The inverse association between serum 25-hydroxyvitamin D and mortality may be modified by vitamin A status and use of vitamin A supplements.

BACKGROUND: Low serum 25-hydroxyvitamin D [25(OH)D] levels have been associated with higher risk of many diseases that affect mortality, including cardiovascular disease (CVD) and cancer. The inverse association between serum 25(OH)D and mortality may be modified by excess circulating vitamin A, due to interactions of vitamin A at the level of the vitamin D nuclear receptor. In this prospective cohort study, we investigated whether the association of 25(OH)D with all-cause, cancer, and CVD mortality was modified by circulating vitamin A or preformed vitamin A intake from supplements. METHODS: We analyzed 15,998 adults in the Third National Health and Nutrition Examination Survey (NHANES III), 1988-1994. Mortality data for all-cause (n = 3890), cancer (n = 844), and CVD mortality (n = 1715) were assessed through December 2006. Serum 25(OH)D was measured using a radioimmunoassay kit, vitamin A biomarkers were measured by HPLC, and information on supplement use was obtained by self-report. Multivariable hazard ratios (HRs) and corresponding 95% confidence intervals (CI) were estimated by proportional hazards regression. RESULTS: Serum 25(OH)D was significantly inversely associated with all-cause mortality (HR 0.93, 95% CI 0.89, 0.97, per 10 ng/mL increase) and also with CVD mortality and mortality due to non-cancer/non-cardiovascular causes, but not with cancer mortality. The observed inverse associations remained statistically significant only among participants with serum retinyl esters <7.0 μg/dL. High intake (>5000 IU/day) of preformed vitamin A from supplements attenuated the inverse association of 25(OH)D with overall mortality. The observed interactions were not statistically significant. CONCLUSIONS: 25(OH)D was inversely associated with overall mortality, CVD mortality, and mortality due to non-cancer/non-CVD causes, but not with cancer mortality. A possible interaction between vitamin A exposure and 25(OH)D concentration appears to be associated with an attenuation of the inverse association between risk of death and quartile of 25(OH)D concentration.