Critique of The China Study – Part 1



The China Study by Dr. T. Colin Campbell examines the health impacts of plant based protein and animal based protein. He also emphasizes the point that the findings of his research are groundbreaking and go against the dogma of mainstream nutrition knowledge. In the beginning of the book, he states that “the American people need to know the truth”, addressing the fact that much of the American population has been deluded by the barrage of pseudoscientific fad diets and studies on the impact of specific chemicals. He claims that if Americans focused more on the more general parts of their diet – namely eating more plant-based foods and less animal-based foods – that they would live much healthier lives devoid of the diseases of affluenza like cancer and heart disease. These dietary recommendations are the core messages that he makes and supports with various pieces of evidence in the entirety of Part 1.

Although I agree with Campbell that most people need to take a bigger picture view of their diet, as they ‘’miss the forest for the trees’, I believe that Campbell makes the same mistake when drawing support for his conclusions against animal protein. The general body of the evidence he provides is compelling and the studies that he presents are ‘’good science’’ when viewed alone, but Campbell tends to cherry pick his evidence while ignoring evidence contradictory to his claims. He also puts too much weight on correlations that he finds, blending them with other weak correlations and specific, limited animal studies to draw claims of causation. This makes it seem like the entire body of scientific evidence strongly supports his assertions of the maladies caused by animal protein when it does not.

Although not central to his main argument, Campbell cites an increasing cancer death rate of the overall general population as a sign of little progress in the War on Cancer. This statistic is misleading because as healthcare technology has progressed, so has our ability to prevent other causes of death occurring earlier in life, thus increasing the average lifespan and therefore the amount of people that will receive cancer because risk of cancer increases with age. Campbell also neglects to mention the constantly declining rates of cancer mortality with some age groups experiencing a decline in mortality rate of more than 20% per decade [1]. This is no doubt due to improvements in medical technology and cancer treatment, like the innovations in new pharmaceuticals and biologics.

More relevantly, Campbell later gives a list of several world class athletes that claim that eating a low-fat plant based diet gives them an edge in performance. However, there are hundreds of thousands of people that claim that a high-fat animal based diet gives them an edge in performance, namely those following the paleo movement (which I do not endorse). Empirically speaking, it is well established in the scientific literature that eating a moderate amount of fats, in order to maintain testosterone levels, and a relatively high amount of protein, in order to increase levels of lean body mass is optimal for athletic performance [2]. Campbell also backs his claim of a low-protein diet improving physical wellbeing with a rat study in which rats fed a high-protein diet exercised more and with less fatigue. In addition to the fact that this study was limited to rats when there are multitudes of human studies examining the effect of macronutrient manipulation on athletic performance, the results can most likely be explained through the decrease in carbohydrates that the high-protein rats received, not an increase in protein.

Before Campbell gets into the bulk of his research in which he provides as evidence, he tells the story of how he was inspired by an obscure study out of India which showed that 100% of rats exposed to aflatoxin while on a high protein (20% casein) diet got cancer, while 0% of rats exposed to the same levels of aflatoxin while on a low protein diet (5% casein). While the study does show interesting results, Campbell neglects to mention the limitations of the study. Aflatoxin was only administered for the first 6 months of the 2 year study because half the rats in the low protein group died of hepatotoxicity after 6 months while all the rats in the high protein group were still alive [3]. Therefore, it seems like the reason why the low protein group received less cancer was because their liver cells were functioning sub optimally and not replicating as often. The trade-off between liver function and cancer seen in this study thus does not provide support for the benefits of a low-protein diet. Consequently, further studies should have studied the difference between low level protein diets that did not impair liver function (>5%), and high protein diets.

