Member-only story

The AHA says Fasting increases cardiac risk by 91%. Are they really that stupid?

Correlation is not causation. Healthy User Bias.

Dr. Jason Fung

--

This week, the American Heart Association presented an abstract that suggested a 16 hour fast is linked to a 91% increased risk of heart disease. This study simply shows a correlation, which is very, very far from proving that fasting CAUSES more heart disease. Unfortunately, this did not stop the AHA from boldly proclaiming that correlation = causation. In this press release, they stray from stating the correlation link to saying ‘may raise’ which clearly implies causation. That’s a huge problem. Because it’s a bald lie.

American Heart Association implies causation

Journalists at various news outlets quickly parroted the view that fasting causes heart disease. To prove causation, that fasting caused heart disease, you need to do a randomized controlled trial (RCT). That is, you randomly give one group one intervention and another one you don’t. This eliminates the inherent problem of correlation studies. This stuff is so basic that I can barely believe I need to write this article.

Major Newspapers around the world parrot the implied causation

Correlation is not Causation.

Let’s start with some basic, basic epidemiology that every person who has taken any entry level statistics course should know. When two factors (call them A and B) are correlated, it means that when one goes changes, the other does too. This does NOT mean that A causes B.

For example, when people eat more ice cream, their death rate from drowning increases directly and significantly. This is a true and strong correlation because people eat more ice cream when it’s hot, and swim more and therefore have more drowning accidents. Ice cream is linked to drowning, but does not CAUSE drowning — obviously a very important distinction.

--

--

Responses (83)

Write a response