62.

25.  Cheng Z, Hug J, King J, Jay G, Campbell TC.  Inhibition of Hepatocellular Carcinoma Development in Hepatitis B Virus Transfected Mice by Low Dietary Casein.  Hepatology. 1997;26:1351-4.

26.  Hu JF, Cheng Z, Chisari FV, Vu TH, Hoffman AR, Campbell TC.  Rperession of hepatitis B virus (HBV) transgene and HBV-induced liver injury by low protein diet.  Oncogene. 1997;15:2795-801.

27.  Srivastava A, Maggs JL, Antoine DJ, Williams DP, Smith DA, Park BK.  Role of reactive metabolites in drug-induced hepatotoxicity.  Handb Exp Pharmacol. 2010;(196):165-94.

28.  Furusho T, Wada M, Yasuhara T, Kataoka E, Kato S, Masushige S.  Tissue specific-distribution and metabolism of vitamin A are affected by dietary protein levels in rats.  Int J Vitam Nutr Res. 1998;68(5):287-92.

29.  Dirige OV, Beaton JR.  Factors affecting vitamin B6 requirement in the rat as determined by erythrocyte transaminase activity.  J Nutr. 1969 Jan;97(1):109-16.

28 Responses to The Curious Case of Campbell’s Rats — Does Protein Deficiency Prevent Cancer?

  1. Stephan says:

    Nice one Chris. The problem is usually in the interpretation, not the data!

  2. ‘Scuse me while I pick my jaw up off the floor.

    I think most China Study skeptics (myself included) have assumed Campbell’s descriptions of his casein research were accurate, focusing instead on the errors of extrapolating it to all forms of animal protein. But it looks like *this* is the real smoking gun. “A deficiency of dietary protein was shown to increase the toxicity of aflatoxin for rats” — why was this not taken into consideration during any of his experiments?! Incredible!

    The final graph (from the “Effect of High and Low Dietary Protein…” study) is particularly fascinating. Lost in the memory hole indeed!

    Stunning job researching this — I think there’s little left of “The China Study” at this point besides a few glowing embers and a smattering of ashes.

  3. No, I think it looks more and more like being neither about interpretation nor data. It seems that the “F” word (the 5-letter one…) is more likely. Unfortunately.
    Thank you for spending time to produce such a thorough analysis!

  4. William Cane says:

    Thank you for this excellent research.

  5. Rachel says:

    Deep down, I have always known there were huge contradictions to the China Study and its implications for human health. Thank you sincerely for such a thorough report!

  6. Emily says:

    Curious indeed. I wonder though at Campbell’s recommendation for 10% protein when the studies you mentioned most often compared 5% to 20% casein diets. Maybe to avoid the malnutrition? The whole thing is just so bizarre.

  7. Great work as usual Chris! More evidence that careful and unbiased interpretation is necessary for true science. It seems today that most health study results in the name of “science” pander to money with expectation strings attached. Scientists are afraid to state the truth because they know they will lose their funding if it doesn’t fit with preconceived expectations. It’s very difficult to do good research without funding.

    I’d like to see animal feeding studies using foods closer to what people might eat rather than fractionated and highly processed “food”. For instance, one simple study would be to feed one group of rats nothing but raw milk from pastured heritage cows, while feeding another group nothing but commercial pasteurized/homogenized 2% milk from a local store. It might be very telling and relatively inexpensive.

  8. DR.BG says:

    I’ve loathed and suspected Campbell the second I heard about the so-called China Study because I’ve grown up with countless diabetic, ill and cancerous Chinese people… (I’m Chinese and my parents are physicians who treat a lot of semi-vegetarian Chinese)

    Casein sucks as a protein source so I am surprised the conclusions in the post are so blatant and discerning.

    Keep up the strong work, Master Chris!

  9. Fiber1 says:

    If your “data” is so valid and thorough. .Then why is the US health system in such termoil? Why does the US have such high rates of disease and what are your causes? Your “data” aparently disregarding The China Study is oh so complex. Well where are your answers to the complex questions Dr. Campbell gave just advice and scientific evidence towards?

  10. Thanks Stephan!

    I agree — in most “flawed” studies the flaw resides in the inferences made from the design rather than the design itelf. In general, Campbell’s animal research was very rigorous. One exception I think was the use in one study of 5ppm aflatoxin for high-casein and 2.5ppm aflatoxin for low-casein rather than just using 2.5 ppm for both groups so the low-casein rats wouldn’t die. All the other studies were well designed. Conclusions in [i]The China Study,[/i] a different matter.