Campbell’s first studies on the effect of animal protein on cancer in involved rat experiments where he tried to elucidate the mechanisms of action in which he believed animal protein caused cancer by investigating the initiation stage. Through his studies he found many mechanisms by which a high protein diet could facilitate the initiation phase of cancer such as ‘’ less aflatoxin entered the cells, cells multiplied more slowly, and less aflatoxin-DNA adducts were formed.’’ Though these lines of evidence draw strong support for a high protein diet facilitating the cancer initiation phase, Campbell seems to ‘’miss the forest for the trees’’, as his own study showed that ‘’In fact, the livers from these latter 20-5 animals resembled these of control (no AFB) animals.’’ This study showed that animals fed a high protein diet (20% casein) had a cancer-protecting effect during the initiation phase, although the high protein diet spurred cancer growth after the dosing period, the promotion period [4]. He then investigated the effects of cancer on foci response, an indicator of cancer development. He found that there was only a dose-response to aflatoxin in the case of a high protein diet, probably because the low protein group had impaired liver function due to only receiving 5% casein.

Campbell then claims that while a high animal protein diet promotes cancer, a high plant protein diet does not. This is backed by his experiment in which he compared foci response between rats that were fed 5% casein, 20% wheat protein, and 20% soy protein. However, Campbell fails to mention that wheat is not a complete protein and is missing the amino acid lysine. When this amino acid is supplemented in rats eating a high wheat protein diet, foci response is comparable to rats eating a high casein diet [5]. Soy protein on the other hand seems to have a protective effect against cancer and reduces the incidence of mammary tumors in rats in the presence of a carcinogen [6]. In fact, the other protein component of milk, whey, is twice as effective as soy in the protection against cancer [7]. It seems like these protective qualities of proteins are not confined to only plant protein, but rather more specific groups of protein. It also begs the question of whether proteins in general promote cancer when consumed in an adequate amount, with some groups of proteins being protective. Campbell’s rat studies certainly support the fact that you would not want to feed rats a high casein diet if you wanted to protect them from cancer, but it would be irresponsible to stretch the claims any further. Perhaps future studies could more closely mimic that of humans using a more varied protein diet. Assuming a person consumes 3 cups of milk worth of food per day, they would only be consuming 19.5g of casein, or 3% of the average 2648 calories (a greatly underestimated number due to the use of self-reported intake) an American male eats per day, a far cry from the 20% used in the rat studies [8].

In the next chapter of the book, Campbell uses for the epidemiological evidence of the China study in order to promote plant protein and demonize animal protein. This study examined hundreds of variables from thousands of people across different counties in rural China. Campbell found many significantly significant associations between different variables which he used to support his hypothesis, but these associations are weak in supporting the causal relationships that he suggests due to the ignorance towards many confounding variables. The most egregious mistake was that Campbell did not account for differences in caloric balance among sample groups. In most cases, Campbell did not associate animal protein consumption directly with cardiovascular disease and cancer, but that of cholesterol levels, which are associated with animal protein consumption. As it turns out, animal protein was not correlated with cancer or cardiovascular disease, based on the data in his study [9], but a sustained caloric surplus is correlated with both western diseases, including cancer and cardiovascular disease [10]. Though Campbell mentioned the fact that the Chinese population had greatly higher calorie consumption, he attributed this to the fact that they had higher activity levels and carbohydrate consumption. He theorized that the high carbohydrate and low protein consumption led to more calories being burned off as body heat which is why the Chinese had a much lower level of obesity. This is plain wrong. High protein diets have consistently been superior for improving body composition and weight loss than low protein diets [2].  The discrepancy can be explained by the fact that the Chinese not only have higher activity levels, but higher resting energy expenditure due to having better body composition because of the increased activity (calorie consumption was standardized by weight). Also, the caloric intake data for Americans is probably highly skewed to a lower level because it is measured through self-reported intake, which those who are obese tend to underestimate – a large proportion of Americans [11]. Given this information, it is likely that Americans have a higher rate of diseases of ‘’affluenza’’ because of a sustained caloric surplus, not because of higher consumption of animal protein. As increased cholesterol is correlated with a caloric surpluses, it is possible that the increased incidence of western disease in counties with high cholesterol was caused by overeating, not overconsumption of animal protein. As for his claims regarding the protective effects of increased vegetable intake, according to a more detailed statistical analysis, this seems to be true [9]. However, it would be interesting to see how much of the effect still exists when controlling for calorie balance as an increase in vegetable consumption also causes a decrease in total calories consumed as vegetables are not very calorie dense.

