from Marks Daily Apple
The second study found that multivitamin usage was positively associated with mortality in older women. Broken down by specific micronutrients, zinc, magnesium, vitamin B6, folic acid, copper, and (most strongly) iron were linked to higher mortality, while calcium was not. Sounds horrible, right?
But then you look at this other, recent “multivitamin and mortality” study, which distinguished between “baseline users,” “baseline non-users,” and “never-users.” A baseline user is someone who had been taking supplements before the start of the study. A baseline non-user is someone who had been taking no supplements before the start, but who began taking them during the study. A never-user is someone who took no supplements at all, neither before or during the study. Baseline users had the lowest mortality (both all-cause and from cancer). Never-users had the next lowest mortality, while baseline non-users who began taking supplements during the study were the most likely to die. Who else is most likely to start taking supplements later in life? Sick people. Unhealthy people. Who’s more likely to die? Sick, unhealthy people. In fact, the authors of the study conclude that a “sick user effect” likely explains the results.
In the original multivitamin study, no distinction was made between baseline users and baseline non-users. They were all packaged together, and I suspect the sick user effect is also at play in this one. Furthermore, as the Jaminets point out in a recent post, the authors also failed to adjust for the age of the participants. That older people tend to die more often than younger people is a pretty basic concept and a function of the aging process, so you’d think it would be included in their “multivariate analysis.” The study also showed that older people tended to take more supplements than younger people. And indeed – when you do adjust for age, select supplement usage actually lowered the risk of mortality.