A Closer Look at the China Study: Dairy and Disease

20 06 2010

Mongolian yaks: A source of Chinese dairy.

I’ll admit it: Out of all the variables in the China Project, dairy is the one I’ve been most eager to analyze. Not because I’m a dairy lover myself (I haven’t touched it in years) or because I’m secretly a billionaire milk tycoon with my own thousand-acre Holstein farm (au contraire; I’m strangely phobic of cows). In his book, T. Colin Campbell makes such a compelling case about casein (a milk protein) as a cancer-promoting agent that I’m left wondering: Does the China Study data shows an equally convincing link between dairy and disease?

After all, the counties studied in the China Project weren’t eating the hormone-laden, antibiotic-stuffed, factory-farmed dairy we find in most stores. Their dairy was from pastured animals—typically sheep, goats, or yaks along with cattle—raised on natural diets in rural areas. As best I can deduce, milk products were neither pasteurized nor homogenized. This means that any connections we find between dairy and mortality variables are probably from dairy itself—not the nastiness that accompanies the dairy Westerners are more familiar with. This could be one of our best opportunities for studying dairy consumption in its raw, natural state. Yeehaw!

(Note: If this is your first visit to my site and you’re on a quest for China Study information, you may want to start with the earlier posts in this series:)

Dairy consumption in rural China

First, the bad news—from a scientific standpoint, at least. Only three out of the 65 counties in the China Project consumed any noteworthy amount of dairy at all. The rest were completely dairy-free or consumed dairy only a handful of times per year. That means our sample size of hardcore dairy eaters is tiny, and drawing any conclusions about dairy consumption is trickier than if we had lots of dairy-eating regions to examine.

But there’s some good news, too. Two of the counties that did eat dairy ate a lot of it—856.5 grams per day for Tuoli (just shy of 2 pounds!) and 147 grams per day for Xianghuang qi (about a third of a pound). In terms of diet, both of these places are proverbial black sheep in China, making them quite interesting as case studies. And because these regions consumed so many milk products in contrast to everyone else, any correlations genuinely tied to dairy should be pretty dramatic.

Baoqing county, the third highest consumer of dairy, ate about 19.1 grams of milk products per day. Not a whole lot, for sure, but we should keep an eye on this county as well when looking for links with disease.

A tad more info on these places for anyone who’s curious:

Tuoli county. Located at the tippy-top of northwest China in the Uyghur region—a straggler county on the map, much farther west than any other area studied in the China Project. The Tuoli get over half of their daily calories from dairy products (holy cow, literally; this is even more than Americans eat) and are quite fond of yogurt. They consume very few vegetables, fruits, or nuts, but do eat a significant amount of wheat flour.

Xianghuang qi county. Located in inner Mongolia. Local cuisine includes mutton and dairy products—especially yogurt, fermented milk, Mongolian milk tea, butter, and cheese—as well as oats and buckwheat. Vegetables haven’t been a big part of their menu until quite recently. You can read more about the diet of this region at China.com.

Baoqing county. Located at the northernmost and easternmost spot out of all the counties studied. These folks are one of our highest egg eaters, eat moderate amounts of meat, and consume wheat as their primary grain.

Dairy consumption in rural China

In China, dairy intake ranged from zero grams per day to 856.5 grams per day, as mentioned earlier. Since so few counties consumed milk products, our correlations are strongly swayed by the habits of the two dairy-eating regions. Those include snuff use (correlation of +98), meat eating (+52) and wheat intake (+35). Dairy-eating people also tended to be heavier and taller than other Chinese citizens (+34 and +23, respectively). Although the additional fat and protein from dairy foods could be responsible for a bigger body size, inhabitants of these regions tend to be ethnic minorities in China, and it’s possible they have larger builds genetically.

Dairy intake correlates positively with HDL or “good” cholesterol (+29), but not with LDL or “bad” cholesterol (-15). And in terms of general diet composition, dairy-eating regions had a low intake of fiber (-28), soluble carbohydrates (-23), and plant protein (-25) but a high intake of total protein (+80), animal protein (+99), total fat (+78), and total calories (+62).

So does all this animal protein and fat equate to more chronic diseases, especially cancer? Let’s take a look-see. Surprisingly, there’s only one statistically significant correlation, so I’ll list even the weak correlations to give the full picture.

