The China Study: A decent yet terrible book within the same covers

I finally got around to reading the China Study, a 2006 publication from Dr. Colin Campbell and Dr. Thomas Campbell (from what I read, father and son). I mainly read this book as it is the polar opposite of what I believe, and the only thing I like more than screwing with the norm by being a devil’s advocate myself is to see other people create a shit-storm and sit by watching. Another reason is that I found the notion of low protein so ambiguous yet backwards that I felt I had to read the book to come to any conclusions on the topic and better address others.

Many things surprised me after summation. Firstly, many people who talk about the China Study talk about the book whereas they are met with people who are addressing the Study itself; huge potential for miscommunication. Secondly, for a study that made so many conclusions from correlation in the media the author himself (Colin Campbell) is quite an intelligent individual and repeatedly declares how correlation does not equal causation, and to take the results with a grain of salt.

I guess this is the first Book Review I am doing on my blog; might as well be a controversial one.

Introduction to ‘The China Study’ (Book)

The China Study is a book written by the two authors above talking briefly about the actual China Study and the preliminary research which led to the China Study. It also brings in other research (with over 700 citations, was happy to see that) to back up his point.

The general topic is that the ‘healthiest’ diet (mostly for cancer prevention, with some notion to diseases of affluence like diabetes, obesity and heart disease) is a low protein one with no animal protein sources (5-20%, although the former is stated more often), low fat, and generally high carbohydrate.

I’d like to say good things about the book (as my title alluded to), but all the good things that came from the book stemmed from my own ignorance on the topics at hand and opening research avenues. Basically, I never looked into the link between animal products and early onset of menses in young females or higher rates of breast cancer in older females and I never looked into a diet’s effects on P450 status. The good of the book was that it opened my eyes to research avenues, but this was merely for being opposite of my views.

I’ll go into the ‘bad’ of the book in the most ironic way possible, by stating my various ‘Beefs’ with this low-protein vegetarian plan.


Beef #1: Too much emphasis on epidemiological research; not enough on interventions

At the beginning of the book, he mentions on ‘Indian Study’ that had two groups of rats fed aflatoxin (a toxin produced by mold) and then fed either 5% protein or 20% protein. Over time there was a 0% and 100% death rate in those groups respectively (see the next section for why this happened). Later on he replicated these results in rats, and then again in mice with hepatitis (similar mechanisms of action to aflatoxin).

Then he went on to the China Study, and pretty much stayed there for the duration of the entire book. Every now and then he would cite other researchers and epidemiological research, but mostly he stuck to the China Study.

I didn’t read about a human intervention with a low protein intake at all; which was surprising given the authors faith in this topic and his research background. I would imagine that a researcher of his prestige can get a group of hepatitis patients and work with hospital staff to administer different protein concentrations in the diet, but the topic never came up.

That, and when he was stumped for citations he continually referenced the rat studies he did in the past. Something about talking mostly about epidemiological research, and then jumping back to his own research in diseased rats, leaves a bad impression on me when he recommends such a drastic change to the norm (Re: protein intake).

Later in the book, when talking about heart disease, many human interventions conducted in already sick populations were conducted and noted. These studies do show improvements on a vegetarian, high carbohydrate, whole foods diet (as is routinely mentioned in the book), but its fairly well accepted that those already into heart disease pathology bode well with diets low in fat and cholesterol. Preventative nutrition (for those already healthy) is another story, but the author(s) are satisfied with showing how their diet lowers and is correlated with reduced serum cholesterol (of which its usage as a non-clinical marker for heart disease has been brought into question as of late; questionable validity in extrapolating these interventions onto healthy persons). So yet again, where a human intervention study matters and would be relevant to a good deal of the population reading the book, it is not given and epidemiological research is relied upon.

Aside from those, the rest of the heart disease section, the entire obesity and diabetes section, and the cancer section were all based on correlational research. There was a surprisingly well written write-up in the autoimmune section on the link between Bovine Serum Albumin (BSA) from cow’s milk and inducing Type I diabetes in infants, although it is irrelevant to the diet’s application unto the population as a whole. If anything, that excerpt shows that the author(s) are easily capable of writing amazing and coherent reviews given they have adequate sources to cite; which may work against the rest of the book that lacks such depth.


