The Bjelakovic meta-analysis (PMID: 173275269) made huge headlines back in 2007 around the world and you easily identified only one of the many big mistakes and distortions they made.
He did not. Meta-analysis of mortality data whether as a primary or secondary endpoint is legitimate. Whether it is
sufficient to arrive at a strong conclusion is an entirely different matter.
It's great for the scientific process if you point out mistakes in that particular meta-analysis, but what about the rest? Individual authors consistently found a null/detrimental effect of vit E on mortality.
Meta-analyses have also been performed on the major endpoints of interest e.g. cancer incidence and mortality from all or some specific cancers; CVD surrogates and outcomes (though, I can't find a more recent meta-analysis than 2004 for anything but vitamin E and hard endpoints).
Individual, well-powered trials have failed to show a benefit; that's a failure even when you want to use a broader approach to causality like Hill's criteria. As per the posted study: two thirds of studies missed their primary outcome.*
(refs. at request)
The following arguments by the authors are problematic:
The argument that the studies failed to find an optimal dose only strengthens the broad consensus: antioxidants from dietary sources are fine (and extremely complex), but supplementation still fails to replicate diet.*
Esp. their argument that studies did not select patients marginal or deficient in those antioxidants is weak. It's tautological: <RDA levels are
defined as those levels likely to be detrimental for some people. There is no need to fund many more studies on this topic.
The bigger issue is that those studies, even when optimistically interpreted, likely won't apply to us. No one denies that benefits are seen in certain subgroups(low baseline intakes, certain diseases, genotypes, etc). But healthy people, even more so life extensionists, will reach several RDAs of most nutrients and supplement those that they are lacking. The results are already disappointing in the general population, the effect on mortality is null/negative and long term effects are unknown. Therefore it is a recipe for disaster to recommend high-dose "shotgun" supplementation of antioxidants to healthy folks.
(better not start me on pooling data from heterogeneous populations with heterogeneous diseases using heterogeneous antioxidants. I don't get the point. For one thing, different primary endpoints are *not* equivalent because there are only a few major killers for healthy people..)
Do you suggest overwhelming incompetence on the part of thousands of authors, a conspiracy or something else?
It would be great if you could explain your position more clearly. Which exogenous antioxidants, at which dose, for whom (etc) based on what evidence do you consider necessary?
*the use of Hill's criteria and an (albeit indirect) comparison of supplemental and dietary antiox. can be found in Mente et al. "A Systematic Review of the Evidence Supporting a Causal Link Between Dietary Factors and Coronary Heart Disease"
Edited by kismet, 08 September 2010 - 02:59 PM.
