Whole grain intake study - dangers of observational studies

[Guest guest]

> Results: The results showed a significant inverse trend between

whole-grain intake and the metabolic syndrome (P for trend = 0.005)

and mortality from cardiovascular disease (P for trend = 0.04),

independent of demographic, lifestyle, and dietary factors.

--------------------------------------

There are hunders of observational studies like this, among them

studies saying coffee is bad, coffee is good, coffee prevents cancer,

coffee causes cancer, etc., etc.

The problem is, the devil is in the details. The people writing such

studies recognize that association is not causality, and they usually

put in a standard statistical disclaimer to this effect, and saying

that their results need to be confirmed by an interventional trial.

But then everyone goes and treats the study without this appropriate

caution.

In observational studies, investigators make an attempt to control for

confounding factors as best they can; e.g., " demographic, lifestyle,

and dietary factors " , but the problem is, that such statistical

control manipulations are far from perfect, and can sometimes even

introduce bias into a study.

I would never change my own dietary preferences based on an

observational study. Otherwise, I would have no peace, since they

often keep coming up with different and conflicting answers.

When you see such a study, you need to ask a common sense question:

OK, why is this one geezer eating whole wheat bread, and other one

eating plain white Wonder bread? These dietary choices don't just

happen for no reason.