[quote]orion wrote:
I would like to say that that would lead you to throw every epidemiological study into the dustbin, but they are of course totally peer reviewed.
Garbage, nonetheless.
Because, correlation does not mean causation, never has, never will, and for next to insignificant correlation in nutritional studies that is twice as true as it would otherwise be.
It works for overwhelming correlations, as in cholera is related to polluted water, but in the case of macronutritients and all kinds of diseases, not so much.
A hypothetical: If people who digest more fiber are healthier, does that mean that fiber is good for you?
If you answer is not a resounding “No! Not necessarily!” you are part of the problem.
That would be, at best, the basis of a clinical study, specifically a ward study, where you substitute with fiber and do nothing else.
Then you would have a point, but that is not how it is done.
What is done is that 90% of the money for such studies is granted by organizations that totally buy into the lipid hypothesis and you either swallow it or you become a cab driver.
You deal with these people on a daily basis, what are they going to choose?
[/quote]
I will be the first to say that the peer review process is not perfect. I see studies all the time that pass peer review and contain major flaws in methodology and conclusions, but it is better than nothing and does make it easier to sort through bullshit claims. Most reputable journals are fairly devoid of bias, at least in physics. That may not be the case in the biological science, but my fiance tells me it is not all that big of an issue and I believe her. Now, scientists working for a company are most definitely biased and their claims should always be taken with a grain of salt. For instance, a study done by a corn farming organization that shows that high fructose corn syrup is healthy should surprise no one, but should be considered suspect no matter what. One of the reasons that we have a tenure system and such things as unrestricted grants is to prevent things like what you describe. It is not fool-proof, but it is mostly effective. I was told once that the health issues caused by hydrogenated oils were known back in the 60’s or 70’s, but major companies managed to convince the FDA to allow their unrestricted use anyway. I don’t know if this is true or not, but stuff like that does happen.
One major problem with scientific research, though, is not so much the research itself, but the way people use the conclusions of that research. Scientific studies, by necessity, tend to be narrowly focused and use very stringent controls and one needs to be careful about claims made by third parties citing that study.
Using your fiber example, let’s say that a supplement cites that study in an article promoting a fiber supplement and let’s assume that the study did show a health benefit from an increase in fiber. What was the health benefit? A person who is thinking about buying that supplement should look into the study and see if the parameters under which the added fiber increased the health of the subjects are applicable to that person. If the study was done on people who had been fiber deficient for a long period of time, the the health benefits would be more pronounced then in a person who already takes in an adequate amount of fiber and thus would not need that supplement. The article did not necessarily lie if it did not lay out the conditions under which their product would be helpful, but supplement articles tend to be very misleading with the scientific evidence for the benefit of their products
EDIT: the mods spend a lot of time breaking up my long ass posts into smaller, easier to read paragraphs so I figured I would do it myself for once.