NEGATIVE CORRELATIONS (more dairy = fewer of these diseases)

Liver cirrhosis: -21
Peptic ulcer: -19
Lymphoma: -19
Other metabolic diseases: -17
Rectal cancer: -14
Penis cancer: -13
Diabetes: -10
Colorectal cancer: -9
Death from all causes: -9
Other heart disease: -9
Bladder cancer: -8
Diseases of the blood and blood-forming organs: -7
Colon cancer: -6
Leukemia: -6
Liver cancer: -4
Neurological diseases: -2

POSITIVE CORRELATIONS (more dairy = more of these diseases)

Hypertensive heart disease: +30*
Stomach cancer: +10
Breast cancer: +9
Stroke: +9
Oesophageal cancer: +8
Death from all cancers: +7
Myocardial infarction: +6
Brain cancer: +4
Cervix cancer: +2
Rheumatic heart disease: +1

Remember that, for both positive and negative correlations, smaller numbers are essentially insignificant. Zero is perfect neutrality, but we rarely get a statistical zero in the real world—especially when we have a maximum of 65 data points to work with. With the exception of hypertensive heart disease, none of the positive correlations appear meaningful, and perhaps only lymphoma and “other metabolic diseases” warrant further study among the negative correlations. (Liver cirrhosis and peptic ulcers, the other marginally strong inverse trends, most likely have non-nutritional causes.)

Hypertensive heart disease

Unlike atherosclerosis, hypertensive heart disease is not caused by plaque building up on arterial walls. With this condition, chronic high blood pressure (AKA hypertension) forces the heart to work harder, leading to a thickening of the muscle. Does something about dairy consumption raise blood pressure and lead to a big heart (the diseased kind, not the generous kind)?

Looking solely at the “hypertensive heart disease” variable, we can see that a few other factors outshine the dairy correlation. Daily salt intake correlates at +37, salt intake plus urine salt correlates at +50, weight correlates at +33, and wheat flour correlates at +54. The “other foods” category—which includes vinegar, MSG, baking powder, tea, and melon seeds—has a correlation of +51. Negative associations include rice (-45), yearly green vegetable intake (-36), steamed bread and pancakes (-57), and daily alcohol consumption (-27).

Since we really only have three counties that consumed significant levels of dairy, it won’t be easy—and maybe not even possible—to pinpoint the role of dairy itself in hypertensive heart disease. Here’s our untampered graph, charting every county that reported hypertensive heart disease  mortality.

Kinda sparse, huh?

There certainly is an upward trend amongst dairy-eating counties, but let’s face it: We’re basing that on two measly dots. Dots that, more importantly, represent complete dietary rebels in this data set. Along with consuming milk products, our two dairy-eating regions have some other nutritional divergences compared to the rest of China. For instance:

  • Both significantly fewer vegetables than any of the other 65 counties in the China Project: Tuoli only consumes green vegetables an average of two times per year (!) and Xianghuang qi consumes them 24 times per year, about twice per month. The average intake for all counties is closer to 200 times per year. Could lack of green vegetables, which we already know correlates with hypertensive heart disease, be raising these county’s rates?
  • Tuoli consumes a whoppin’ average of 172.5 grams of protein per day, 134.55 of which come from animal sources. Even a gym-rat bodybuilder might consider that excessive. No other county comes close to that intake; the next highest is 90.8 grams per day. Could an uber-high protein intake contribute to hypertension?
  • Tuoli also consumes far more fat than any other county: 185.8 grams per day. Wowza. The average for all of China is 44.2. Is a high fat intake playing a role?
  • As you’d expect from a very high-fat diet, the Tuoli eat more calories per day than any other county studied in the China Project: 3704 on average. No, that’s not a typo. They’re also the heaviest county among the 65. Do these folks have higher rates of obesity, which surely is a risk factor for hypertensive heart disease?
  • Both dairy-eating counties had a higher levels of dietary and urine sodium than average. This one’s a no-brainer; excess sodium is a well-known cause of high blood pressure and, consequently, hypertensive heart disease.
  • Both dairy-eating counties eat significant quantities of wheat flour, which—as you shall soon see—has some really mind-bogglingly crazy associations with cardiovascular disease. Seriously, it’ll knock your socks off. But that’s a few more posts away.

