Background/objectives: This study aims to investigate the reproducibility and relative validity of the Dutch food frequency questionnaire (FFQ), to estimate intake of dietary phylloquinone and menaquinones compared with 24-h dietary recalls (24HDRs) and plasma markers of vitamin K status.
Subjects/methods: In a cross-sectional study among 63 men and 58 women, the FFQ was completed three times over a 1-year period and the reproducibility was calculated over these measurements. Twelve-monthly 24HDR were collected to estimate relative validity. In addition, the relative validity of the FFQ, compared with plasma phylloquinone and desphospho-uncarboxylated matrix Gla protein (dpucMGP), was assessed cross-sectionally among 507 postmenopausal women.
Results: Intraclass correlations showed a good reproducibility, with correlations ranging from 0.65 to 0.83. The relative validity for phylloquinone intake compared with 24HDR was lower for women (rs=0.28) than men (rs=0.40). The relative validity, compared with 24HDR, for intake of short-chain menaquinones were ranging between 0.30 and 0.34. Long-chain menaquinones showed good relative validity (rs=0.60-0.69). Plasma phylloquinone concentrations were weakly correlated with phylloquinone intake (rs=0.16 (0.07-0.24). Plasma dpucMGP was negatively but weakly correlated with phylloquinone intake (rs=-0.09 (-0.18; -0.01)) and long-chain menaquinones (rs=-0.13 (-0.21; -0.04)), but not with short-chain menaquinones (rs=-0.04 (-0.13; 0.05)).
Conclusions: The FFQ is reproducible to rank subjects for phylloquinone and menaquinone intake.The relative validity of our FFQ, compared with 24HDR, to estimate intake of phylloquinone and short-chain menaquinones was low, but the relative validity for long-chain menaquinones was good. The relative validity of our FFQ, compared with plasma phylloquinone and dpucMGP, was relatively low for both phylloquinone and menaquinone intake.