Beef#2: Biochem was introduced as a necessity, and dropped way too early in the book

For those wondering, there is sort of a legitimate link between protein and cancer. I will try to summarize below in text and the bullet form.

There is a family of proteins in the body called cytochrome P450. These are a large family (60+) of proteins that are given the task of metabolizing unknown compounds that enter the body. Compounds expected in the diet (Thiamin, dietary protein, calcium, etc.) have their own custom-made metabolic pathways, but compounds that shouldn’t really exist unless humans made them, or otherwise are not part of a normal human diet, are subject to the wide-reaching effects of the P450 system.

Due to the lack of specificity of the P450 system, accidents happen. For starters, the enzyme is not a gentle one and it pretty much generates a super radical in order to force a reaction in a compound by pretty much Falcon-Punching it with the oxygen radical. This may generate some free radicals (or in the case of the toxin Paraquat, a shit-ton of free radicals), and sometimes it forces a reaction onto an otherwise stable molecule.

(To note, the reason it forces a reaction is so other enzymes can ‘tag’ the newly formed site with a functional group designed for urinary excretion and thus get said compound out of the body. Compounds that cannot be attacked by P450 include the toxins Dioxin and PCBs, which build-up in the body)

By activating an otherwise stable compound (nitrosamines, heterocyclic amines, etc.) P450s can actually turn inert compounds into carcinogens. Its a necessary and accidental evil really.

One of these compounds is the aflatoxin, probably one of the most potent carcinogens naturally produced (by mold) and one of the only things known to cause liver damage (the other being hepatitis). P450s activate aflatoxin into a toxic metabolic that does wreak havoc on the liver.

The whole reason for all of this? Low (>5% in rats) protein diets seem to drastically suppress P450 formation and thus bioactivation of pre-carcinogens. I would go out to say that it is quite reliably shown in the literature that a very low protein diet can protect against carcinogenesis that is a result of P450 activation.

And that was it. The rest of the book was correlational research. This left a bad taste in my mouth (so to speak) for numerous reasons:

  • What about carcinogens that are legitimately detoxified through the P450 system? Would not lessening their activity then cause those carcinogens to exert more profound effects?
  • Why are you advocating such an unsustainable low protein diet (have you ever eaten >5% protein? I haven’t, because I had to hit calories and my diet was more than just lettuce), when you could easily advocate certain carcinogen avoidance?
  • What replaces the protein exactly? You also recommend a low fat diet and do not advocate caloric restriction. Do you expect people to eat the most non-nitrogenous sources of carbs known to man? Practicality!
I should note that he briefly mentions that casein is the most ‘carcinogenic’ protein (in this case, casein just has the greatest influence on P450 levels; hell, it could be greatly anti-carcinogenic if you look at different carcinogens) and that plant based proteins (grains and soy) did not influence carcinogenesis at all, even when nitrogen-matched with animal proteins.
Beef#3: His notes on Obesity
There is somewhat of a neverending debate as to which of the energy macronutrients (carbohydrate or fat) is better for long term anti-obesity (to clarify, weight loss is a short term goal to reduce body fat stores; anti-obesity is a long term preventative measure aimed at preventing adipose accrual). However, both the low carb and the low fat groups do tend to revere protein. This made me very confused when the author recommended his ‘high carb, whole foods’ diet as the solution for weight loss.
What made me even more confused was when he cited a study (pg.142) where vegetarians had a higher resting metabolic rate during rest. After searching for the study, it turns out that there is negligible difference between the resting metabolic rates (specifically, two measures were non-significant and the third benefited meat eaters) and vegetarians actually had less TEF from the meal. The only other real study the author(s) cited in this section was one showing that rats fed 5% protein were more active (and had greater NEPA, or non-exercise physical expenditure) than did rats consuming 20%.
This entire section, and his entire argument, was a bunch of pages dedicated to pandering to his audience by denouncing some vague ‘diet industry keeping you fat to earn money’ and the advice that ‘diet is a marathon, not a sprint’. Nothing was put forth to even say whether his diet was good for fat loss or control because, relative to a good diet that includes meat, it really isn’t.

