OK, this is the graph of heart deaths plotted against fat intake, produced by Ancel Keys in 1953. It's a beautiful curve, utterly convincing to any Congressperson looking to find fame by funding a cure for heart disease:
Unfortunately Excel doesn't do nice sweeping curves, so I've mangled it by eye to a straight line thus:
Which is still very convincing. Slightly less convincing is when the choice of different countries from the same data bases suggests that dietary fat has nothing to do with heart disease and that heart disease is very rare anyway:
Of course choosing a few other countries might have given a negative correlation:
And, with a little effort, we can see that if only we could just get everyone's fat intake above 40% of calories heart disease would be a thing of the past, just so long as we chose the correct countries to look at:
So let's stop playing and look at the whole database from which Keys carefully selected his six countries:
OK, there IS a correlation. It's pathetic, especially compared to the original line swept in by Keys. Of course things get worse if you add in the Masai, the Inuit, the Rendile, the Tokelau and a few others, shown as red dots:
At this point you would have thought that the name Keys would have become a joke and people would simply have ignored him as a self publicising evangelist with scant respect for the truth. But Keys was nothing if not determined. In 1958, the year after Yerushalmy and Hillebo had shown him to be a charlatan, out came this paper explaining the mechanism by which the high fat diet apparently caused heart disease, even if it didn't. Serum total cholesterol!!!!!! Picture credited to Dr Ravnskov (couldn't get the original).
Of course adding a few data points, which Keys had conveniently forgotten, gives this plot, again credited to Dr Ravnskov:
Anyone for a straight line? These studies are the core of the lipid hypothesis. They are where it all came from and the "science" is total junk.
As a final comment on between nation studies this figure, credited to Dr Kendrick, shows the rates of death from CHD plotted against the percent of a given population who are frankly hypercholesterolaemic. He extracted the data from the little discussed MONICA study:
Obviously living in Russia, Northern Ireland, the Czech Republic or Lithuania is bad news for CHD. Forget cholesterol.
That's as much as there was time for on between nations studies. Essentially they provide absolutely no support for dietary fat or serum total cholesterol as causes of coronary heart disease. It seems impossible that they should have spawned the current climate of cholesterol psychosis, but they did.
Within nations studies comes next.
OK, time to knuckle down to five very long days at work...