And despite eating almost two pounds of dairy a day, Tuoli’s hypertensive heart disease mortality (far right dot) is still surpassed by two other counties—both of which eat no dairy at all. And the disease rate for Xianghuang qi county is right smack dab in the middle of the data set range, ranking behind six dairy-free counties. Even if dairy is a factor in developing hypertensive heart disease, which we can’t say for certain, it’s clearly not the only contributor.

At any rate, I’m not interested in vilifying nor vindicating dairy here; I’m just exploring alternative possibilities for the correlation between milk products and hypertensive heart disease.

Bottom line:

  • China’s dairy eaters don’t have significantly more cancers, myocardial infarction, stroke, and so forth than the dairy-free regions.
  • Dairy’s only significant mortality correlation, hypertensive heart disease, may be related to any number of variables we don’t have enough data to tweeze apart. (Lack of vegetables, excess sodium, high body weight, and high caloric intake, to name a few.)
  • Despite T. Colin Campbell’s findings with the milk protein casein spurring cancer in lab rats, there does not seem to be a correlation between high dairy consumption and cancer in the China Study data.

Are you as fascinated as I am with the Tuoli’s incredibly high-fat, high-protein, vegetableless diet? Are you wondering what their disease rates are, how long they live, and whether they’re any healthier or sicker than the Chinese eating plant-based diets? I sure am. Up next will be a mini-post about the Tuoli specifically, including answers to the aforementioned questions. This should be pretty interesting!

About these ads

Actions

Information

31 responses

20 06 2010
Raw.la - Raw Food in The News and Around The Web

[...] A Closer Look at the China Study: Dairy and Disease [...]

21 06 2010
Sue

Can’t wait to hear about the Tuoli population. Thanks for the dairy report!

23 06 2010
Tuoli: China’s Mysterious Milk Drinkers « Raw Food SOS: Troubleshooting on the Raw Food Diet

[...] China’s Mysterious Milk Drinkers 23 06 2010 As I mentioned in the previous post on dairy consumption and disease in China, there’s a fascinating little county by the name of “Tuoli” situated in northwest [...]

7 07 2010
The China Study: Fact or Fallacy? « Raw Food SOS: Troubleshooting on the Raw Food Diet

[...] You can read more about the connection between dairy and hypertensive heart disease in the entry on dairy in the China Study. [...]

10 07 2010
Dana

It’s odd that you don’t want to take a closer look at the wheat consumption. That’s got a higher correlation with hypertensive heart disease than even salt does. And I think salt’s rather suspect as an alleged disease agent. There seem to be some people genetically encoded to respond to high sodium with hypertension, but that doesn’t seem to be the normative human experience until there is already some damage done to the body’s BP-regulating system–and that damage is usually done by insulin, from what I’ve heard, not salt.

Wheat’s relatively new to the human diet, we haven’t had time to evolve to meet its challenges, and there are a LOT of people walking around with undiagnosed gluten intolerance of some degree.

You could say milk’s relatively new in the diet too, but many of the components of milk are familiar to our bodies in a way that wheat can never be; they’re animal proteins and fats, after all. Even lactose is fairly closely related to glucose; it can be turned into galactose which then converts into glucose.

Wheat proteins are quite another matter and on top of that, wheat contains compounds that chelate minerals, not something you want to happen unless you planned for it. Paleopathologists can tell whether a thousands of years old skeleton belonged to a hunter-gatherer or a farmer just by how well-developed the skeleton is and how many lesions are in the bones (an indicator of disease). The farmers were usually shorter and sicker.

The Maasai consume milk on a regular basis if following their traditional diet and they do have hardening of the arteries but do not seem to have any significant heart disease risk despite that fact.

Oh, something else: Just because a person consumes a high-calorie diet does not mean they will become obese. Again, the wheat consumption alone explains the weight gain of these groups. You should look into the experiences of people who deliberately follow a higher-fat, moderate-protein, lower-carb diet and do not restrict calories, it’s an eye-opener.

