Vitamin K is integral to haemostatic function, and in vitro and animal experiments suggest that vitamin K can suppress production of inflammatory cytokines. To test the hypothesis that higher vitamin K status is associated with lower haemostatic activation and inflammation in community-dwelling adults, we analysed the cross-sectional association between serum phylloquinone (vitamin K1) with haemostatic and inflammatory biomarkers in 662 participants in the Multi-Ethnic Study of Atherosclerosis (MESA) [mean (SD) age=62 (10) years; 46% female; 37% Caucasian, 25% African-American, 25% Hispanic, 13% Chinese-American]. Following adjustment for demographic and lifestyle characteristics, medication use, triglycerides and body mass index, those in the highest quartile of serum phylloquinone had significantly lower circulating interleukin-6 [adjusted mean (SEM) pmol/l: quartile 4 (Q4)=1.22 (0.07), quartile 1 (Q1)=1.45 (0.07); p-trend<0.01], C-reactive protein [adjusted mean (SEM) mg/dl: Q4=1.57 (0.11), Q1=2.08 (0.18); p-trend=0.02], soluble intercellular adhesion molecule-1 [adjusted mean (SEM) ng/ml: Q4=247 (11), Q1=288 (11); p-trend=0.02], and plasmin-antiplasmin complex [adjusted mean (SEM) nmol/l: Q4=4.02 (0.1), Q1=4.31 (0.1), p-trend=0.04]. We detected an interaction between age and serum phylloquinone with respect to factor VIII and D-dimer (interaction p-values=0.03 and 0.09, respectively). Among participants ≥70 years, serum phylloquinone was inversely associated with factor VIII activity (p-trend=0.06) and positively associated with D-dimer (p-trend=0.01), but was not associated with either marker among participants <70 years (both p≥0.38). In contrast, dietary phylloquinone intake was not associated with any inflammatory or haemostatic biomarker evaluated (all p-trend>0.11). These findings are consistent with laboratory-based studies that suggest a possible anti-inflammatory role for vitamin K. Whether or not these associations predict clinical outcomes linked to elevated inflammation or haemostatic activation remains to be determined.
Keywords: Haemostasis; epidemiological studies; inflammation; nutrition.