Multivitamins, multiminerals, meta-analysis, and mortality

The study authors used a research method called meta-analysis to combine 21 controlled clinical trials on multivitamins and mortality--risk of death due to any cause. Only trials in which participants took a multivitamin-multimineral supplement every day, and which had a minimum duration of one year, were part of the analysis.

The 21 trials created a total sample of 91,074 adults who took multivitamin-multimineral supplements for an average of 43 months (3.6 years). Participants' average age was 62 years, and 8,794 deaths occurred during the studies. From this large pool of data, the researchers concluded that compared with adults assigned to take multivitamin-multimineral supplements, those who did not take the supplements experienced:

• no increased or decreased risk of all-cause mortality,
• no increased or decreased risk of death due to vascular causes (heart disease and stroke), and
• no increased or decreased risk of death due to cancer.

Considering multivitamins, finding balance

This large, comprehensive study found that death rates were no different for older adults taking multivitamin-multiminerals compared with adults not taking these supplements. Does this mean you should ditch your multivitamins? Not necessarily. The researchers also noted a trend toward decreased mortality in the supplement group. This finding isn't statistically significant, but still, it suggests that contrary to previous observational studies, multivitamins may offer benefit. And it's reassuring to note that unlike the previous studies, this large meta-analysis did not note any increase in mortality among supplement users. Our tips can help decode the multivitamin puzzle as it applies to you and your family:

• Account for age. The study focused on older adults starting supplements later in life. The results may not apply to younger adults or children.
• Consider duration. Vascular disease, cancer, and other chronic diseases develop over decades; the short study timeframes--an average 3.6 years--may not be long enough to accurately capture the relationship between multivitamin-multimineral supplements and mortality.
• Underlying health. This study focused on generally healthy adults. Multivitamin-multimineral supplements may benefit people who have a poor diet or compromised nutritional status for other reasons, such as an existing illness or inability to consume a healthy, varied diet.
• Supplement smartly. According to lead study author Dr. Helen MacPherson, PhD, "As most commercially available multivitamins approximate the recommended daily value, excessive intake may be more likely in those who use multiple dietary supplements than in those who take [only] a daily multivitamin." To minimize the risk of overdoing it, read labels carefully, and avoid loading up on multiple supplements with the same nutrients.
• Keep skepticism intact. The study authors note that previous, highly publicized reports from large observational studies, "have led to considerable concern regarding potential harm associated with multivitamin-multimineral use." This meta-analysis included only controlled clinical trials--the gold standard of evidence-- and it suggests that this level of alarm may be unwarranted. Multivitamins do not appear to increase risk of death in older adults.
• Aim for balance. Many people like to take a multivitamin as nutritional insurance, to fill in the gaps when they are eating a less-than-perfect diet. If you do decide to take a multivitamin, this study offers support that this choice is safe. For optimal benefit, avoid mega-doses of any one nutrient, and select a product that offers about 100% of the daily value for most nutrients.
• Assess personal needs. The study results don't address specific health needs. For example, if osteoporosis is your concern, you may want to focus on particular nutrients, such as vitamins D and K, calcium, and magnesium. Talk to your doctor or a registered dietitian to determine what you need to stay in tip-top health.

(Am J Clin Nutr 2013;97:437-44)