By Perry Holman, Executive Director, Vitamin D Society

There has been a lot written this past month about the new study results from a large vitamin D randomized placebo-controlled trial, the VITAL study, by Manson et al. which was published in the New England Journal of Medicine. Media coverage has been particularly confusing for the public with some articles proclaiming the benefits of vitamin D while others reporting vitamin D does not reduce CVD or cancer and still others taking a more middle ground. For example:

 CBC News

Vitamin D lowers risk of cancer death, fish oil reduces odds of heart attack, study shows

 ABC News

STUDY: Fish Oil And Vitamin D Pills No Guard Against Cancer Or Serious Heart Trouble

 Washington Post

Big studies give mixed news on fish oil, vitamin D

Confusing! Which one is correct? All perhaps, to some degree. The Vitamin D Society will try to bring clarity to this issue.

The VITAL study is the largest ever randomized, placebo-controlled trial of vitamin D, and took place over 5.3 years. It was funded by the National Institutes of Health in the USA. It recruited 25,871 participants, men over age 50 and women over age 55, with a mean age of 67, in the United States. The participants were randomized and split into 4 equal groups. The groups received either 1) vitamin D 2000 IU/d and 1 g of omega-3; 2) vitamin D 2000 IU/d; 3) 1 g of omega-3; 4) placebo. For vitamin D reporting purposes the study authors combined groups 1) and 2) and compared to groups 3) and 4).

The study reported:

- 793 incidents of cancer in the vitamin D group vs 824 in the placebo group.

- when split by race, dark skinned participants in the vitamin D group had a 22% drop in cancer with 98 incidents vs 126 for the placebo group

- 154 cancer deaths in the vitamin D group vs 187 in the placebo group (a 27% reduction)

- in a sub-analysis excluding  cancer deaths from the first 2 years based on the assumption that these cancers would have probably been present within the participant at the start of the study, they found a statistically significant 25% reduction in cancer deaths from the vitamin D group (112 vs 149)

- major cardiovascular events took place in 396 people from the vitamin D group and 409 from the placebo group

- no excess risks of hypercalcemia or other adverse events were identified with the dose of 2000 IU/d

The study concluded: “Supplementation with vitamin D did not result in lower incidence of invasive cancer or cardiovascular events than placebo.”

Note: Even though the vitamin D group showed less incidents and deaths by actual number the difference was not large enough to be statistically significant in scientific terms and therefore could be due to chance. This is why the study reports that vitamin D did not result in lower incidence even though it had overall less reported incidents of cancer and cardiovascular events as shown above.

We noticed some serious limitations in the VITAL study design and execution that would affect the results:

(The study provides 25(OH)D measurements in the US standard of ng/ml. We have converted them to international standards of nmol/L (multiply by 2.5).)

- Vitamin D testing was limited and incomplete.

At baseline only 15,787 (61%) participants were tested out of a total of 25,871 in the VITAL study. In year 1, a subgroup of 1,644 (6.4%) participants were re-tested. For all participants in years 2-5 no further vitamin D testing was completed. The study failed to show and report the 25(OH)D levels for each group by year? It also failed to test and report the 25(OH)D levels for everyone who had a targeted disease outcome. Research has shown that providing the same vitamin D dose to similar people can produce wide ranging and very different 25(OH)D levels. This is why it is important to analyze and present results of vitamin D studies by 25(OH)D levels.

- Starting vitamin D blood levels or 25(OH)D levels for participants entering the study were very high.

“Among the 15,787 participants who had blood samples that could be analyzed, the mean serum total 25-hydroxyvitamin D level at baseline was 30.8 ng/ml (77 nmol/L).” A starting vitamin D level this high would eliminate a large portion of the potential disease prevention effect of taking someone with sub-optimal vitamin D blood levels to optimal. In addition, the starting level was significantly higher than the target level recommended by the Institute of Medicine in 2010 of 20 ng/ml (50 nmol/L). A recent US study reported that 60% of people are below 30 ng/ml (75 nmol/L) compared to the 50% found in the VITAL study. In Canada 77% of people are below 75 nmol/L. Therefore study participants had a very high starting vitamin D level which was not representative of the current population.

- The number of participants starting the VITAL study with deficient 25(OH)D blood levels of under 20 ng/ml (50 nmol/L) were low, at 12.7%.

Other US population studies such as Kroll 2015 have found that 33% of Americans have 25(OH)D blood levels below 20 ng/ml (50 nmol/L). Again, having very few people starting the study below 20 ng/ml (50 nmol/L) would significantly reduce the potential positive results of increasing vitamin D blood levels through supplementation.

- The vitamin D dosage given, 2000 IU/day, was relatively low and only increased mean 25(OH)D blood levels by 12 ng/ml (from 29.8-41.8 ng/ml) or 30 nmol/L (75-105 nmol/L).

For adults, studies have shown that a vitamin D intake of 4000 IU/d and greater is required to maintain a 25(OH)D blood level of between 40-60 ng/ml (100-150 nmol/L) throughout the year.

- The placebo group, and all groups in the VITAL study were allowed to consume up to an additional 800 IU/d of vitamin D supplement per day.

