Low 25-Hydroxy Vitamin D Levels and Mortality Among U.S. Adults: Results of NHANES 18-Year Mortality-Linked Data

Introduction: A previous study reported an increase risk of death in patients with low serum 25(OH)D (vitamin D) concentration. Unfortunately, this study was limited by small number of events due to short follow-up time. To gain further insight into those findings, we now extend that work, using additional mortality data to examine the association of all-cause and cause-specific mortality with different quartiles of vitamin D concentrations among U.S. adults aged 20 years and older. Methods: Data from the third national health and nutrition examination (NHANES III) and the NHANES III mortality-linked files were used. To determine the cause of death, the national center for health statistics linked NHANES III participants to the national death index registry through December 31, 2006. Analyses included 15,917 adults. Cox proportional regression models were used to examine mortality across quartiles of vitamin D concentrations. Results: Study subjects were evenly distributed in each quartile of vitamin D: 24.5% in Q1 (i.e. >32.0 ng/mL), 24.9% in Q2 (24.3-32.0 ng/mL), 25% in Q3 (17.7-24.2 ng/mL), and 25.3% in Q4 (<17.6 ng/mL). In multivariate analyses, female sex, low HDL, smoking, all races other than white, excess alcohol use, and older age were independently associated with increase odds of being in Q4 (<17.6 ng/dL). Having a GFR of >60 ml/min, <12 years of schooling, below poverty index line, higher albumin, being physically active, elevated cholesterol, and non-winter seasons were independently associated with a decrease odds of being in Q4. The median follow-up time was 14.7 (range 1.5 to 18.1) years. There were total of 4,076 (25.6%) deaths with 45.1% due to cardiovascular diseases. The overall 18-year Kaplan-Meier survival was 79.2%. survival differed by vitamin D quartiles; 74.0% for <17.6 ng/mL, 74.0% for 17.7-24.2 ng/dL, 79.0% for 24.3-32.0 ng/mL, and 84.0% for >=32.0 ng/mL. There was a 30% increase in all-cause mortality (HR 1.30; 95% CI 1.12, 1.50) among individuals in the lowest quartile of vitamin D (vitamin D>32.2 ng/mL being reference) in adjusted analyses including smoking and physical activity. Excluding individuals with hypertension and diabetes from the multivariate model resulted in an even higher risk (HR 1.35, 95% CI 1.16, 1.57). Being in lowest quartile of vitamin D was associated with increased cardiovascular (HR 1.25, 95% CI 1.05, 1.49) and diabetes (HR 3.27, 95% CI 1.11, 9.65) related deaths. There was no association between being in lowest quartile and cancer, liver, respiratory and infectious diseases related deaths. Conclusion: Lower 25(OH) D levels (<17.8 ng/ml) are strongly associated with all-cause, cardiovascular and diabetes related deaths in the U.S. general population.