Negative studies have piled up, but are they a fair test?
Good news about vitamins is hard to find these days. Headline after headline — including a few in this newsletter — blare negative results. B vitamins don't prevent heart attacks. Vitamin E doesn't benefit people with Alzheimer's disease. Vitamins A, C, and E — no cancer protection there.
Lack of benefit is one thing. But mixed in with the null findings are a few that suggest large doses of vitamins might cause some real harm. One of the first came in 1994, when results from a high-profile Finnish study linked beta carotene, a form of vitamin A, to an increased risk of lung cancer in male smokers. Several years later, high intake of retinol — another form of vitamin A — was linked to hip fractures.
More recently, folic acid has come under some suspicion. Studies have identified a possible connection between high intake of this B vitamin and increased risk of cancer, particularly colon cancer. And because flour and other grain products are fortified with folic acid, most of us are consuming more folic acid than we realize.
Taking large, "megadoses" of individual vitamins has been a dubious proposition for quite some time (although exactly what constitutes a megadose is ill-defined). There are, though, at least two notable exceptions. First, pregnant women — indeed, all women of childbearing age — should get extra folic acid (400 micrograms daily) to protect against birth defects. Second, there's growing evidence that most Americans — particularly those in the northern two-thirds of the United States — would benefit from getting between 800 international units (IU) and 1,000 IU of vitamin D daily, which is twice as much as the official recommendation.
What is new is that the multivitamin has become a harder call these days. For years, even the most mainstream of doctors have been comfortable recommending them. They're not hugely expensive: taking a Centrum Silver every day for a year costs about $30. They're an easy way to fill in the nutritional gaps left by our less-than-ideal eating patterns. And where we don't have gaps, they supply modest surpluses — a nice hedge just in case extra amounts of some vitamins (and minerals) do turn out to have health benefits.
But as the negative findings of studies of vitamins have piled up, there's been some pulling back. In 2006, a conference of experts convened by the National Institutes of Health came to the conclusion that there was insufficient evidence to recommend multivitamins. And in 2009, results from a large study of postmenopausal women were published that showed multivitamin use had little to no effect on cancer and cardiovascular risk.
Counterarguments come from Dr. Walter C. Willett, chair of the Harvard School of Public Health's nutrition department and a member of the Health Letter's editorial board. Dr. Willett says the problem isn't so much with vitamins, but the way they've been studied. The randomized clinical trial, considered the gold standard in medical research, has some limitations when it comes to long-term prevention efforts like taking vitamins, he says. (More on that below.) Dr. Willett also sees more nuance, a mix of good and bad news, than headlines allow. He's still inclined to give daily multivitamins the benefit of the doubt and recommends that people take them — and look for brands that contain lots of vitamin D.
Vitamins are organic (carbon-containing) compounds that we must ingest in small amounts for our bodies to function properly. That distinguishes them from carbohydrate, fat, and protein, which are needed in relatively large amounts, and minerals, which aren't organic molecules. Thirteen vitamins have been isolated and identified: eight different B vitamins and vitamins A, C, D, E, and K.
Historically, the problem with vitamins was that people didn't get enough of them. The vitamin-deficiency diseases have colorful names like beriberi (lack of vitamin B1), rickets (lack of vitamin D), and scurvy (lack of vitamin C).
But as the food supply became more reliable and diets improved, the focus shifted from absence to addition and whether consuming vitamins in abundance might fend off disease and perhaps even prolong life. Test-tube and animal experiments showed that vitamins A (in its beta carotene form), C, and E were potent antioxidants that could "soak up" unstable molecules capable of altering fats and proteins and damaging DNA — the kind of damage that eventually manifests itself as heart disease and cancer. In several cases, the antioxidant findings were buttressed by results from long-term studies like the Harvard-based Nurses' Health Study. Linus Pauling, the two-time Nobel Prize winner, brought star power to the cause with his ardent promotion of vitamin C as a cure-all. In 1994, the Dietary Supplement and Health Education Act (DSHEA) classified vitamins (and herbal medicines) as dietary supplements so they didn't need to meet the same safety and efficacy standards as drugs. Perhaps the demand was already there, but DSHEA took the lid off and allowed vitamins and herbal supplements to become a big business.
...and the bust
But amid all the vitamin mania, researchers were quietly running vitamins through the gantlet of randomized trials. Hundreds, if not thousands, of results have now been published. Not all the results are negative by any means, but there's a persuasive tilt in that direction.