This was due to ethical considerations. It is not ethical to withhold the recommended daily amount of vitamin D from participants. This is one key reason that randomized controlled trials of vitamin D tend to show null results. In this case 800 IU is 40% of the tested intervention of 2000 IU/d. And 43% of participants (all groups including placebo) were already supplementing with vitamin D during the study. In addition, after 5 years, 10.8% of the placebo group were taking greater than allowed 800 IU/d.

We asked our expert vitamin D scientific advisors for their thoughts on this research paper and this is what they had to say.

“VITAL was not designed to evaluate the vitamin D-cancer hypothesis. Rather it was designed to see whether giving a moderate (not high) dose of vitamin D₃ to healthy people could reduce the risk of cancer and cardiovascular disease,” says Dr. William Grant, Director Sunlight, Nutrition and Health Research Center. “While the results for the entire group failed to find a significant benefit, reductions in cancer mortality rates for all and cancer incidence for blacks and those with low BMI were found.”

Dr. Susan Whiting, Professor of Nutrition and Dietetics, College of Pharmacy and Nutrition, University of Saskatchewan, said “The study reported in the New England Journal of Medicine by Manson et al is a study displaying what can go wrong with a randomized trial of vitamin D efficacy conducted in the USA.  At baseline, the average vitamin D status (as measured using 25-hydroxyvitamin D metabolite levels in blood) of the subjects enrolled was above 30 ng/ml (75 nmol/L), a level already meeting what will reduce risk of adverse health effects. In other words, half of the study population was already getting enough vitamin D, thus negating the planned comparison of high versus low vitamin D.” 

In the end, looking objectively, was VITAL a fair test to determine the cancer and cardiovascular health effects of vitamin D? In my opinion, for the points raised above, I believe it was not.

Dr. Robert Heaney wrote a research paper recommending guidelines for optimizing design and analysis of clinical studies and nutrient effects. It highlighted the fact that the baseline starting status of the nutrient must be low on the response curve and the intervention must be large enough to span much or all of the response region. This was not true for the VITAL study. 

Should nutrients such as vitamin D be evaluated on a randomized controlled trial (RCT) study basis that was developed for drug trials?  And is it fair that the placebo group can have access to the nutrient being tested? This is further complicated for vitamin D because not only could placebo participants take 800 IU/d of vitamin D supplement, they could also make vitamin D naturally through their skin from sun exposure which was not measured and adjusted for. How do you control or limit this effect?

Would a drug manufacturer doing a drug trial allow for the placebo group to take 40% of the dose of the drug being tested? Of course not. That is ridiculous. That would negate the effect and results for the drug.

Another thing to consider when looking at the results of vitamin D studies - is there a difference in how the vitamin D is received by the body? Is there a difference between supplementation or sunlight? The VITAL study was a supplementation based study. Would the health effects have been different if the vitamin D was formed naturally from UVB sunshine exposure? Does vitamin D from sunlight have some additional health benefits over vitamin D supplements?

Naturally occurring vitamin D is produced in your skin from UVB exposure. This is how the human body evolved to receive vitamin D. Dr. Michael F. Holick in his book - The Vitamin D Solution, said “if the body could say which method it prefers to get its daily dose of vitamin D, it would hands down give a standing ovation to sunlight sources of vitamin D rather than a bottle.” But trying to set-up a study on vitamin D received exclusively from sunshine would be a lot harder than providing a participant a vitamin D supplement. And again how do you limit the placebo group from receiving sunshine? Put people in a submarine? There would be ethical issues.

Vitamin D is a powerful steroid hormone made by your body when UVB in sunlight hits the skin. It was mis-named a vitamin. Vitamins are organic compounds that cannot be made by the body. Hormones are very powerful and control most bodily functions

What this study is really saying is that if you start with a fairly high vitamin D blood level of 31 ng/ml (77 nmol/L) and if you take a vitamin D supplement of 2000 IU/d more than you are currently taking, there is a slight benefit in reduced cancer events and cardiovascular events vs someone taking vitamin D supplements of 800 IU/d or less, but it’s not scientifically significant.  In addition, death from cancer would be reduced by 25%.

With the limitations that were imposed by this study its surprising that any positive results were found. Bottom line – VITAL did not give vitamin D the fair opportunity to show what it can do for your health.

What should you do? Vitamin D is a powerful hormone that tells your cells what to do. Vitamin D has been shown to have a positive effect in reducing the risk in numerous serious diseases from thousands of research studies. The recommendations from an expert group of 48 vitamin D researchers, doctors and scientists in their Scientists Call to D*action recommend that everyone, all ages, maintain an optimal vitamin D levels of between 40-60 ng/ml or 100-150 nmol/L. These scientists have nothing financially to gain from you. The only way you will know if you have optimal vitamin D blood levels is to have it tested. Vitamin D is cheapest health insurance you could ever have. You can choose what source is right for you. Moderate levels of non-burning sunshine in the warmer months, artificial UVB sources through the winter or a vitamin D supplement of up to 4000 IU/d. You decide. Its your health.