Sunday, May 30, 2010

Eating Processed Meats May Raise Risk of Heart Disease and Diabetes?

Dr John Ray comments:

This is rubbish. You can get anything you want out of a meta-analysis and the fact that the conclusions below were based on only 20 out of 1600 relevant studies certainly facilitates that. Biased selection of what to include is the bane of meta-analyses. I have seen it at work in my own field -- where "awkward" results were ignored.

Furthermore, a meta-anaysis of epidemiological conclusions is just tells you what the usual assumptions are, nothing more. They do have the grace below to admit that they have NOT established cause and effect but then go on to talk as if they have!

I wonder how many of the studies they used controlled for social class? That is the big blind spot in most epidemiology

Another huge blind spot that certainly applies here is the decision to look at just one or two diseases in isolation. Something that increases one sort of disease may decrease the risk of another disease. So overall mortality is what we need to know and that is seldom looked at.

And it would seem likely that the researchers DID have data on many diseases among meat-eaters. Did they just pick out the diseases that suited their demonization of processed meats? Were some other diseases REDUCED among eaters of processed meats?

What a crock!

The rest here

Posted via email from Garth's posterous

No comments: