Data for the Number-Crunchers (Updated 7/31… It’s Coming!)

Update #3 regarding upcoming response:

Yeah, yeah, yeah. It’s now Saturday. Gotta love being a multiple-offense deadline breaker. (I tend to value thoroughness over timeliness–so anyone out there who was thinking of hiring me for any time-sensitive job, you’ve been duly warned.) I’m currently adding a final section on wheat to Campbell response #2, and then this puppy WILL be ready to post. Pinky swear! Thanks for bearing with me.

–(end update/start of older post)–

I’m excited to see quite a few people take interest in the China Study data (huzzah, numbers!), and even more excited that some of you are already posting the results of your analyses. To quote reader and blogger Ned Kock:

I hope more people will do their own analyses on the original data, like we have been doing. Then the discussion will move away from X or Y are saying this, to something more like “the data” is saying this.

Right on.

While I’m finishing a fairly laborious (you’ll see what I mean later) response¬† to Mr. Campbell, I thought I’d post some of the data I already have typed up for those of you who are gettin’ antsy. I’ll be updating this entry frequently as I upload more files, but here’s the first batch.

I’ll also use this post to link to anyone who has posted their results somewhere on the ‘net. Those will be right after the links to the data.

Also feel free to request any variable(s) you’re interested in analyzing, and I’ll type them up when I have a spare moment.


Myocardial infarction/coronary heart disease:

(includes total cholesterol, HDL cholesterol, green vegetable consumption, animal protein, plant protein, dairy variables, egg variables, meat variables, and fish variables)

(Note: included are the variables “amount of green vegetables consumed” and “frequency of green vegetables consumed” to illustrate the Green Veggie Paradox.)

Colorectal cancer:

(includes cholesterol, schistosomiasis, plant protein, and animal protein)

(A shout out to eds. Chen Junshi, T. Colin Campbell, Li Junyao, and Richard Peto for making this stuff available in book form.)

Reader links:

So far, we have two posts from Ned Kock:

  1. The China Study again: A multivariate analysis suggesting that schistosomiasis rules!
  2. The China Study one more time: Are raw plant foods giving people cancer? (This one’s particularly interesting: Ned used a nonlinear regression analysis on the data with no schistosomiasis infection, and uncovered a U-curve in the relationship between cholesterol and colorectal cancer. In other words, the counties with the lowest cholesterol and highest cholesterol had higher rates of colorectal cancer than the groups with more mid-range cholesterol, who appear the most protected. Ned offers a great hypothesis for this result in his post. Additionally, while animal protein consumption correlated strongly with total cholesterol, animal protein itself correlated inversely (beta = -0.31, p<0.10) with colorectal cancer, while plant protein correlated positively (beta = 0.47, p<0.01). Remember, of course, that correlation doesn’t equal causation, and this is just a sampling of the dizzying number of variables recorded in the China Study.)