[mrkeen]

    A: saturated fats

    B: LDL cholesterol

    C: heart disease

studies show saturated fats increase LDL cholesterol

    A => B

People who get heart disease are more likely to have high LDL cholesterol

    C => B

> Researchers have thus drawn a logical conclusion: a lot of saturated fat in the diet produces more cholesterol, which in turn increases the risk of heart disease.

    A => B /\ B => C

> But this has been surprisingly difficult to prove.

You don't say?

[robwwilliams]

Great book on this topic: Judea Pearl’s Book of Why. He discusses causal models in great depth. And a highly amusing read.

Mendelian randomization can sometimes “rescue” causality from large epidemiological studies. Pearl explains how.

[andybak]

If I'm reading you correctly - this is the classic false syllogism?

[mrkeen]

I think so.

The author wrote "People who get heart disease are more likely to have high LDL cholesterol". Had she written "People who have high LDL cholesterol are more likely to get heart disease" then the argument would have been valid.