Are People Who Take Vitamins More Likely To Die?

A recent study in the Archives of Internal Medicine has added fuel to the debate regarding vitamin use in healthy adults .  The study showed that vitamin users were more likely to die of any cause over several decades than non vitamin users.

The past several years has seen several journal articles about vitamins and minerals that have provided a wide array of information about their use to the general public.  The information these studies have given usually has been confusing, vague, often contradictory and the result of study methods that are confounded.    The general message of these studies has been that vitamins have not been proven to have great health benefits to healthy adults and may in fact be harmful.  This study does little to add concrete information to the debate about the benefits or harms of vitamin use.  The same questions that existed about vitamins last month persist today. 

 

 

Medicine still doesn't really now much about the long term use of vitamins in healthy adults and the debate rages on about what the effects actually are. As the authors of the recent paper pointed out "Although dietary supplements are commonly taken to prevent chronic disease, the long term consequences of many compounds are unknown."

This particular study looked at a large group of woman from Iowa over several decades to assess what if any relationship vitamin use had on the woman's overall mortality.  Data was collected on over 40,000 woman who completed surveys regarding vitamin use in 1986,1997 and 2004.    The results seemed to carry a daunting message regarding vitamin use. All the vitamins and minerals studied (with one exception) were associated with a higher risk of all cause mortality.


Specifically the authors looked at Multivitamins, Folate, B12, B6, Magnesium, Zinc, Copper, Iron and Calcium. Of these only Calcium use was associated with a lower rate of mortality. All the other vitamins studied showed an increased rate of death.  The magnitude of the association in all cases was small but statistically significant.


So healthy adults are again deluged with another piece of information about vitamin use and are left to wonder if vitamins are associated with a increased risk of death and if healthy adults should stop taking vitamins?


Before we can assess this we need to look at the old paradox of association and causation. We can't assume that the association we see in this study is because vitamins somehow contributed to early death. A very likely explanation for some or all of these findings observed is that people who take vitamins are generally at baseline less healthy and more likely to have chronic disease than those who don't. By studying vitamin users the authors selected a group of patients who were by definition sicker and more likely to die than the general public for reasons that had nothing to do with vitamins.

It is imporant to note the authors did correct for Cardiovascular Disease and Diabetes in the results and the observed association held up. However they did not correct for many other chronic diseases that make people more likely to die and may also lead to vitamin and supplement use such as kidney disease, cancer, anemia, HIV, mental illness and many, many other medical problems.  

We can use iron as an example of the flaws in this study.  If you read the paper you will notice the strongest association the authors found is between iron supplemention and death rates. They even found a dose response relationship between iron and death rates, with the highest doses of iron leading to a 65 percent increase in death risk.  Does this mean iron is a dangerous thing to take? 


As always the true answer is unclear. Iron may contribute to mortality but some of the sickest individuals in the world today are on very high doses of iron because of severe chronic illness leading to low blood counts.   The authors speculate that perhaps iron leads to some "pro oxidative" state that contributes to disease and explains the results.  To me this is complete speculation and almost irresponsible to put in a paper. It may be worth investigating if iron causes disease and death based on these results, but it is hardly the most likely explanation.  The most likely explanation for why this study showed an association between iron use and death rates is that very sick people are often advised to take iron to help manage chronic, debilitating disease (because of low blood cell counts). It should not be a suprise to anyone that more of those people died over 2 decades of follow up.   I would expect to find similar relationships contributing to  (but not neccesarily invalidating) the observed results for all the vitamins and supplements looked at, including calcium which showed a beneficial effect on death rates

So where does this leave us for vitamin use? Can we completely ignore these results because the methodology of the study was questionable?  



While there is no doubt that vitamin use has been shown to be helpful in specific medical conditions (such as Vitamin D and Calcium preventing osteorporosis in at risk individuals) for the general public if you don't fall into a specific group that needs to take a vitamin or supplement for a medical reason than the wisdom of taking vitamins is questionable.  More and more research has failed to find a health benefit to vitamin use in healthy adults (including this study) and some have shown a potential harm.

It is possible that all the studies showing harms of vitamins had such flawed methodologies that they actually masked a truly beneficial relationship for vitamin use in healthy adults but this is unlikely.  It seems more plausible that evidence supporting benefits to vitamin use in healthy adults without medical problems is scarce and the harms may be real.   There are specific medical conditions where vitamins are definitely beneficial but until more is known healthy adults should have conversation with their physician about if vitamin use is appropriate for them.