Dietary and nondietary determinants of vitamin K biochemical measures in men and women.

Few epidemiological studies that rely on the food frequency questionnaire (FFQ) for dietary assessment have measured biomarkers of vitamin K intake to independently confirm associations between self-reported dietary vitamin K intake and disease risk. Associations were examined between two sensitive biomarkers of vitamin K status, plasma phylloquinone and serum percent undercarboxylated osteocalcin (%ucOC), and self-reported usual phylloquinone intake as estimated from a FFQ. The influence of other dietary and nondietary factors on plasma phylloquinone concentrations was also examined. Dietary phylloquinone intake was estimated using a FFQ in 369 men and 468 women of the Framingham Offspring Study. The prevalence of high %ucOC concentrations (>/= 20%), suggestive of a low vitamin K status, was 44% in men and 54% in women, respectively. After multivariate adjustment, the odds of a high %ucOC was 2.5 greater for women (odds ratio: 2.5; 95% confidence interval [CI]: 1.2-5.1) and almost three times greater for men (odds ratio: 2.8; 95% CI: 1.3-5.9) in the lowest dietary phylloquinone intake quintile category compared to the highest quintile category. Fasting triglyceride concentrations, smoking status and season were associated with plasma phylloquinone concentrations, independent of dietary phylloquinone intake. Phylloquinone and green vegetable intake was linearly associated with plasma phylloquinone, after adjustment for potential confounding factors. There were limitations in the use of the FFQ to predict plasma phylloquinone, evident in an observed plateau effect and required nondietary adjustment factors. Despite these caveats, these findings support the use of a FFQ for a relative assessment of vitamin K status in population-based studies.