Vitamin C was first to travel from vaunted to vanquished. Pauling's assertion that taking large doses (500 mg or more daily) would prevent colds was undercut by results from several studies in the 1970s. Later, a Mayo Clinic trial upended claims that the vitamin could be used to treat colon cancer. In late 2008, Harvard researchers added another finding to the list that vitamin C does little, if anything, to prevent heart attacks and strokes.
Vitamin E looked promising because of its strong antioxidant effects — and doctors believed it, too: a 1999 survey found that almost 40% of cardiologists took vitamin E. But randomized trial results haven't been kind to E, either, and 2005 was a particularly rough year. Alzheimer's disease, heart disease, cancer — reports that year made the vitamin out as ineffectual against all three.
And it's been a bumpy ride for the B vitamins, too. The hope was that a trio of them — B6, B12, and folic acid — might be an easy way to rake back heart disease risk because they lower levels of homocysteine, an amino acid in the blood. In randomized trials, they've reduced homocysteine levels, but that reduction hasn't translated into lower heart disease risk. An exception to the gloom: a major meta-analysis found that folic acid supplements do reduce the risk of stroke.
What doctors do
The Health Letter editorial board is split on multivitamins. Dr. Walter C. Willett takes them and says we should too, as a precaution against poor nutrition and as a way to boost our vitamin D intake. The newsletter's other nutrition expert, Dr. Bruce Bistrian, doesn't and depends on a good diet and lots of fish (five servings a week) to protect his health, although he believes others might benefit from multivitamins.
When we surveyed the other doctors on our editorial board, it was close: eight takers, seven nontakers. One doctor said he took them occasionally. Another is in a blinded study, and he doesn't know if the pill he is taking is the multivitamin or the placebo. The multivitamin takers said they took them for the vitamin D content and as an insurance against the holes in their diets. Most of the nontakers said taking multivitamins was unnecessary because they eat a healthful diet with enough fruit and vegetables.
Limits of randomized trials
Much of this bad news for vitamins has come in the form of negative results from randomized trials, which in medical circles is damning evidence indeed. The randomized trial is considered the true test in medicine — the final arbiter of whether or not something works.
In simplified terms, a randomized trial involves enrolling a group of people into a study and randomly assigning them to receive either the intervention under investigation (vitamins in this context, but it can be almost anything) or the control (classically, a placebo pill). Random assignment means letting chance decide which group each person goes into. After a prescribed period of time, researchers look for differences between the intervention and control groups. The power of these trials comes from the fact that randomization — if done carefully — works to make comparison groups alike, so the only explanation for a difference in outcome is the effect of an intervention.
But when it comes to evaluating how well vitamins — and other preventive measures — protect us against chronic diseases like cancer and heart disease, the randomized trial may have some blind spots. For one thing, it may take decades for conditions like cancer or dementia to develop. So even if a trial lasts many years, that might not be long enough to pick up on the effect of an intervention, particularly if it's subtle. In the Harvard-based Physicians' Health Study, beta carotene had no effect on cognitive function after 12 years, but six years later, some modest benefit seemed to have kicked in.
There's also a question about applying narrow randomized trial results to everyone. Strictly speaking, the results of any trial are applicable only to the sorts of people included in the trial, so the negative result from, say, a trial of older men who took a vitamin for five years are relevant only to older men, not to the population as a whole. Randomized trials give answers for slices of the pie, not the entire thing.
Results from randomized trials can also be misleading if the findings for a subset of people differ from the overall ones. For example, in 2005, the high-profile Women's Health Study dealt another setback to vitamin E when it showed no cardiovascular benefits from taking the vitamin. But when the researchers analyzed women ages 65 and older separately, they found that taking the vitamin was associated with a 26% reduction in cardiovascular events. Subgroup analyses can be abused by researchers "data dredging" for provocative results, but they can also show that trial results aren't uniform.
That elusive bottom line
Despite all the back-and-forth, there are a couple of areas of agreement. Added vitamins have lost their sheen, and there are more doubts than ever about taking them in pill form. Large doses of single vitamins aren't a good idea: the benefit is doubtful, and some can cause harm. The notable exceptions are vitamin D, and folic acid for young women.
And what about multivitamins? The doses they contain aren't likely to pose any risk, but the leap of faith that we're getting benefit from them has definitely gotten longer. On the other hand, it's a reasonable choice to take a multivitamin — or not to, if you have a well-rounded diet.