Beef#4: Author(s) pretty biased, due to involvement in the China Study themselves and writing style

It made me cringe to see the authors talk about others; they had the tone of bullies really. Stating that Dr.Atkins was (paraphrased from pg. 97) an obese man with heart disease who became one of the richest snake oil salesman in history. The author(s) also seemed to have a tone of superiority in their text, always touting how “They have done tons of legitimate research” and how other nutritional researchers and authors “Haven’t spent the time (we) have conducting true research”(paraphrased). It may just be my opinion (as a naive student with what we call ‘hope’) but I would imagine that a doctor of such affluence could criticize their (authors of other diet books) science and ideas rather than attack their character and body.

For that, my views of the authors diminished. It would be hypocritical for me to judge them now, but if I had to it would not be promising. If the pain has become too much to manage, you should seriously consider seeing an experienced back pain doctor. A doctor should be a teacher and a student, neither of which attack others personally.

Their views on research validity were, weird, to say the least. When they needed to talk about research validity they did so splendidly. Many times in the introduction they admitted and even emphasized the limitations of animal models of disease and epidemiological research. I was happy to see this.

Then they wrote a book recommending lifestyle interventions based on said epidemiological research.

They never once controlled for veggie or grain intake either...


In reading it, it really seems that the author(s) took a tunnel-visioned approach to the topic. Once they found benefits of their touted ‘High carbohydrate, whole foods’ diet they immediately went to find out more effects of this diet without wondering what caused the effects seen. Numerous times they even stated that it was futile and silly for researchers to try and find the factor in the diet that causes an effect because ‘the effects seen in a diet are due to synergy between food nutrients’. Although I have written about this earlier and wholly agree with it, the author(s) uses it in such a manner that I believe it is just a cop-out.

If he attempted to control for variables, perhaps he could have distinguished between ‘carbohydrate’ (the molecule) and plant based compounds like polyphenols (which can act as digestive enzyme inhibitors) and bioflavonoids (which can inhibit aromatase in food bound form, and must be considered in the cancer segments). Fiber was mentioned, but not highly controlled for nor given much credit.

So basically, he never once showed proof that the diet was good because of the exclusion of meat or because of the inclusion of vegetables and phytonutrients. All comparisons were made between his vegetarian/vegan high carbohydrate, whole foods diet against the Standard American Diet, which is shit.

I should also mention that the vast majority of his 700+ citations are from the 60s-90s, and not much post-2000 research was cited at all despite publication in 2006. The studies are also divided by chapter, so his rat study and the China Study itself, since mentioned in multiple chapters, have multiple citations in that total. It was impressive at the onset, but not so much when I looked into it.

In summation

The China Study (book) is one that relies much too heavily in epidemiological research and correlation. It then adds to the validity of said epidemiological research with selective intervention studies. The problem lies, however, in the fact that the author(s) only use rat studies they performed themselves or human intervention studies in diseased populations. They then use selective intervention studies and correlational research to recommend drastic changes to the diet, and never once attempt to place causation on components of the diet (low protein, high vegetable, high fruit, high carb, low fat, fat ratios, vegetarian versus animal, etc.). The book touts a high carbohydrate diet and never one espouses how nice grains are (since they really aren’t for the topics the book looked at) and only compared the final ‘High carbohydrate, whole foods diet’ (of which both high veggie and high carb are included) against the Standard American Diet.

This book is one talking about how awesome a high vegetable intake is; not how awesome a low protein and fat intake is. Their argument based on reducing protein was weak and carcinogenic-specific (of which we have no control over) and their argument on reducing fat was wholly based on correlational research. They had no specific reason to tout such a high carbohydrate intake aside from ‘the other two are evil’, and that premise is faulty at best, and completely backwards at worst.

Connect with Facebook

Speak Your Mind