Low vitamin D levels were associated with increased all-cause mortality, a comprehensive population-based study found.
Among individuals whose 25-hydroxyvitamin D levels were below 20 ng/mL, the adjusted hazard ratio for death was 2.37 (95% CI 1.87-3), reported Daniel Dudenkov, MD, of the Mayo Clinic in Rochester, Minn, in a poster at the American Society for Bone and Mineral Research annual meeting.
“Levels of serum 25-hydroxyvitamin D below 20 ng/mL and above 50 ng/mL have been associated with chronic adverse events, including mortality,” he said. For instance, one study found levels of 25(OH)D deficiency or insufficiency had increased cardiovascular, cancer, and respiratory disease mortality.
But previous studies have not been population based, so to clarify those observations in a large cohort, he conducted a retrospective study of individuals age 18 and higher in Olmsted County, Minn., who were enrolled in the Rochester Epidemiology Project from 2005 to 2011.
Participants were followed from 30 days after their first vitamin D measurement until leaving Olmsted County, death, or Dec. 31, 2014.
Vitamin D levels were categorized as less than 12 ng/mL, 12-19 ng/mL, 20 to 50 ng/mL (reference range), and higher than 50 ng/mL.
The multivariate analysis adjusted for age, sex, race, Charlson comorbidity index, and month of the year when the vitamin D level was obtained.
The study included 11,022 individuals, whose mean vitamin D level was 30 ng/mL. A total of 643 participants had vitamin D levels below 12, 1,605 had levels from 12 to 19, 8,210 had levels of 20 to 50, and 564 had levels above 50 ng/mL.
Mean age was 54, more than three-quarters were women, and almost 90% were white. Baseline Charlson index was 3.37.
During a median follow-up of 4.8 years, there were 723 deaths.
Compared with the reference range, the unadjusted hazard ratios (HR) for the overall cohort were:
Less than 12 ng/mL: 2.57 (95% CI 2.05-3.23)
12 to 20 ng/mL: 1.28 (95% CI 1.04-1.56)
Higher than 50 ng/mL: (95% CI 0.75-1.56)
“We found a significant interaction by race (P=0.0015),” he said. For instance, for whites whose vitamin D levels were below 12 ng/mL, the HR was 2.83 (95% CI 2.21-3.62), while for nonwhites, the HR was a nonsignificant 1.76 (95% CI 0.93-3.31).
For whites whose levels were 12 to 19 ng/mL, the HR was 1.42 (95% CI 1.14-1.77), whereas for nonwhites, again the HR was a nonsignificant 1.52 (95% CI 0.92-2.51).
For white men, the HR for levels below 12 ng/mL was 4.46 (95% CI 3.33-5.98), and for nonwhite men, it was 2.77 (95% CI 145-5.31). For white women, the corresponding HRs were 2.83 (95% CI 2.21-3.62) and 1.76 (95% CI 0.93-3.31).
Differences were also seen for whites and nonwhites according to age. For example, for participants who were 70 and older and whose vitamin D levels were below 12 ng/mL, the HR for whites was 1.30 (95% CI 1.01-1.67) compared with 0.81 (95% CI 0.43-1.53) for nonwhites.
The study also differed from previous observations in showing that high levels were not associated with increased mortality.
‘”Maybe for nonwhites the normal reference range should be different,” he said.
“We know vitamin D plays a role in bone metabolism, and has various effects on blood vessels, diabetes, and has some effects in the brain,” he told MedPage Today.
As to why all-cause mortality should be linked with low levels of 25-hydroxyvitamin D, he suggested that it was probably through effects on bone, with falls and fractures, and the cardiovascular system, with calcification of blood vessels.
A limitation of the study was that only a single measurement of serum 25-hydroxyvitamin D was obtained.