Posted on 01/06/2014 at 02:18:03 PM by Student BloggerBy Colby Vorland
Several weeks ago an editorial in the Annals of Internal Medicine declared the last word on multivitamins after new trials failed to demonstrate benefit. "Enough is enough," the authors wrote. “[W]e believe that the case is closed- supplementing the diet of well-nourished adults with (most) mineral or vitamin supplements has no clear benefit and might even be harmful.” A review in Advances in Nutrition also looked at recent developments in multivitamin/mineral research. It is true that many studies have demonstrated no benefit- a few even suggest a slight risk of harm from supplementing specific vitamins or minerals. However there is a nuance missing from such strong viewpoints worth highlighting.
The authors reiterate several times that the population is well nourished. “This message is especially true for the general population with no clear evidence of micronutrient deficiencies, who represent most supplement users in the United States and in other countries (9).” That reference is to an NCHS Data Brief that showed “no clear evidence of micronutrient deficiencies” though it is of course true that micronutrient deficiency symptoms are not common in the U.S. But are we truly a well-nourished population?
To add to the discussion, I graphed the proportion of the population that falls short of the Dietary Reference Intake (DRI) for each of 21 micronutrients for four groups: males and females age 19-30, and males and females over 70. This is 2009-2010 NHANES data. The DRIs are set by committees named by the Institute of Medicine to review literature on each nutrient in its relation to various disease endpoints (or to achieve balance). If sufficient research is available, an Estimated Average Requirements (EAR) is set, a level that is estimated to meet the needs of 50% of the population. If an EAR can be established, a Recommended Dietary Allowance (RDA) can be set, estimated to cover the needs of 97-98% of the population. If not, an Adequate Intake (AI) is set.
It is clear that a majority of Americans fall short of DRIs for many nutrients. Below that graph is the same data assuming everyone in the population took a Centrum Adults multivitamin (this brand is used is several trials). Much of the population now achieves recommended intakes where they didn't before.
Further, I calculated that only 0.5% of Americans in these age groups achieve all 21 DRIs, and the average person achieves about 11.
So with such strong plausibility that multivitamins would mitigate disease, why are trials mostly showing no effect?
It helps to first understand what endpoints are used when setting DRIs. Here are a few examples of what research has been considered when setting the DRI (just picking some from each):
● Vitamin E: hydrogen peroxide-induced hemolysis, LDL-oxidation (cardiovascular disease), diabetes, diabetic neuropathy, cancer (prostate), immune function, cataracts, Parkinson's disease, Alzheimer's disease
● Folate: Vascular disease (homocysteine), megaloblastic anemia
● Iron: anemia, reduced physical work capacity, delayed psychomotor development in infants, impaired cognitive function
It should be noted that for some nutrients, like zinc, we simply don't have sufficient research yet to set DRIs based on specific disease endpoints, so they are set largely by rate of excretion (to achieve balance). It is likely that many nutrient-disease relationships simply remain undiscovered. It should also be noted that many nutrients (like zinc as well) are cofactors in many enzymes important for many critical biological pathways, which is why this is plausible.
The takeaway here is that for each nutrient, there are multiple disease/endpoints that influence the DRI.
In contrast, multivitamin (and individual nutrient) trials so far have focused mainly on cardiovascular disease and cancer. Here are disease endpoints specific to multivitamin trials (3 or more micronutrients): cardiovascular events, cognitive function, cancer/CVD/mortality in women, acute respiratory tract infections. Contrary to what the editorial proclaims, there are some trials that have suggested multivitamins do have positive effects: (small) reduced cancer risk in men, reduced cancer and all-cause mortality in men (but not women), cancer and total mortality (in an undernourished population), reduced age-related macular degeneration, reduced cataract in men (but not AMD), cataract progression, reduced risk of immune decline and morbidity in HIV.
And that isn't breaking down significant variation in trial differences by primary vs secondary prevention, issues such as trial length (cancer and CVD are long latency diseases after all), genetics (e.g. MTHFR/folate and riboflavin), age, etc. For instance, two different epidemiological cohorts suggested that it takes over 10-15 years of supplementation to reduce colon cancer risk. Most clinical trials don't do this long, and it isn't practical to do many trials to test against all plausible outcomes.
Of course, it is ideal to get nutrients from food. But as an example, Centrum Adults is just over 7 cents per day, an inexpensive and simple intervention, though we do need more research on possible behavioral licensing effects. It is also true that the supplement industry is an unregulated mess and there are safety issues and many claims for micronutrients that go way beyond the data. But this is a separate issue, and it is important not to let that influence our interpretation of research. Given the significant population below recommended micronutrient intakes, I think it is very premature to close the door on multivitamins.