Low vitamin D levels not useful as predictive risk marker for mortality

American Heart Association Scientific Sessions 2010

CHICAGO — Low levels of vitamin D were not predictive of increased all-cause mortality or for CVD mortality among participants in the Women’s Health Initiative, results from a study suggested.

Researchers for the study enrolled participants in the Women’s Health Initiative (WHI) recruited from 40 centers between 1993 and 1998. The analytic cohort included 2,429 postmenopausal women with 25(OH) vitamin D levels ascertained in the WHI. Mortality was assessed at annual and semi-annual follow-up, and was collected through a number of methods, including a review of the National Death Index and self-reporting by family members, as well as adjudication of hospitalizations, ED visits, autopsy and coroner’s reports. For the purposes of the study, CVD mortality included coronary disease, cerebrovascular disease, pulmonary embolism, HF and other CV causes. Serum 25(OH) vitamin D was assessed using an immunoassay.

Participants were divided into quartiles of seasonal-adjusted vitamin D to account for differences in sun exposure from season to season, with quartile 1 having the lowest levels of vitamin D and quartile 4 having the highest levels.

According to the study results, there was an approximately 62% increased risk for all-cause mortality from the lowest quartile compared with the highest quartile (HR=1.62; 95% CI, 1.11-2.36), but the relationship did not attain statistical significance when adjusting for confounding variables such as age, ethnicity, hypertension, smoking, CVD, diabetes and others (HR=1.27; 95% CI, 0.81-1.99). For CV mortality, the same relationship was observed between the lowest and highest quartiles prior to adjustment (HR=1.92; 95% CI, 1.03-3.58), but also was not statistically significant after adjusting for confounding variables (HR=1.30; 95% CI, 0.83-2.03).

Limitations of the study included potential misclassification of cause-specific mortality, non-random sampling, unmeasured confounding bias and possible misclassification of 25(OH) vitamin D exposure.

“Low vitamin D levels are not an independent prospective risk marker for increased all-cause, CVD or cancer mortality in the WHI,” Charles B. Eaton, MD, of Brown University in Pawtucket, RI, said in a presentation. “Waist circumference appears to be an important confounding factor and shows effect modification of the relationship of vitamin D and all-cause mortality.” – by Eric Raible

For more information:

Eaton C. AOS 211.06. Presented at: American Heart Association Scientific Sessions 2010; Nov. 13-17; Chicago.

Follow CardiologyToday.com on Twitter.