Abstract 10585: Vitamin D and Risk of Cardiovascular and Non-Cardiovascular Mortality: A Prospective Cohort Study and Meta-Analysis
Background: Low circulating vitamin D concentrations have been linked with higher risks of cardiovascular (CVD) mortality, but whether this association is causal is uncertain. We report 13-year data from the Whitehall prospective study of older men, together with a meta-analysis of similar studies, to assess the relevance of circulating concentrations of 25-hydroxyvitamin D (25[OH]D) not only for CVD but also for non-CVD mortality.
Methods and Results: Data were obtained on plasma concentrations of 25(OH)D in 5409 older men (mean baseline age 77 years) and related to cause-specific mortality over a 13 year follow-up period. At baseline, median 25(OH)D concentration was 77 nmol/L (interquartile range 69 to 89 nmol/L). During follow-up, 3215 men died (annual death rate 6.4%), including 1358 (2.7% per year) from vascular causes and 1857 (3.7% per year) from non-vascular causes. After adjustment for age and seasonality, proportionally higher plasma concentrations of 25(OH)D, within the range of values studied, were inversely and approximately log-linearly related to both vascular and non-vascular mortality. After adjustment for age, prior diseases and a range of other vascular risk factors (smoking, blood pressure, blood lipids, alcohol consumption, markers of inflammation and employment grade), a doubling in 25(OH)D concentration, achievable by high dose vitamin D supplements, was associated with a 23% (95% confidence interval [CI] 13-32%) lower CVD mortality rate and a 30% (95% CI 22-37%) lower non-CVD mortality rate. When the results for all-cause mortality were included in a meta-analysis of 18 prospective cohort studies (involving over 11 000 deaths), participants with 25(OH)D concentration in the top quarter of the distributions had, on average, a 32% (95% CI 28-36%) lower death rate compared with people in the lowest quarter.
Conclusions: Higher concentrations of 25(OH)D are consistently associated with lower death rates from both vascular and non-vascular causes. Reliable assessment of the causal relevance of low vitamin D levels to disease requires further research, including large-scale randomized trials using adequate doses of vitamin D to achieve the necessary changes in 25(OH)D levels associated with differences in mortality.
- © 2011 by American Heart Association, Inc.