22 07 2010
julianne

The Masai consume milk A2 not A1 milk. There is an entire book devoted to the association of health problems associated with A1 milk.
I knew the scientist looking at the data – he said the correlation between heart disease and A1 milk was so strong – if you told him the percentage of A1 to A2 cows in a country he would predict the heart disease rates with accuracy.

The book is “The Devil in The Milk”
http://thebovine.wordpress.com/2009/03/20/the-devil-in-the-milk-dr-thomas-cowan-on-how-a2-milk-is-the-answer-to-the-mystery-of-why-even-raw-milk-sometimes-does-not-seem-to-be-enough-of-an-improvement-over-store-bought/

26 08 2010
kelly

Denis, you are is right to say there’s no correlation to show milk’s and meat’s link to increase in cancer (if you just compared Tuoli’s diet to the other China counties). I think it’s also due to the fact that these Chinese – who eat a lot of dairy and meat products – eat naturally fermented products, mushrooms and tea. In my opinion, whatever ‘damage’ that’s been done to meat-eaters, were healed by those foods.

You may say my diet may be far worse than that of a vegetarian, because I am a vegan. My body age according to a bio scan, is 13 years younger than my real age.

26 08 2010
julianne

I’m a paleo eater and an inner scan tells me my bio age is 12, my real age is 50. My health has never been better since cutting grains.

26 08 2010
kelly

I apologise for the typo, should be “Denise”. :)

27 08 2010
Ed Terry

I developed a blood pressure problems when I drastically reduced carbs and increased fat. Part of the problem is that I eat only whole foods now, so my sodium intake is less than 1,600 mg/day. However, when my intake was 3,000 mg a day, I also had this problem. At a carbohydrate intake of less than 70g/day, my blood pressure dropped too low. I now have to add sea salt to some of my food in order to raise it to 120/70.

At 5’10″ and 182 lbs, when I drop my caloric intake below 3,000 calories I lose weight. I also consume far less protein than I used to, although I do eat three eggs every day. 10 years ago, eating a lower fat diet, my HDL was 32. Eating a high-fat diet (>260g/day) it’s now 76.

7 09 2010
Dairy and breast cancer

[...] cancer, oesophageal cancer, brain cancer, cancer of the cervix, and deaths from cancer overall. http://rawfoodsos.com/2010/06/20/a-c…y-and-disease/ In any case, the source of 'dairy' Miss Minger is referring to is yaks, not the (crossbred, [...]

22 10 2010
Alan

Ed Terry

what is a “whole foods” diet….. you eat the feathers with the chicken? The hoofs with the steer? The thorns with the raspberries? The scales with the fish?

22 10 2010
Ed Terry

Alan, if you consider feathers, hoofs, scales, and thorns “food”, then by all means go for it.

30 10 2010
David

You say, “This one’s a no-brainer; excess sodium is a well-known cause of high blood pressure and, consequently, hypertensive heart disease.”
I would have thought you would be familiar with the work of Gary Taubes. He dispelled the myth that sodium is a significant factor is hypertension, except for a small portion of the African American population. His article in Science, “The (Political) Science of Salt”, traced the history of how this idea became a “fact” without any quality evidence.

7 02 2011
Another Halocene Human

That’s funny. I can manipulate my bp very effectively with salt intake/outflow (sweating makes it go bye-bye, of course), and to the best of my knowledge me and my burn-in-the-shade Irish hide are not of near-term African ancestry.

I’ve read 1/3 of the US population are sensitive to sodium in their diet. If you are part of the insensitive 2/3, more power to you. (But neener neener, I can use salt in place of aspirin!!)

18 12 2010
Andrés

Thanks for your blog Denise!

I am glad that you have enough brain-power left to stop “purifying” yourself on time.

While we wait for that high expected wheat analysis of yours, I would like to know if you or anyone else have received new information about the cooking/fermenting methods for it. I know something have being said in some other comments in some other entries of this blog, but it doesn’t make sense to me yet how it may be these correlations on hypertensive heart disease have emerged: “wheat flour correlates at +54″, “The “other foods” category—which includes vinegar, MSG, baking powder, tea, and melon seeds—has a correlation of +51″ –I know, baking powder is only one of many but…– and over all “steamed bread and pancakes (-57)”.

