Dec 19 2013

OK – But Should I Take a Vitamin

In a follow up to my post earlier this week, I have received (in the comments and elsewhere) multiple questions about whether or not it is still reasonable, in some circumstances, to take a multivitamin. This is ultimately an issue of how to apply scientific evidence to specific individuals – something which physicians have to do every day.

What studies of vitamin supplementation have generally found is that taking a multivitamin does not reduce the risk of cancer, reduce the risk of vascular disease, or prolong survival overall. The question is – can we generalize from the populations in these studies to the general population – to you? Perhaps you are in a subgroup that would benefit from vitamins.

Another common question is whether or not it is reasonable to take a vitamin for “nutritional insurance” in case your diet is lacking in one or more vitamins.

The scientific evidence we have is imperfect and incomplete, otherwise we would have a definitive answer to any such questions. Scientific evidence in medicine is always incomplete – we cannot study every possible permutation of physical conditions and interventions.

At this point, however, the evidence is robust and we can make fairly confident statements about some basic conclusions. If you are a member of the general population, which means you have no diseases or conditions which result in a restricted diet, poor absorption of nutrients, or greater than usual nutritional needs, despite many studies there is no evidence of any health benefit from routine supplementation.

This does not rule out a small benefit, or a subgroup benefit, but it does mean that any possible benefit is highly likely to be minimal. Further, the subgroup argument is not helpful, because you won’t know whether or not you are in a subgroup that may benefit (by definition, otherwise you wouldn’t be considered part of the general population. In other words, known subgroups have already been identified and removed.)

You can make the subgroup argument about any failed clinical trial. This generally does not pan out as a way of rescuing a failed treatment. It’s mostly not worth considering (it’s mostly special pleading) unless you have a good reason to suspect that a specific subgroup might benefit, in which case you should study the subgroup.

The nutritional insurance argument also does not fly – this is exactly what the studies have looked at. Taking a vitamin routinely as nutritional insurance is of no benefit.

A number of studies have compared dietary changes to vitamin supplementation for the prevention of cancer or vascular disease. These studies generally find that dietary changes are helpful, while supplements are not.

Often critics of these negative studies argue that the control group had a good diet, and this obscured any benefit from the vitamins. But that is exactly the point – if you have a good diet, you don’t need vitamins. And – if you have a bad diet, taking vitamins does not make up for it. You need to improve your diet.

The evidence shows that, either way, you are better off improving your diet, and taking vitamins does not help.

Yesterday on Science Based Medicine I wrote about a recent analysis showing that just recommending that people eat an apple a day – essentially adding a portion of fruit to their diet (without increasing their overall caloric intake) could save 8,500 lives a year in the UK, about as many as would be saved from taking a statin.

The same basic power and type of clinical and observational studies that show no benefit to vitamins do show that adding a portion of fruits and vegetables to your diet has a measurable health benefit.

Therefore, if you want some nutritional insurance, eat an apple.

What are the known subgroups that could benefit from some supplementation? I cannot give an exhaustive list, as there are many specific conditions, but here are a few common ones:

– Woman who are trying to get pregnant should take prenatal vitamins

– Recovering from surgery or injury

– Any malabsorption disorder

– Older patients may have a decreased ability to absorb vitamin B12, and should have their level checked periodically.

– Anyone on a restrictive diet should be familiar with which nutrients their diet may lack.

– Certain medications may deplete specific vitamins, in which case your physician will likely give you a supplement to take with the medication.

Dieting for weight loss may be a situation where it is reasonable to take a supplement. However, you should not be restricting to the point that you cannot get adequate nutrition through your food. Further, adding lots of vegetables to your diet is a good way to both lose weight and get your vitamins.

What about growing children? Here again it is optimal to encourage a well rounded diet. However, I admit to being on the fence here. If a child is going through a picky phase, and has a narrow diet despite your best efforts, a multivitamin may be reasonable as they are growing.

But – do not fall into the trap of thinking this is adequate and relieves you of the need to encourage a well-rounded diet. That would still be better.

If your child is not having a problem with obesity, then fruit is usually a good solution. My daughters love strawberries and apples, so we make sure there are always plenty of both in the house, and they can eat as much as they want. Strawberries in particular are packed with vitamins. And yes, you have to nag them to eat their vegetables.

You are better off giving them good eating habits for life, rather than just giving them a supplement.


I wish it were not the case. I really do wish that I could stave off disease and prolong life – in my patients, my loved-ones, and myself – simply by taking a supplement. The evidence, unfortunately, just doesn’t show it.

The good news is that, for most people, the best advice is actually quite simple. Eat a well-rounded diet with 5 portions a day of fruits and vegetables and exercise regularly. When it comes to vitamins and supplements – save your money. Spend it on food or a stationary bike.

If you have reason to suspect that your nutritional needs may be special, then consult your physician. It is easy enough to check your vitamin levels.

Also, the one vitamin that is perhaps the exception to the above is vitamin D. Research on this is still preliminary, but vitamin D insufficiency (if not outright deficiency) may be very common. The darker your skin and the further from the equator you live, the greater the chance you may need to supplement your vitamin D, since we get most of it from the sun. The consistent use of sunscreen, while good for preventing skin cancer, also decreases vitamin D levels.

But again – get your levels checked and supplement accordingly. Primary care doctors are doing this routinely now.

43 responses so far