In October 2011, the Archives of Internal Medicine published a study called “Dietary Supplements and Mortality Rate in Older Women.” Researchers followed 39,000 older women over the span of more than 20 years and found that 40.8 percent of those who took a daily multivitamin had died by the end of the study, but only 39.8 percent of those who didn’t supplement had died. Cue media panic. The headlines from newspapers to blogs screamed that vitamins can kill you — all because of one dinky little percentage point reported in a journal with a history of supplement bashing.
Will taking a multivitamin kill you? Deeply, desperately unlikely. And a staggering number of flaws in this study illustrates why sometimes studies don’t tell us anything except what researchers them to.
• When this study began, in 1986, the mean age of the subjects was 61. That means, that in 2008, when scientists looked at death rates, the mean age of the subjects was 83. The “mean age” means that some subjects were younger and some were older than 83 — and some were dead. Surprising? Hardly.
• The way researchers identified which subjects took vitamins was by asking them in questionnaires sent in 1986, 1997 and 2004. There were no blood tests to measure vitamin or mineral levels, there was no surveillance to make sure that subjects did indeed take their supplements every single day over those many years, and there was no cognitive test to ensure that subjects remembered what they did during the previous 11 years. Do you remember what you were taking 11 years ago?
• Correlation does not equal causation. Remember this. It means that just because two traits are seen to go together does not mean that one caused the other. Researchers did not establish that it was specifically the multivitamin supplements that killed the subjects. In fact, it could simply have been old age that was the true culprit.