A recently published paper from the Iowa Women’s Health Study (Arch Intern Med. 171:1625-1633, 2011) concluded that “In older women, several commonly used dietary vitamin and mineral supplements may be associated with increased total mortality risk; this association is strongest with supplemental iron.” For many, this report is alarming, especially since approximately half of the U.S. population takes at least one dietary supplement. However, we found insufficient evidence to support the claims of the authors that dietary supplement use is associated with increased risk of mortality in older women.
First, it is important to note that the Iowa Women’s Health Study is an observational study. In observational studies, scientists examine associations between certain dietary and lifestyle factors and the incidence of disease in the study population, which is referred to as a "cohort." Although observational studies provide valuable ways to generate hypotheses that can be further tested in randomized, placebo-controlled trials, they cannot establish cause and effect. Observational studies instead establish associations between a given trait and disease outcome, but they do this only after mathematically adjusting for multiple confounders—characteristics that might confound or confuse the association—such as age, gender, exercise habits, and body weight. Further, despite the statisticians' best efforts to take all of these confounding factors into consideration, there are numerous additional factors that haven't been discovered yet or were not measured in the study. This phenomenon is called "residual confounding" and is a major reason why observational studies can only generate hypotheses.
In the Iowa Women’s Health Study, 38,772 postmenopausal women were asked to report their supplement use at only three time points many years apart (1986 at baseline, 1997, and 2004), and their self-reported information was then related to the occurrence of death during the corresponding interval. The authors of the study emphasize that the most consistent findings were a dose-dependent increase in mortality risk with supplemental iron and a dose-dependent decrease in mortality risk with supplemental calcium.
After closer examination of the data, we note several important limitations:
1. Healthy person or survivor bias. The supplement users were healthier and lived longer overall. As a result, a higher percentage of them stayed in the study, while many non-users were lost to follow-up and their mortality data could not be assessed. The authors report that supplement users were, in fact, naturally "healthier" compared to non-users at baseline, characterized as having a lower prevalence of diabetes mellitus, high blood pressure, and smoking status; having a lower body mass index (BMI) and waist-to-hip ratio; being more physically active; and being more likely to have healthful dietary patterns. This differential retention during the study period biases the results, contributing to a spurious association between supplement use and mortality.
2. Reverse causality. People may begin taking supplements as they age or when they are diagnosed with disease. As a result, death is likely due to old age or disease, not dietary supplement use. In fact, the data reveal that self-reported supplement use increased substantially at each study interval: 62.7% in 1986, 75.1% in 1997, and 85.1% in 2004. As the study cohort aged, more women began taking dietary supplements. Furthermore, subjects with cardiovascular disease, diabetes mellitus, or cancer were excluded from the analysis at baseline but not afterwards. Considering that the follow-up period spanned 19 years and the mean age at baseline was 60.1 years, it is likely that disease emerged during the study period. The authors do not state the cause of death; therefore, it is misleading to blame the supplement—rather than a serious chronic disease—for mortality.
3. Recall bias. Self-reported use of dietary supplements is prone to recall error, and such data tend to be selective and inaccurate. Importantly, there is no information on supplement dose (with the exception of calcium and iron), formulation, or frequency provided for the analysis. Daily supplement habits can vary considerably over time, and multivitamin/mineral supplements are not uniformly formulated. Many contain amounts of reactive metals like iron that are not needed as supplements by postmenopausal women (or men).
4. Generalizability issues. The cohort consisted of only white, postmenopausal women with higher than average rates of supplement use. This is not a representative population and the applicability of the findings to premenopausal women, men, other ethnic groups, and the general population is not appropriate.
According to data from the National Health and Nutrition Examination Survey (NHANES), many Americans are not meeting current intake recommendations for several micronutrients. Even when considering micronutrient intake from fortified foods and dietary supplements, an estimated 70% of the U.S. population aged 2 years and older do not meet the estimated average requirement (EAR) for vitamin D, 60% for vitamin E, 45% for magnesium, 38% for calcium, 34% for vitamin A, and 25% for vitamin C. Multivitamin/mineral supplements are a simple, inexpensive, and safe way to help fill these nutritional gaps and improve micronutrient status.
As part of its "Rx for Health," LPI recommends a daily multivitamin/mineral supplement as nutritional insurance to meet micronutrient needs. Some multivitamin/minerals may provide excessive iron or vitamin A, which can have untoward health effects. Therefore, LPI recommends that men and postmenopausal women take a multivitamin/mineral without iron and containing no more than 2,500 International Units (IU), or 750 mcg, of preformed vitamin A (usually labeled vitamin A acetate or vitamin A palmitate) and no more than 2,500 IU of additional vitamin A in the form of beta-carotene. For more information on multivitamin/mineral supplements, see the article in the Micronutrient Information Center.
In summary, it would not be prudent to discontinue the use of all vitamin and mineral supplements based on a single, flawed study. Each micronutrient fulfills one or several specific biological functions in normal metabolism and good health. It is not sensible to discourage the use of micronutrient supplements when so many people in the U.S. and around the world have insufficient dietary intakes of many vitamins and minerals.