Flashbit
"The soul never thinks without a picture." Aristotle
Exploring Nutrition and Disease with the China Study
The China-Cornell-Oxford Project is a nutrition study that
has been called "the most comprehensive snapshot of diet, lifestyle and disease ever taken." [1]
It's the basis of a best selling book whose website says "The science is clear. The results are unmistakable." But it's a holy
war. Vegans and vegetarian writers say that the study proves that we should all be vegetarian, others say that it proves
that we should eat meat. The data is available here and
field descriptions here.
I've downloaded it and am creating visualizations so that you can explore the data itself.
The first visualization shows correlations of staple foods to causes of mortality. Each line is a staple food.
A line that goes above zero is a positive correlation between that food and the cause of mortality.
The list of staple foods is from this website, the Diet survey intake.
For this first, simple visualization, I'm including the staple foods, and two categories of causes of mortality. I chose the
causes of mortality that were described in [2] as Group A (More Typical of Developing Countries) and Group B (More Typical
of Western Countries) also known as diseases of poverty and diseases of affluence. A correlation number, however, is just the
beginning. To understand the significance of a correlation, go to the scatterplots below and click on the axis labels to change
what's being compared. In the scatterplots each dot is a county/gender/year combination, colored by year.
White dots are from 1983, blue dots from 1989. Mouse over to see what it represents, drag to select.
[1] TC Campbell, TM Campbell The China Study: The Most Comprehensive Study of Nutrition Ever Conducted And the Startling Implications for Diet,
Weight Loss, And Long-term Health
BenBella Books 11 May 2006
[2] T.Colin Campbell, Banoo Parpia, Junshi Chen
Diet, lifestyle, and the etiology of coronary artery disease: the Cornell China Study
The American journal of cardiology 26 November 1998 (volume 82 issue 10 Pages 18-21)