  11. Hey Denise–

    Thanks! I hope your jaw has recovered by now. 🙂

    Campbell sometimes took the increased toxicity of aflatoxin into account in his papers, but it’s amazing that 1) this disappeared from his papers in the 1990s and consequently in [i]The China Study[/i] and that 2) he [i]never[/i] investigated the protective effect of the 20% casein diet during the initiation period.

    You burned chapter four down and even set the spark to chapter three with what you found on lysine, but I guess I gave it a good huff and puff of well oxygenated breath and a lil’ fuel.


  12. Stan,

    You’re welcome!

    Campbell does seem to have reported the genuine data as is. Had he not, I would not have been able to write this post. Certainly, however, much cherry-picking was involved in designing the course of the research program and in writing [i]The China Study[/i].


  13. Dear William and Rachel,

    You’re welcome, and thanks for your comments!


  14. Hi Emily,

    I don’t think Campbell clearly advocates 10% protein — he says this is a common figure but the true requirement is less — but to the extent he favors this amount as ok it is probably because in his most extensive dosing study he found that changes in casein intake above 20% or below 10% had little to no effect on cancer promotion, but that changes between 10 and 20% had a dose-responsive cancer-promoting effect.

    Unfortunately, however, he never did such an extensive dose-response study to find where the protective effect of a high-casein diet begins and ends when fed during the initiation period. Could 10% provide maximal protection during both periods? Maybe. But maybe that value would be 15%. A single study could have clarified it up, but Campbell chose to ignore the protective effect during initiation and forever condemn high-protein diets for their effects during the promotion period.


  15. Hey Bryan,

    Thanks! I agree on the whole foods approach, although I think the “reductionist” approach is also important. I would like to see these studies followed up with addition of specific nutrients to study protein-nutrient interactions and then with whole food sources of protein such as meat. For specific inferences about milk, I would like to see the same thing for casein-whey interactions and interactions between casein and common milk lipids and micronutrients, and then a final followup using whole milk, raw and pasteurized, and with other variations in species and strain as well (rat versus cow, A1 versus A2 casein, etc).


  16. Dr. BG,

    Thanks! I will.

    I agree there are problems with casein, and heretofore I had assumed, like Denise writes below, that Campbell’s main interpretive flaws were extrapolating from casein to other animal proteins. After reading through these studies, however, it is evident that these affects have been shown for a broad spectrum of animal and plant proteins and that it has nothing to do with the specific protein source and everything to do with the provision of a complete protein.

    So what is truly bizarre beyond belief is how Campbell can say that plant protein is protective when he clearly showed this dichotomy to be non-existent.


  17. Hi Fiber1,

    Thanks for commenting! It’s good to see people from a variety of viewpoints reading each other’s material and kindly responding.

    I agree this issue is complex but this data is not mine. All of the data presented are from Dr. Campbell’s papers and indeed huge portions of my blog post are quotes from Dr. Campbell and his co-authors themselves, set off as block quotes. If you follow the linked references, you can find the pubmed indexing for these papers. Some of them are available free full text and any others you can get through interlibrary loan. If you check the data, I believe you will confirm that I reported it accurately.

    Of course Campbell makes many absolutely excellent points about how we spend the most on health care and are the least healthy. There are many complex reasons for this and I have written a lot on the issue of health on this site as well as on In sum, however, the American diet is attrocious because it is rich in modern, refined foods. The animal foods are factory farmed, the grains are refined, genetically engineered and not prepared properly, the fruits and vegetables are grown in poor soil, and on top of these things the food selection minimizes fruits and vegetables, leaves out organ meats entirely, and emphasizes refined grains. There are other important influences on health such as physical activity, mental, emotional, and spiritual well being, and the American lifestyle often does not do to great in these areas either.

    Again, thanks for writing,

  18. Hi Chris,

    Re: [i]Campbell does seem to have reported the genuine data as is. … much cherry-picking was involved in designing the course of the research program and in writing The China Study.[/i]

    That’s right but I think in science “cherry-picking” is undistinguishable from “fraud” and mocks the concept of the scientific method. Falsification by omission KNOWINGLY is still a falsification. It seems to be the same pattern of falsification with his “China Study” book as it is with his rats papers!

