Taking multivitamins to prevent a heart attack or stroke? They re probably not doing you any good, according to a study in this week s JAMA.
Dr. Howard D. Sesso, of Brigham and Women s Hospital and Harvard Medical School, analyzed data from the Physicians Health Study II, which followed 14,641 male physicians from 1997 through 2011. The study was a randomized, controlled trial, with approximately half the participants receiving a popular multivitamin supplement, while the rest received placebo pills. The researchers found no significant effect of multivitamin supplementation on major cardiovascular events, total myocardial infarction (heart attack) or total stroke.
ACSH s Dr. Josh Bloom, meanwhile, has a new op-ed in the Boston Herald examining recent claims that multivitamins might help prevent cancer. Dr. Bloom tries to educate the public on how the word significantly might sow confusion (perhaps intentionally so) when used by journalists describing a scientific study:
Clinical trials generate enormous volumes of data, and sophisticated calculations are required to process the data and draw any conclusions. An essential parameter called statistical significance determines the mathematical probability that the conclusions derived from the data in the study are real, as opposed to being coincidental. When that probability reaches 95 percent or higher a universally accepted value statistical significance has been reached.
Thus, when the authors say modestly but significantly, what they really mean is modestly and mathematically acceptable.
But just about everyone else will see significantly and read a lot, which is not even remotely the case.
ACSH s Dr. Ruth Kava adds, Indeed, a result may be statistically significant (or real ), yet have little or no significance for human health.
