The China Study | Science-Based Medicine (original) (raw)

The China Study
One of our readers asked that we evaluate a book I had not previously heard of: The China Study: Startling Implications for Diet, Weight Loss and Long-Term Health, by nutrition researcher T. Colin Campbell, PhD, with his non-scientist son Thomas M. Campbell II. The China Study was an epidemiologic survey of diet and health conducted in villages throughout China and is touted as “the most comprehensive study of nutrition ever conducted.” The book’s major thesis is that we could prevent or cure most disease (heart disease, cancer, diabetes, autoimmune diseases, bone, kidney, eye and other diseases) by eating a whole foods plant-based diet, drastically reducing our protein intake, and avoiding meat and dairy products entirely.

Opinions of the book

There’s a lot of praise for this book on the Internet. It was named VegNews Book of the Year. PETA loves it (not surprisingly). Heather Mills McCartney calls it inspirational. It was featured on Oprah.com and endorsed by two of her favorite doctors: Mehmet Oz and Dean Ornish. Its author was even interviewed on Coast to Coast AM.

But I also found this critical review which makes some excellent points and accuses the authors of misrepresenting the findings of the study. And this commenter on an Amazon.com forum also charges Campbell with misrepresenting the data from the study and points out numerous flaws in his reasoning.

Problematic references

I didn’t look at the praise or criticism of others until after I read the book, and the following represents my independent impressions. I approached the book as I do any book with scientific references: I read until I come across a statement of fact that strikes me as questionable and then I check the references given for the statement. This immediately got me off on the wrong foot with this book. In the first chapter I found the statement:

Heart disease can be prevented and even reversed by a healthy diet.

The end notes listed two references in support of this assertion. The first reference was not about diet alone, but about a combination of several lifestyle interventions (Dean Ornish’s intensive program). 28 patients were assigned to an experimental group (low-fat vegetarian diet, stopping smoking, stress management training, and moderate exercise) and 20 to a usual-care control group. The experimental group had less artery narrowing.

The second was a study of 22 patients with severe heart disease in a single physician’s practice. Their disease was arrested or reversed over a 5-year period not by diet alone but with a combination of a very low fat diet and cholesterol-lowering drugs. These studies were published in 1990 and 1995 respectively and as far as I know have not been replicated. And they were small preliminary studies of the kind that should be used only to guide further research, not the kind of definitive studies that can be used to guide clinical decisions.

Neither reference supports the claim that diet (by itself) can reverse heart disease and neither of them has anything to do with preventing heart disease – all the subjects in both studies already had the disease.

Reading further, I found that this was not an isolated oversight, but part of a pattern. For instance, Campbell asserts that diet is an effective treatment for melanoma and supports that claim by citing a Tijuana study of the discredited Gerson protocol, which includes coffee enemas and other non-dietary interventions. Patients allegedly cured of cancer by this method were tracked down by a naturopath who found that 5 years later all but one had died of their cancer and the only one still alive was not cancer free.

Actually there are a couple of more credible references listed on PubMed suggesting that a low-fat diet might be beneficial for melanoma. Why did the author not cite those studies but pick a disreputable one?

Sloppy citations like these do not disprove the author’s thesis, but they throw doubt on his scientific rigor and reasoning abilities.

The China Study

The China Study involved 100 adults in each of 65 counties in China. Only those between the ages of 35 and 64 were studied; for mortality rates they eliminated death certificates of those over the age of 64 as “unreliable.” They pooled blood samples from everyone in a village so they would have large enough samples to measure over 109 nutritional, viral, hormonal and other indicators in blood. They also measured 24 urinary factors, mortality rates for more than 48 diseases, 36 food constituents, 36 nutrient and food intakes, 60 diet and lifestyle factors, and 17 geographic and climatic factors. All in all, they studied 367 variables and made 8,000 correlations. I’ll leave it to others to comment on the study design and the statistical analysis. (Addendum: someone has done just that)

The Chinese eat far less animal protein than Americans and far less total protein than even American vegetarians. They eat more calories per kilo of body weight than Americans, apparently without weight gain even among the subset of Chinese who were least physically active (Campbell attributes this to thermogenesis from carbohydrate metabolism as compared to protein metabolism, and he claims that vegetarians feel more energetic and naturally exercise more, using up extra calories). They found that Chinese cholesterol levels are far lower than Western levels and decline as the amount of protein in the diet declines. They found a strong dose-effect relationship between the amount of animal protein in the diet and the rates of many diseases like heart disease and cancer.