    [rant] Did you notice how almost everywhere, did real scientists get pushed out of the mainstream academic institutions and replaced with “Cornell” type of characters? I have seen the same thing going since the 1980-ties in other sciences. Very sad, that’s why so much is falling apart in this culture. Regards and keep up your fascinating research. Stan

  19. Hi Stan,

    I think there are a lot of great researchers out there, and a lot of junk. If you have any interesting historical/socio-political insights into the changes you’ve observed I’d be interested to hear them.


  20. PhilM says:

    This whole drama around China Study begs the question of whatever happened to practicing science? I can easily understand not being able to repeat his experiments but surely, his paper could have been read and data analyzed independently? Did it really need an “obscure English-literature-major girl blogger” to do the even basic checks on a study of such importance? I mean, come on! Isn’t there a single person who worked on this study that has some integrity and question the conclusions?

    BTW, I admire Denise Minger’s work – girl blogger notwithstanding 🙂

  21. Hi PhilM,

    I think there are scientists who have refuted some of Campbell’s claims — Denise has quoted some of them — but in general scientists respond to papers within the peer-reviewed literature. Responding to a book is a much bigger task, and of course even responding to a chapter of a book like I have done here and like Denise’s work has mostly done is clearly a ginormous task in itself, too broad to address within the scientific literature.


  22. Roline says:

    Chris, thank you for your detailed articles. I am a vegetarian, and recently read about the movie and Dr Campbell’s research. Even as a vegetarian I was absolutely not convinced of his conclusions as described by the website of the movie and found them to be biased. Thank you for your research into the topic.

  23. Janie says:

    I don’t know. I think the china study is flawed and I think the above article are flawed. I think when we try to break everything down it is difficult and I believe it when people say Foods are greater than the sum of their parts. Maybe it isn’t each individual vitamin that is good for us, but how they all interact as a food together… and perhaps differently in each of our bodies.
    As many studies as you find that say one thing, you can find one that says something else.
    Even in science there is nothing as an unbiased opinion. I did also find the book by dr. caldwell esselstyn interesting, too, tho, maybe moreso.
    Anyway…enjoy your food!

  24. Scott says:

    On September 24, 1907, the New York Times published an article entitled “Cancer Increasing among Meat Eaters,” which described a seven-year epidemiological study showing that meat-eaters were at high cancer risk, compared with those choosing other staples. Focusing especially on immigrants who had abandoned traditional, largely planted-based, diets in favor of meatier fare in the U.S., the lead researcher said, “There cannot be the slightest question that the great increase in cancer among the foreign-born over the prevalence of that disease in their native countries is due to the increased consumption of animal foods….

  25. Scott says:

    ‘Animal Products Linked to Cancer
    December 27, 2013
    Consumption of meat and other animal products is strongly linked to several types of cancer, according to an article published in the journal Nutrients. The author analyzed data on 21 different cancers in 157 countries and found that certain factors, especially diet, were associated with risk for developing specific cancers. Specifically, the association between animal product consumption and cancer was as strong as that linking tobacco and cancer. Possible mechanisms for risk include animal products’ promotion of growth and high iron and fat content. The author notes that while detection and treatment are important, animal product consumption has been recognized as a cancer risk for more than a century and needs to be addressed in order to prevent the deadly disease.
    Grant W. A multicountry ecological study of cancer incidence rates in 2008 with respect to various risk-modifying factors. Nutrients. 2014;6:163-189.

  26. William Raed says:

    First read the entire china study book and all other research related to ANIMAL PROTEIN before claiming it to be false. HAVE YOU DOBE ANY RESEARCH YOURSELF? OR GOOGLE IS YOU MAIN TOOL? Before calling a study “OBSCURE” (The indian study about protein) you are showing how bias you are.

    Read some more at :
    ( A milk protein, casein, as a proliferation promoting factor in prostate cancer cells.)

    FOR THE LAZY READER, HERE IS THE RECULT ===>>CONCLUSIONS: The milk protein, casein, promotes the proliferation of prostate cancer.


  27. Michael says:

    Funny – There is not shortage of empirical evidence that meat eaters have a h9igher rate of cancer (among other health issues), yet not one study that showing vegetarian children with stunted grown or high childhood death rates. It is one thing to point out what other researches may have not discussed in their research, but it is another to concoct fiction results of what was overlooked.
    In hindsight, most of your opinions are flat out wrong.

Leave a reply

© 2015 The Weston A. Price Foundation for Wise Traditions in Food, Farming, and the Healing Arts.