Evidence for the vitamin D hypothesis: The NHANES III extended mortality follow-up.

Background And Aims: Emerging evidence suggests that low levels of vitamin D may be an important risk factor for multiple chronic diseases and mortality, but the evidence is mixed. Vitamin D levels are associated with sun exposure, diet, and metabolic status. One potential explanation for the lack of consistent findings amongst various studies is that low vitamin levels may be associated with poor diets or other risk factors that we were not adequately controlled for in different analyses.

Methods: Prospective analysis of adults over the age of 35 in NHANES III data (1988-1994) with 20 year mortality follow-up. Sequential Cox proportional hazard models quartiles of 25OH vitamin D adjusted for age, season, geography, sociodemographic (SD), CVD risk factors (CVD) and nutritional factors (NF) were performed.

Results: Gender, race, diabetes, anti-hypertensive meds, income, taking vitamin D supplements, physical activity, alcohol consumption, region, body mass index, blood pressure, creatinine, albumin, CRP, thyroxine, iron, RBC folate, vitamin A, E, alpha-carotene, and lycopene were all associated with different quartiles of vitamin D and as well as CHD and all-cause mortality and thus are important potential confounders of this relationship. Adjusting for the confounding factors, higher levels of vitamin D demonstrate an inverse relationship with all-cause mortality, but only the top quartile of vitamin D shows an inverse relationship with CHD mortality.

Conclusions: The highest quartile compared to the lowest quartile of 25OH vitamin D levels is inversely associated with CHD and all-cause mortality adjusting for multiple confounders. Whether supplementation of individuals with low vitamin D will result in similar benefits will require a randomized clinical trial.