Nonprofit, nonpartisan journalism. Supported by readers.


UCare generously supports MinnPost’s Second Opinion coverage; learn why.

Dietary supplements associated with increased risk of death, a U of M study finds

Taking vitamin and mineral supplements is associated with an increased risk of death in older women, a new study by researchers at the University of Minnesota has found.

The study findings, which were published Monday in the Archives of Internal Medicine, add to the growing body of research that suggests the routine use of dietary supplements may be harmful rather than beneficial to our health.

Yet, despite such findings, Americans are taking dietary supplements in ever-increasing amounts. One-half of all adults in the United States take some kind of vitamin or mineral supplement. In 2010, annual U.S. sales of these products reached more than $28 billion.

Supplements do help a small minority of individuals with a specific deficiency condition (anemia, for example), but most people who take them are well nourished. Individuals turn to supplements primarily because they hope the extra nutrients will prevent chronic health conditions or somehow make them feel better.

“In some cases, they are medically prescribed, and there are good reasons for taking them,” said Lisa Harnack, one of the study’s co-authors and a professor of epidemiology and community health at the University of Minnesota’s School of Public Health, in a phone interview Tuesday. “But your average healthy person may not need to be taking a dietary supplement.”

Lisa Harnack
Lisa Harnack

“We found, that for older women, it looks like there’s not much benefit from taking most dietary supplements — and for some women, there is potential for harm,” she added.

The study’s details
For their study, Harnack and her colleagues used data collected from 39,772 women participating in the U of M’s long-running Iowa Women’s Health Study. The women, whose average age was about 62 at the start of the study in 1986, were asked about their use of dietary supplements three times: at the study’s onset and again in 1997 and 2004.

When the study started, 62.7 percent of the women reported using at least one supplement daily. By 2004, that number had increased to 85.1 percent.

The study found, however, that taking supplements did not reduce the women’s total death rate. In fact, those women who supplemented their diets with multivitamins, vitamin B6, folic acid, magnesium, zinc, copper and iron died at higher rates during the study’s 19 years than those women who didn’t take supplements.

The absolute increases in risk ranged from 2.4 percent for multivitamins to 18 percent for copper.

Only calcium supplements were associated with a lower absolute risk (3.8 percent).

The authors are questioning that calcium finding, however.

“We are skeptical about this finding because clinical trials have been done and not shown benefit,” said study co-author David Jacobs, an epidemiologist and professor of public health at the U of M, in an e-mail exchange with MinnPost.

David Jacobs
David Jacobs

Indeed, as a commentary that accompanies the U of M study points out, a recent meta-analysis of randomized controls trials found that calcium supplements are associated with an increased risk of heart attacks. Other research has found an association between high calcium intake and an increased risk of prostate cancer.

(The commentary incorrectly states that the U of M study found that vitamin D was associated with a lower risk of death. The study found no such effect, according to Jacobs and Harnack.)

The current study is an observational one, which means it can’t prove a direct cause-and-effect relationship between dietary supplements and increased risk of death. It can only show an association between the two.

Although the researchers adjusted the data for a variety of confounding factors, such as age, diet, physical activity, and diagnosed diseases (like high blood pressure and diabetes), something not yet identified may explain the study’s findings.

Some of the women in the study, for example, may have begun taking the supplements because they weren’t feeling well — perhaps because they had a yet undiagnosed disease. “We can’t know that from this data,” said Harnack.

So what should consumers do?

“Supplements should be taken for an indicated medical issue for which they are known to be a solution, most likely with consultation with a physician,” said Jacobs. “Our observational findings do not suggest benefit for most supplements in terms of longevity of older women. Nevertheless, we do not rule out specific benefits related to function or other aspects of life. However, for most supplements, the amount of research that has been performed is not adequate to be sure what they do.”

In other words, stay tuned.

Comments (7)

  1. Submitted by Paul Brandon on 10/11/2011 - 12:32 pm.

    Good points!
    But as usual, the study (at least the summary available online) cites the usual second order statistics, which make it hard for a consumer to make an informed decision.
    For instance, to decide whether or not to take calcium, one should know (at least for one’s age and population category) what the odds of living another five years are with and without calcium supplements. This would incorporate all known risk factors as well as benefits associated with consumption of the supplement.
    THAT would be useful.

  2. Submitted by Pat McGee on 10/11/2011 - 03:06 pm.

    Followed out over time, we all die. That said. Are the ones taking supplements with doctor advise doing so icw other conditions that are the “real” cause of death? Without a control group, this study is just a bunch of malarky.

  3. Submitted by Paul Brandon on 10/11/2011 - 05:56 pm.

    The point of scientific research is to rule out alternative explanations of observed relationships, such as that between calcium intake and mortality.
    A control group is one of the better ways to do this, but hardly the only one.

  4. Submitted by Media ANHUSA on 10/12/2011 - 09:45 am.

    Big pharma profits down, no new drugs in the pipeline, running out or people to over medicate, time to attack vitamins! The bottom line. This study is a classic example of scientific reductionism being used to fulfill a particular need. In this case, it’s supplement bashing. Read our rebuttal

  5. Submitted by Sam Sewell on 10/12/2011 - 11:28 am.

    Here We Go Again! – The factual, scientific, and intellectual flaws in vitamin studies published by the media.
    Please feel free to download this report and add your own contact information at the bottom. Please give credit to the source: THE NATURAL ADVOCATE

    There has been a controversy swirling around the world about nutritional supplementation for decades. It seems like nearly every week you read stories in the news claiming a particular vitamin demonstrates NO effect on the prevention of a certain disease.

    Today’s news media went at it again. Here are some ideas for turning that misleading information into a way to introduce people to a healthier life for everyone & a better life for anyone.

  6. Submitted by James Blum on 10/12/2011 - 03:39 pm.

    I will ignore the spam from obviously self-interested parties above, but this study still sets off my BS detector. We have the usual poor science reporting going on here, in this case accompanied by some really specious conjecture on the part of the study authors. Apparently almost every science writer requires a daily (or perhaps hourly) reminder that CORRELATION DOES NOT EQUAL CAUSATION. For example, this study shows a CORRELATION between supplement-taking and death, but that correlation is not strong and there are so many other factors that influence overall health over time in an aged popolation (not to mention the issues with self-reporting of diet, exercise, etc.) that it becomes another expensive but essentially meaningless exercise. Does supplement-taking result in increased incidence of death, or does rational fear of impending death result in increased supplement-taking?

    Can we please do a study on the quantity and quality of education about statistics among science writers? Never mind, I have a pretty good idea what it will say without having to waste the money and time to gather the data.

  7. Submitted by Susan Perry on 10/13/2011 - 07:51 am.


    Did you read the entire post? I specifically describe the caveats you raise.


Leave a Reply