Damn if PZ didn’t beat me to this one:
A federal panel concluded yesterday that there is not enough evidence to recommend for or against use of multivitamins and minerals — the popular dietary supplements taken by more than half of American adults in the hope of preventing heart disease, cancer and other chronic illnesses.
Americans spend an estimated $23 billion annually on various multivitamins and multi-mineral supplements, the 13-member panel found. One of the latest federally funded national surveys showed that 52 percent of adults reported taking multivitamins. Slightly more than a third of those surveyed said that they take these products regularly.
Several studies suggest that consumers believe that taking a multivitamin promotes health and can prevent some common diseases, but the evidence for both is “quite thin,” said the panel’s chairman, J. Michael McGinnis, senior scholar at the National Academy of Sciences’ Institute of Medicine.
That’s no big surprise, as the evidence that supplementing one’s diet with additional vitamins will do any good at all if you’re already eating a balanced diet has been shaky at best for a long time.
Indeed, the panel only found three situations the evidence indicates a clear benefit from vitamin supplementation: prevention of birth defects of the brain and spine by giving women of child-bearing age folic acid; reduction of macular degeneration by supplementation with vitamins C and E, beta carotene and the minerals zinc and copper; and lowering the risk of bone fractures in post-menopausal women through calcium and Vitamin D supplements.
In addition:
But the panel also found that multivitamins “are virtually unregulated” by the federal government. It advised Congress to expand the Food and Drug Administration’s authority and resources to require manufacturers to disclose adverse events, ensure quality production
What was most amusing was the response of the vitamin industry:
The dietary supplement industry criticized the panel for not considering less scientifically rigorous studies that point to multiple benefits from multivitamin and mineral use.
“They have done exactly what they set out to do, which is a review based solely on randomized controlled trial data, which results in a misleading picture,” said Annette Dickinson, past president of the Council for Responsible Nutrition, an industry trade association.
Damn those scientists for insisting on rigorously designed, adequately powered randomized clinical trials to answer medical questions! How dare they? What do they think this is, science? (Hmmm. I wonder why the vitamin industry would say such a thing.)
“There’s huge evidence that these products are safe,” Dickinson said.
Mrs. Dickinson, Mr. Strawman. Mr. Strawman, Mrs. Dickinson. That’s not what the panel said. The panel said that the evidence for the value of vitamin supplementation is very thin, except in a few situations. They also expressed a reasonable concern that the industry is essentially unregulated and that people may be getting more vitamins than they need, some of which can cause problems when taken in excess. Their conclusion was that, except in the cases above, there’s inadequate data to support or reject the use of vitamin supplementations.
Note also that the industry isn’t exactly disputing the actual conclusions of the panel; it’s only complaining that the panel didn’t look at less rigorous scientific studies. Sorry, Annette, but when there are decent randomized blinded clinical trials available addressing a medical question, they almost always trump less rigorous trials. The only exception is when a flaw in the RCT exists that invalidates its results. Indeed, physicians should only make decisions based on lower levels of evidence than RCTs when no good RCTs addressing a clinical question exist. If well-designed, well-powered clinical trials show a result for a question, it would have been irresponsible of the panel to put much stock in less rigorous studies that are prone to more biases and problems that RCTs.
Yep, all that sciency-ness can be frustrating.
