About half of U.S. adults take some sort of nutritional supplement. (My wife and I are in the other half.) Many supplements have a crucial specific function (vitamin C, for example, is the best and only treatment for scurvy) but their benefits to the general population are unproven.

Vitamin D is frequently prescribed to prevent and treat osteoporosis. It is critical in calcium metabolism and in bone health. Some observational studies have found lower rates of death from cancer and cardiovascular disease in regions with greater sun exposure than in areas with less sun exposure, leading some to speculate that vitamin D protects from cancer and heart disease.

Fish oil is frequently prescribed to decrease triglycerides. And in people with high triglycerides who are at high risk for a heart attack and are already taking a statin, there is some evidence that fish oil prevents heart attacks. As with vitamin D, some observational studies have associated fish oil with protection from cancer and heart disease.

(See this previous post for a quick explanation on what an observational study is and what a randomized study is.)

In November, the New England Journal of Medicine (NEJM) published the results of a study that attempted to find a definitive link between vitamin D use, fish oil use, and the incidence of cancer and heart disease. The study enrolled over 25,000 volunteers who were men 50 years of age or older, or women 55 or older. They were randomly assigned to four groups. One group was given vitamin D 2,000 units and omega 3 fatty acids (fish oil) 1 gm daily. The second group took vitamin D and a placebo daily. The third group took fish oil and a placebo daily. And the last group took two placebos daily. They were followed for an average of five years. They were tracked for the development of invasive cancer of any type, heart attack, stroke, and death from cardiovascular causes.

The good news is that none of the groups had any serious side effects, which reinforces our previous knowledge that vitamin D and fish oil are quite safe. The bad news is that there was no difference between groups in incidence of cancer or cardiovascular events, which means neither fish oil nor vitamin D protect the general population any better than placebo.

There are three general lessons here. The first is that many supplements may have very specific effects that benefit specific patients. That shouldn’t lead us to think that they are “good for us” in some vague general way, and that everyone should take them. For example, iron supplements are just the thing for people with iron-deficiency anemia, and maybe for people with restless leg syndrome. But that’s not a reason for everyone to take them. The second lesson is that we don’t really know anything until a randomized trial is done. Observational trial results generate hunches, not conclusions. That’s why we know very little about the effects of diet on health. Almost all dietary studies are observational. So people pontificate about foods that decrease inflammation, or foods that protect from cancer, or foods that make you taller and richer, but until someone randomizes thousands of volunteers to test it, we shouldn’t believe them.

The third lesson, and the hardest one, is another reminder that null hypothesis is usually right, and in the absence of convincing data we should always assume that any two things have no effect on each other.

Learn more:
Vitamin D And Fish Oil Supplements Mostly Disappoint In Long-Awaited Research Results (NPR Shots)
Fish Oil and Vitamin D Supplements May Not Help Prevent Heart Attacks and Cancer, Study Says (Time)
Vitamin D Supplements and Prevention of Cancer and Cardiovascular Disease (NEJM article, abstract available without subscription)
Marine n−3 Fatty Acids and Prevention of Cardiovascular Disease and Cancer (NEJM article, abstract available without subscription)
VITAL Signs for Dietary Supplementation to Prevent Cancer and Heart Disease (NEJM editorial, by subscription only)

My prior post explaining the difference between observational and randomized studies:
Without A Randomized Study Your Results May Vary