Conflicting data

I found a number of studies in PubMed that reached very different conclusions. I’ll quote from one typical example and provide links to a couple of others:

Vegetarians form a non-homogenous group consisting of semivegetarians (plant food, dairy products, eggs and fish), lacto-ovo vegetarians (plant food, dairy products, eggs) and vegans (plant food only). According to pure vegetarian ideologists, people consuming vegetarian diet have better health and live longer than nonvegetarians, because persons consuming milk, dairy products, meat, eggs and fish are at health risk. In fact the most healthy people in Europe are inhabitants of Iceland, Switzerland and Scandinavia, consuming great amounts of food of animal origin. Meta-analysis of several prospective studies showed no significant differences in the mortality caused by colorectal, stomach, lung, prostate or breast cancers and stroke between vegetarians and “health-conscious” nonvegetarians. In vegetarians, a decrease of ischemic heart disease mortality was observed probably due to lower total serum cholesterol levels, lower prevalence of obesity and higher consumption of antioxidants. Very probably, an ample consumption of fruits and vegetables and not the exclusion of meat make vegetarians healthful.

And there’s this one and this one.

Campbell criticizes all Western studies of low-fat and low-protein diets because the Western versions of those diets still have far more fat and protein than the average diet in China. The Nurse’s Health Study found no connection between breast cancer and the amount of fat in the diet, but Campbell points out that it really only compared carnivorous nurses to slightly less carnivorous nurses. The “low-fat” group was still eating a very high-fat diet by Chinese standards. This is a very valid criticism, and it also applies to the relatively ineffective Western efforts to lower cholesterol and reduce the risk of heart disease with diet.

Observations from other countries tend to contradict the correlations found in China. The African Maasai eat a diet high in animal protein (meat, milk and blood from their cows) – yet they have low blood cholesterol levels and low rates of heart disease. Among the Eskimos (who ate an animal-based, very high protein, high fat diet) heart disease was practically unknown.

Campbell doesn’t attempt to explain a glaring exception to his data: stomach cancer rates are higher in China than elsewhere in the world – he doesn’t even mention that fact.

He cites all kinds of research to support his hypothesis that animal protein is bad. Toxins like aflatoxin and nitrosamine cause rats and mice to develop cancers, but carcinogenesis is prevented by feeding them a low protein diet. Casein (one specific animal protein in milk) has been linked to some human diseases. Links between fat, animal protein, vitamin D and other nutrients confuse the issue. The incidence of many diseases varies with latitude – is it the difference in sun exposure, the blood levels of vitamin D, the fact that people at higher latitudes eat more fat, that they eat more meat?

Diet recommendations

He marshals a lot of evidence, but is it sufficient to support his recommendation that everyone give up animal protein entirely, including dairy products? I don’t think so. There are legitimate concerns that such a diet may not be without risks. Even Campbell recognizes that strict vegetarians are likely to need vitamin B12 supplementation. If cow’s milk is prohibited for growing children and osteoporotic adults, they will likely need a supplemental source of calcium and vitamin D. Without careful nutrition guidance, children deprived of milk might end up malnourished. Breast milk is animal protein – should we avoid breast-feeding too?

He criticizes conventional recommendations for a diet with 45-65% of calories from carbohydrates, 20-35% from fat and 10-35% from protein, showing how the following menu satisfies those requirements:

Breakfast
1 cup Froot Loops
1 cup skim milk
1 package M&M milk chocolate candies
Fiber and vitamin supplements

Lunch
Grilled cheddar cheeseburger

Dinner
3 slices pepperoni pizza
16 oz soda
1 serving Archway sugar cookies.

But that’s a bit of a straw man argument. In reality, most current nutritional advice makes very much the same recommendations Campbell does except for his strict prohibition of animal protein. For instance, for cancer prevention the American Cancer Society recommends a diet high in fruits, vegetables, whole grains, and legumes and low in red meat and alcohol, along with regular exercise and weight control.

Except for proponents of the low-carb diet fad for weight loss, almost everything I have read recently in the way of diet advice is consistent with the ACS recommendations. More veggies, less red meat, fewer calories.

Conclusion: Health is more than just diet

It would be wonderful if we could prevent cancer and all those other diseases by avoiding animal protein. It would have the extra added benefit to the environment of increasing the productivity of agricultural land and reducing the greenhouse effects of gassy cows. I look forward to future well-designed studies investigating the effects of very low protein and animal-protein-free diets. Meanwhile, The China Study makes a good case, but the case isn’t quite good enough.