Vitamin D and cancer

The current New England Journal of Medicine contains an article by Manson, Mayne and Clinton on Vitamin D and cancer prevention. They all served on the Institute of Medicine Committee which recently raised target levels for Vitamin D for the prevention of bony disease. The committee concluded that for outcomes beyond bone health, including cancer prevention, the evidence available is inconsistent and inconclusive. No cause-effect relationship can be established. There has been no randomized controlled trial conducted where cancer has been the primary end point. Most evidence is derived from laboratory studies, ecological correlations, and observational studies. Low vitamin D levels are also linked with confounding factors that are themselves associated with high risk of cancer; for example: obesity, lack of physical exercise, dark skin pigmentation and diet.

Reverse causation may also be a problem - poor health can lead to less sun-exposure or a poorer diet. Vitamin D is like many micronutrients that have been linked with cancer such as beta carotene and selenium; randomized clinical trials for these have been uniformly negative.

The theory that vitamin D can help prevent cancer is biologically plausible. The vitamin D receptor is widely expressed and test tube studies demonstrate that vitamin D can promote cellular differentiation, inhibits cancer cell proliferation, and has anti-inflammatory, proapoptotic and antiangiogenic properties. This might suggest a role in cancer prevention, but, of course, proves nothing.

There have been three vitamin D trials including one that compared vitamin D + calcium to calcium alone that addressed the occurrence of new cancers or cancer mortality as secondary end points, but the results showed no difference. A trial at Oxford of 2686 individuals 833 U/day showed a relative risk of 1.09 (CI 0.86-3.36), a trial in Nebraska of 1179 postmeopausal women showed a relative risk of 0.76 (CI 0.38-1.55) and a big American study of 36,282 postmenopausal women showed a relative risk of 0.98 (CI 0.91-1.05). One study showed that women with the lowest intakes of Vitamin D had the lowest incidence of breast cancer and those with the highest intakes had the highest risk of cancer and in this study both the figures were statistically significant.

For colorectal cancer, observational studies do support a link. A meta-analysis of 5 studies suggests that patients with serum levels of 25-hydroxy vitamin D of 33 ng/ml or higher had only half the risk of colorectal cancer of those with levels below 12 ng/ml. There have been two studies since; a European prospective study found a similar association, but a Japanese study did not find such an association. Randomized trial evidence is limited. A British study did not find a change in incidence in individuals treated with vitamin D and the WHI trial of vitamin D plus calcium similarly found no reduction in incidence or reduced mortality.

Observational studies for prostate cancer have not supported the idea that vitamin D deficiency is associated with prostate cancer and nested case control studies similarly have been non-supportive of the hypothesis. There has been a large scale Vitamin D pooling project for rarer cancers. These show no association between higher levels of vitamin D and reduced risk of endometrial, esophageal, gastric, pancreatic, or ovarian cancer or NHL. Indeed there are suggestions that higher vitamin D levels might be associated with an increased risk of pancreatic cancer.

The Committee concluded that despite widespread enthusiasm for the idea, the evidence is inconclusive and inconsistent. There are new trials assessing moderate-to-high vitamin D supplementation and these should give us an answer in 5-6 years time.