It seems that many of the claims that Campbell makes are unfounded due to the misinterpretation of the data that he presents. He stretches the evidence he provides far beyond their foundations. The limitations of the rat studies prevent the knowledge we learn from them from being applied to a more practical setting, and the associations he draws from the China Study are weakened by the obliviousness to many confounding variables. Campbell also tends to cherry pick data that support his assertions while ignoring those that conflict with his views. For example, Campbell failed to mention that wheat flour consumption correlated more strongly with cancer and cardiovascular disease than did increased consumption of animal protein [11]. It would be interesting for further studies to investigate the limitations that have been discussed in this paper. Animal studies should use diets more aligned with the context of the average human diet because food compounds exist in a complex of many other bioactive chemicals, not just in isolation. As many studies have shown increased red meat consumption to be associated with increased cancer incidence, it would also be wise to examine this connection while controlling for confounding variables such as adequate vegetable intake and caloric balance. Although Part 1 of the China Study does not answer many questions, it brings to light many interesting ones.

Contact me @


  1. Kort, Eric J.; Nigel Paneth; George F. Vande Woude (2009-08-15). “The Decline in U.S. Cancer Mortality in People Born since 1925”. Cancer Research 69 (16): 6500–5.
  2. Kreider, R. B., Wilborn, C. D., Taylor, L., Campbell, B., Almada, A. L., Collins, R., … & Antonio, J. (2010). ISSN exercise & sport nutrition review: research & recommendations. J Int Soc Sports Nutr, 7(7), 2-43.
  3. Madhavan, T. V., & Gopalan, C. (1968). The effect of dietary protein on carcinogenesis of aflatoxin. Archives of pathology, 85(2), 133-137.
  4. Appleton, B. S., & Campbell, T. C. (1983). 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 research, 43(5), 2150-2154.
  5. Schulsinger, D. A., Root, M. M., & Campbell, T. C. (1989). Effect of dietary protein quality on development of aflatoxin B1-induced hepatic preneoplastic lesions. Journal of the National Cancer Institute, 81(16), 1241-1245.
  6. Hakkak, R., Korourian, S., Shelnutt, S. R., Lensing, S., Ronis, M. J., & Badger, T. M. (2000). Diets containing whey proteins or soy protein isolate protect against 7, 12-dimethylbenz (a) anthracene-induced mammary tumors in female rats. Cancer Epidemiology Biomarkers & Prevention, 9(1), 113-117.
  7. Bounous, G., Batist, G., & Gold, P. (1991). Whey proteins in cancer prevention. Cancer letters, 57(2), 91-94.
  8. Smith, L. P., Ng, S. W., & Popkin, B. M. (2013). Trends in US home food preparation and consumption: analysis of national nutrition surveys and time use studies from 1965–1966 to 2007–2008. Nutr J, 12(1), 45.
  10. Unger, R. H., & Scherer, P. E. (2010). Gluttony, sloth and the metabolic syndrome: a roadmap to lipotoxicity. Trends in Endocrinology & Metabolism, 21(6), 345-352.
  11. Lichtman, S. W., Pisarska, K., Berman, E. R., Pestone, M., Dowling, H., Offenbacher, E., … & Heymsfield, S. B. (1992). Discrepancy between self-reported and actual caloric intake and exercise in obese subjects. New England Journal of Medicine, 327(27), 1893-1898.

You may also like...

Leave a Reply

Read more:
Study says flexibility is mental, not physical. What?

I've seen circus acts, like the video below, where some contortionist demonstrates flexibility to the point that its probably more...

Grapefruit & Medication – A Deadly Combination?

  The number of prescription drugs that can interact with grapefruit is increasing. Grapefruit contains furanocoumrarins which are capable of...

Trans Fats — are we be wrong about them?

Trans fats have been so vilified that nutritional labels list how many grams of trans fats food contains. Whatever did...

(Asap) Science explains how to engineer a zombie virus

The key is to engineer a virus similar to rabies that attacks the olfactory nerve, inducing "zombified" behavior leaving other...