Vitamin D levels and cancer incidence in 217,244 individuals from primary health care in Denmark.

Vitamin D has been linked to cancer development in both pre-clinical and epidemiological studies. Our study examines the association between serum levels of vitamin D and cancer incidence in the Capital Region of Denmark. Individuals who had vitamin D analyzed at The Copenhagen General Practitioners Laboratory between April 2004 and January 2010 were linked to Danish registries with end of follow-up date at Dec 31st 2014, excluding individuals with pre-existing cancer. Cox regression models adjusted for age in one-year intervals, sex, month of sampling, and Charlson Comorbidity Index were applied. The study population of 217,244 individuals had a median vitamin D level of 46 nmol/L (IQR 27-67 nmol/L). Non-melanoma skin cancer was the most frequent form of cancer, followed by breast-, lung-, and prostate cancers. No associations were found between increments of 10 nmol/L vitamin D and incidence of breast, colorectal, urinary, ovary or corpus uteri cancer. However, higher levels of vitamin D were associated with higher incidence of non-melanoma (HR 1.09 [1.09-1.1]) and melanoma skin cancer (HR 1.1 [1.08-1.13]) as well as prostate (HR 1.05 [1.03-1.07]) and hematological cancers (HR 1.03 [1.01-1.06]), but with lower incidence of lung cancer (HR 0.95 [0.93-0.97]). In our study, vitamin D levels are not associated with the incidence of several major cancer types, but higher levels are significantly associated with a higher incidence of skin, prostate, and hematological cancers as well as a lower incidence of lung cancer. These results do not support an overall protective effect against cancer by vitamin D.