Vitamin D


Cauley JA reported an analysis from the Women's Health Initiative that found a higher risk of fracture in Black women whose serum vitamin D level was higher than 17 ng/ml. This is very different from White women who had a higher risk of fracture with low vitamin D. This suggests that recommendations for vitamin D should be different in White and Black women. Black women with vitamin D between 20 and 30 ng/ml had 48% more fractures than those with levels below 20.

In November 2010 the Institute of Medicine released their consensus report, which is very similar to the recommendations that were already on this web page. This long and detailed report is available on-line. Dietary Reference Intakes for Calcium and Vitamin D. I have updated some of my recommendations based on this report. There are still many unanswered questions about vitamin D, and many clinical trials are in progress. Details are at One large study is sponsored by the NIH and will enroll 20,000 subjects. Here is a link to their study website: VITAL study.

Vitamin D requirements

There are three sources of vitamin D: natural sunlight, fortification of dietary foods, particularly dairy products and some cereals and oily fish. The radiation that converts vitamin D in the skin is the same wavelength that causes sunburn, so careful application of sunscreen can inhibit vitamin D production. At northern latitudes, there is not enough radiation to convert vitamin D, especially during the winter. After the age of 70 the skin does not convert vitamin D effectively (MacLaughlin J). Vitamin D levels depend on the pigmentation in the skin (darker skinned people have lower vitamin D levels), the altitude (higher levels in the mountains), and body weight (obese people have lower levels than thin people).

Vitamin D supplementation
Children and adults with poor sunlight exposure 600-1000 units/day
Adults older than 70 800-1000 units/day
Patients with cystic fibrosis 800-1000 units/day
Patients with malabsorption Up to 50,000 units/day, check levels
Patients with liver disease May need active metabolites
Patients with kidney disease Need active metabolites
Patients with sarcoidosis
or kidney stones
Be careful not to give excess
Check levels, don't exceed 30 ng/mL

The new report lists the upper limit of safety at 4,000 units/day. However, this is based on limited evidence. Studies of high doses have lasted only a few months.

Many vitamin D supplements also contain high contents of vitamin A - and recent studies show that vitamin A can increase bone resorption. The labels do not have to list the vitamin A, but it is often there, especially if it comes from cod liver oil.

Cholecalciferol is preferred because it sustains blood levels for a longer time (Armas, Glendenning P).

Vitamin D levels in serum

25 (OH) D Levelng/ml
(used in USA)
Deficientless than 8less than 20
Toxicgreater than 90greater than 225

Binkley N, Low Vitamin D status despite abundant sun exposure.
J Clin Endocrinol Metab 92:2130, 2007. copyright 2007, The Endocrine Society

The above graph shows results from 93 healthy 24-year-old Hawaiian surfers. The authors of this paper concluded that: "the goal of vitamin D replacement therapy should be no greater than the maximum that appears attainable, a serum 25(OH)D concentration of approximately 60 ng/ml." However, many laboratories currently have listed their "sufficient" range as 32-150 ng/ml. There is no evidence that levels above 50 ng/ml are beneficial.

There is disagreement about whether the lower limit of the optimal vitamin D levels should be 20 or 30 ng/ml, but none of the experts in the field still think that levels lower than 20 ng/ml are desirable. This is really different from ten years ago, when levels above 15 ng/dl were considered sufficient. The recent Institute of Medicine Report finds that levels above 20 are adequate.

In another paper, Binkley N found differences among the laboratory tests for 25(OH) vitamin D of around 5 ng/dL, and approximately 10% error within the same lab. These errors must be taken into account when interpreting the vitamin D levels.

Vitamin D metabolism

Here is a little Flash slide show about vitamin D formation and actions:

Measure 25(OH) vitamin D, NOT 1,25(OH)2 vitamin D

1,25(OH)2 vitamin D [1,25-D] is more difficult and expensive to measure than 25(OH)D; moreover, it is not a good measure of vitamin D status. When patients are vitamin D deficient, the parathyroid hormone increases and drives the renal 1-alpha-hydroxylase, so that 1,25-D levels increase. Only in severe deficiency, when substrate is depleted, does the 1,25-D become low. Partially treated vitamin D deficiency also results in marked elevations of 1,25-D levels.

Some doctors, thinking they are sophisticated because they know that 1,25-D is more active, order the wrong measurement. Do not fall into this trap and waste money on this expensive but often misleading test! There are only a few situations where you would actually want to know the 1,25-D:
    unexplained hypercalcemia (looking for granulomatous disease or lymphoma),
    suspected genetic childhood rickets,
    suspected tumor-induced osteomalacia,
    some cases of nephrolithiasis or hypercalciuria.

Patients with stages 4-5 chronic kidney disease have decreased 1,25-D levels but even in those patients the 25(OH) vitamin D is a better test of the stores of vitamin D, and the PTH is a better indicator of mineral abnormalities.