Just to not add another post elsewhere, I must point out to you that in order to stop the wikipedia “censorship” (certainly not everybody is able to replicate your analysis or Ned’s –none at the veganism side for now–, so it seems legit to be delegated to the peer-review process) the course of action is clear. I think that yourself and Ned Kock (and perhaps Chris Masterjohn also) should team up and submit to a peer reviewed magazine. Just choose any non biased one.

Live long and prosper :-) (it seems to me that this veganism stand seeks a pure oversimplified ethically superiority too, while the reality is more Babylon-5 like).

19 12 2010
neisy

Hi Andrés,

If you’re talking about the wheat eaten in the China Study, the only specific info I’ve come across is from one of Campbell’s papers (full text here) that mentions Tuoli was eating nan (baked flat bread) and Jiexiu was eating “noodles made from wheat.” I don’t think anything was recorded at all about cooking methods or fermentation, and I’m not too optimistic about that information surfacing any time soon. Jane from the comments here contacted Richard Peto (one of the China Study researchers) a while back about wheat, and he wasn’t even sure whether it was eaten whole or refined, much less how it was prepared.

The correlations with “other foods” are a little misleading, because only a handful of counties had an intake greater than zero for that category — and most of those counties were wheat-eating regions with higher heart disease rates, which is the real reason the correlations with “other foods” were so strong. When you stratify the data to account for wheat consumption, “other foods” loses its correlations with vascular diseases. It’s a bad idea for a category anyway since it lumps so many different things together — I think it was just a catchall for miscellaneous items that didn’t fit anywhere else but that the researchers wanted to account for anyway.

Regarding the censorship: I’m so afraid of jinxing this, but I’m working with a very brilliant, published, math-savvy woman right now to see if we can get a reanalysis of the China Study data peer-reviewed. Shhh, don’t tell anyone yet. :) (Posting this on the internet is a good way to keep it secret, right?) I certainly wouldn’t be opposed to teaming up with Ned or Chris to do something as well if they’re willing!

20 12 2010
Andrés

Excellent! Just to avoid jinxing it too, I won’t comment with anyone else that piece of info ;-).

About the wheat issue, a posteriori it is clear that those pieces of information should have been recollected. It is really a pity that finally it won’t happen. I hope wheat effects are ameliorated by fermentation and by pumping up fish intake since I am not giving up bread yet :-). Nevertheless I am going to bring down pasta consumption a little more too, just in case. I keep eating almost anything but these days my diet is more similar to a paleo one (lot of nuts, steamed broccoli, dairy products –I just love cheese– and so on).

10 04 2011
Susan

This discussion points out one important fact. There are so many variables that go into causation of complex diseases like cancer and heart disease, that it becomes impossible to isolate any one variable as “the” cause. It’s clear to me that the correlations do not prove Campbell’s theory that the only rational approach to eating is to become a vegan. I wonder why Campbell chose to publish for the public rather than for a scientific publication which would have subjected his work to critical peer review. Without that his conclusions are inherently suspect. Having listened to friends who are enamoured with his theories–highly intelligent people–draw completely unjustified conclusions, I can see how easily the general public is taken in.

15 05 2011
Forks Over Knives Review | Ready State Fitness

[...] A Closer Look at the China Study: Dairy and Disease (Associations the “dairy” variable has with various diseases in China.) [...]

4 08 2011
Where Fruit Meets Ice Cream « Big Apple Sighter

[...] “We are not really sure what the problem is.” Maybe it was the recently popular China Study link between dairy and disease, or my own lactose intolerance and general congestion after eating [...]

29 08 2011
Gordon Wayne Watts

I looked at the graph and did an estimated linear regression in my head (since I can’t get my word processor to do it – LOL).

Here is my analysis:

#1 -Apparently, with such a wide variation in heart disease among the non-dairy-consuming regions, is is obvious that there are factors OTHER than dairy to contribute — such as maybe stress or other dietary factors (or even sleep and/or exercise factors).

#2 – Nonetheless, I can see some correlation, and this graph, taken together with other data in my head from other sources, produces some VERY small p-Value, on the order of like 0.0001 or so.

#3 – The ‘R’ (relatedness) Value is not as large for heart disease vs dairy as it is with other things vs dairy, as evidenced by the facts in #1 above.

