Association of dietary vitamin K intake with coronary disease and cardiovascular mortality
| Study | Sample size | Patient population | Study methods | Study findings |
| Geleijnse et al91 | 4807 | Dutch subjects >55 years without prior MI at baseline (1990–1993) Sex: 38% male Age: 67 | Design: prospective, population-based Intervention: vitamin K1 and K2 using FFQ Follow-up (years): 7 | ↓Cardiovascular mortality (RR=0.43 (0.24 to 0.77)) and all-cause mortality (RR=0.74 (0.59 to 0.92)) with K2 but not K1 in upper tertials of energy-adjusted intake |
| Erkkilä et al113 | 40 087 | Healthcare workers aged 40–75 years, free of CVD, stroke and cancer at baseline (1986–2000) Sex: male only Age: 53 | Design: prospective, population-based Intervention: vitamin K1 using FFQ Follow-up (years): 14 | ↓Incidence of total CAD (RR=0.84, p trend=0.05) |
| Gast et al87 | 16 057 | Subjects aged 49–70 years free of CVD, recruited from the European Prospect-EPIC cohort (1993–1997) Sex: female only Age: 67 | Design: prospective, population-based Intervention: vitamin K2 (subtypes MK-7–9) using FFQ Follow-up (years): 8 | ↓Risk of incident CAD (HR=0.91 (0.85 to 1.00)) |
| Juanola-Falgarona et al114 | 7216 | Community-dwelling adults enrolled in PREDIMED trial, without baseline CVD but with either type 2 diabetes or ≥3 cardiovascular risk factors Sex: 57% female Age: 67 | Design: prospective, population-based Intervention: dietary vitamin K1 and K2 (subtypes MK-7–9) using FFQ Follow-up (years): 5 | ↓All-cause mortality with increasing intake of vitamin K1 or K2 (HR=0.57 and HR=0.55, respectively; p<0.05) |
| Cheung et al115 | 3401 | Non-hospitalised participants ≥20 years of age with CKD from the NHANES III Study (1988–1994) Sex: 67% female Age: 62 | Design: prospective, population-based Intervention: vitamin K1 and K2 using 24-hour dietary recall Follow-up (years): 13 | ↓All-cause (HR=0.85 (0.72 to 1), p=0.047) and CVD mortality (HR=0.78 (0.64 to 0.95), p=0.016) |
| Zwakenberg et al121 | 33 289 | Dutch subjects aged 20–70 years without baseline CVD, diabetes or cancer recruited from the EPIC-NL cohort (1993–1997) Sex: 26% male Age: 49 | Design: prospective, population-based Intervention: vitamin K1 and K2 using FFQ Follow-up (years): 17 | Borderline ↓ CHD mortality with higher long chain vitamin K2 intake (HR=0.86 (0.74 to 1.00), p trend=0.06) No association of vitamin K1 or K2 with CVD or all-cause mortality |
| Haugsgjerd et al112 | 2987 | Healthy Norwegian subjects aged 46–49 years without baseline CAD recruited from the Hordaland Health Study (1997–1999) Sex: 57% female Age: 48 | Design: prospective, population-based Intervention: vitamin K1 and K2 using FFQ Follow-up (years): 11 | ↓Risk of incident CAD with higher vitamin K2 but not K1 (HR=0.52, p trend=0.03), attenuated by adjusting for dietary confounders (HR=0.58, p trend=0.16) |
Both ages and follow-up times are presented as mean years, or median years in Cheung et al115 and rounded to the nearest whole number. P values <0.05 denote significance.
CAD, coronary artery disease; CHD, coronary heart disease; CKD, chronic kidney disease; CVD, cardiovascular disease; FFQ, Food Frequency Questionnaire; MI, myocardial infarction; MK, menaquinone; NHANES III, National Health and Nutrition Examination Survey; RR, relative risk.