Finally, some cells generate intracellular 1,25-D to fight tuberculosis, suppress cancer growth, or modify immune response. They need adequate substrate, as measured by 25(OH) vitamin D.

Vitamin D and Mortality

Melamed ML followed 13,331 adults in the US for 8.7 years to see if the vitamin D level was related to mortality. Notice that the mortality increases with either high or low levels of vitamin D (just like some of the other steroid hormones).
copyright American Medical Association, used with permission

Michaelsson K found that in elderly men the mortality rates were increased at both high and low vitamin D levels, and the lowest mortality was seen with vitamin D between 24 and 34 ng/mL.

Another study (Dobnig H) from Germany found lower mortality in people in the top 25% of vitamin D levels, compared to those with the lower 25%. The overall levels, however, were quite low; almost everybody was lower than 33 ng/ml.

A meta-analysis of clinical trials of vitamin D (average dose was about 500 units/day) found an overall reduction in mortality of 7% (Autier P).

Click to enlarge this Forest Plot of mortality from randomized trials of vitamin D. Overall there was a non-significant decrease of 7% in the vitamin D groups compared to placebo (Chung) This report was sponsored by the US Agency for Healthcare Research and Quality.

Vitamin D and fractures

This table shows overall results of studies of vitamin D and fracture rates. Most of the studies did not find a signficant benefit of vitamin D.

Click on the table to download an Excel file with more details.

Recently there have been several meta-analyses of this topic, which show different points of view. After reading these studies and analyses, I conclude that vitamin D therapy can reduce fractures in frail elderly patients, especially when they have vitamin D insuffiency, and that adequate calcium intake is also needed. Otherwise, the difference in fracture rates is really quite modest in the vitamin D clinical trials.

Vitamin D and cancer

Byers T. Anticancer vitamins du Jour--The ABCED's so far. Am J Epidemiol 2010;172(1):1-3. The July issue of this journal was devoted to careful analysis of the relationship between vitamin D and various cancers. This editorial should be read by everybody who has been advocating huge doses of vitamin D.

The Vitamin D Pooling Project included 10 cohorts around the world, and studied between 500 and 1300 cases of less common cancers: endometrial, kidney, lymphoma, ovarian, upper GI, and pancreatic. There was no benefit of higher vitamin D in any of these cancers. Of concern, however, was an increased risk of pancreatic cancer when the serum vitamin D level was greater than 40 ng/mL (100 nMol/L).

Vitamin D can suppress growth of cancer cells grown in culture. Therefore, researchers hoped that vitamin D could be used to prevent or treat cancer. Early studies were promising, but recent larger studies do not find the same encouraging results. Click to see recent large prospective observational studies relating vitamin D to colon, breast, or prostate cancer:

Colon cancer

Vitamin D probably protects against colon cancer (Wei MY, Giovannucci E,Freedman DM). Click to enlarge this Forest Plot of colon cancer from studies of vitamin D. (Chung) This report was sponsored by the US Agency for Healthcare Research and Quality. Since then, a large nested case-control study in Europe found an increased risk in men with low vitamin D (Jenab M; results are shown on the image above). Another study looking at different ethnic groups also found lower risk of those in the lowest quintile of vitamin D (less than 16.8 ng/ml), but no difference in the other 80% of patients (Woolcott CG).

Breast Cancer

(Abbas S, Agborsangaya CB, Almquist M, Bertone-Johnson ER, Crew KD, Engel P, Freedman DM, Crew KD, McCullough ML, Rejnmark L)

In the Women's Health Initiative randomized clinical trial of calcium and vitamin D, which included 36,000 women, there was no difference in the incidence of breast cancer between those assigned to placebo or to calcium and vitamin D (Chlebowski RT). Click on the thumbnail to enlarge the graph from that study. However, if the intake of vitamin D included the supplements taken by the subjects, then those with the highest intakes had higher risk of breast cancer.

Prostate Cancer

Early studies suggested the men with low vitamin D levels were more likely to get prostate cancer, but newer, larger studies do not find this. Click to enlarge this Forest Plot of prostate cancer from studies of vitamin D. (Chung) This report was sponsored by the US Agency for Healthcare Research and Quality. In the large Prostate, Lung, Colorectal and ovarian cancer screening trial (PLCO), the risk of aggressive prostate cancer was higher in men with higher vitamin D levels, as shown in the colored graph above (Ahn J). Two newer studies within the last year continue to show no significant benefit of vitamin D - one was a case control study of men with different ethnicities, including 329 cases (Park SY) and the other a large nested case control study from the European Prospective Investigation into Cancer and Nutrition (EPIC) study, which included 652 cases ( Travis RC).

More about vitamin D.

Updated 8/30/11