#4 – I suspect that the lower association (‘R’ value) with diseases in general is that less processing of the milk is done in China vis-a-viz other countries, though I do not know this for a fact. (Highly processed milk is MUCH worse for your health, or so I infer.)

#5 – For anyone new here, I have posted a more detailed analysis on both my own blogs’ Health sections:

http://GordonWatts.com/#health
or:
http://GordonWayneWatts.com/#health

-as well as other comments threads of Denise’s blog here, and she was kind enough to let me. I address the correlation and causation relationship, and admit that a common cause, though unlikely or small in effect, could be a factor.

Gordon W. Watts
A.S., Electronics Technology, UEI, Valedictorian
B.S., Biological / Chemical Sciences: double major with honours
The Florida State University

29 08 2011
Gordon Wayne Watts (@Gordon_W_Watts)

What! — Stupid Facebook app did not take my post — it appears only in that Browser –so, I apologise if it double-posts… – trying the Twitter app this time:

… as I was saying…

I looked at the graph and did an estimated linear regression in my head (since I can’t get my word processor to do it).

Here is my analysis:

#1 -Apparently, with such a wide variation in heart disease among the non-dairy-consuming regions, is is obvious that there are factors OTHER than dairy to contribute — such as maybe stress or other dietary factors (or even sleep and/or exercise factors).

#2 – Nonetheless, I can see some correlation, and this graph, taken together with other data in my head from other sources, produces some VERY small p-Value, on the order of like 0.0001 or so.

#3 – The ‘R’ (relatedness) Value is not as large for heart disease vs dairy as it is with other things vs dairy, as evidenced by the facts in #1 above.

#4 – I suspect that the lower association (‘R’ value) with diseases in general is that less processing of the milk is done in China vis-a-viz other countries, though I do not know this for a fact. (Highly processed milk is MUCH worse for your health, or so I infer.)

#5 – For anyone new here, I have posted a more detailed analysis on both my own blogs’ Health sections as well as other comments threads of Denise’s blog here, and she was kind enough to let me. I address the correlation and causation relationship, and admit that a common cause, though unlikely or small in effect, could be a factor.

Gordon W. Watts
A.S., Electronics Technology, UEI, Valedictorian
B.S., Biological / Chemical Sciences: double major with honours
The Florida State University

22 09 2011
24 09 2011
Debunking Myths About An Animal Based Diet | EUPHORY

[...] Dairy and disease [...]

28 11 2011
‘The China Study’ and ‘Forks over Knives’, links to critiques | Julianne's Paleo & Zone Nutrition Blog

[...] A Closer Look at the China Study: Dairy and Disease (Associations the “dairy” variable has with various diseases in China.) [...]

27 03 2012
Quora

What are the health effects, if any, of long-term protein powder consumption?…

I’m not aware of any long-term studies on protein powder consumption. I still supplement with protein powders sometimes, but I’ve substituted away from them and towards real food for a few reasons: First, a Consumer Reports study [1] found high level…

21 09 2012
“Forks Over Knives”: Is the Science Legit? (A Review and Critique) | Justin Penoyer

[...] Dairy and disease That should cover it, right? Moving on… [...]

28 12 2012
Čínska štúdia a teória krvných skupín | Žrádielko

[...] obe teórie vyvolávajú (použité vedecké metódy podporujúce teóriu krvných skupín, či robenie záverov o nevhodnosti mlieka napr. podľa testovania na potkanoch [...]

12 02 2013
China Study: A Voice for Veganism | Josh Mitteldorf

[...] have bet the farm by including dairy along with meat in their conclusions. The problem here is that only three of the 65 counties had any dairy at all in their diets, and zero relied primarily on dairy for protein. The Campbells [...]

29 12 2013
Mike Kaplan

Parts of western China are a high plateau with elevation above 10,000 ft – some over 15,000 ft. I believe this includes the Uyghur region. High elevation regions have lower air pressure and less oxygen. Hypertensive heart disease could come about as the heart pumps harder to get oxygen to the body. It would be interesting to see the correlations after this variable is controlled for.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s




Follow

Get every new post delivered to your Inbox.

Join 7,177 other followers

%d bloggers like this: