Validation of the assessment of folate and vitamin B12 intake in women of reproductive age: the method of triads

Abstract

To validate the folate and vitamin B12 intakes estimated by a food-frequency questionnaire (FFQ) designed to be used in a case-control study on the association between maternal dietary intake and the risk of having a child with a congenital heart defect. The FFQ was filled out by 53 women of reproductive age. Immediately thereafter, blood samples were taken to determine serum folate, red blood cell (RBC) folate and serum vitamin B12 concentrations. Subsequently, three dietary 24-h recalls (24HR) were completed during a period of three successive weeks and used as a reference method. The recalls comprised two weekdays and one weekend day. Using the method of triads, validity coefficients were calculated by comparing nutrient intakes derived from the FFQ and 24HR with the corresponding nutritional biomarkers in blood. The validity coefficient is the correlation between the dietary intake reported by the FFQ and the unknown 'true' dietary intake. The comparison of B-vitamin intakes reported by the FFQ and the mean of the 24HR revealed deattenuated correlation coefficients of 0.98 for folate and 0.66 for vitamin B12. The correlation coefficients between the B-vitamin intakes estimated by the FFQ and concentrations of serum folate, RBC folate and serum vitamin B12 were 0.20, 0.28 and 0.21, respectively. The validity coefficients for serum folate, RBC folate and serum vitamin B12 were 0.94, 0.75 and 1.00, respectively. The estimated folate and vitamin B12 intakes were comparable with the results of the most recent Dutch food consumption survey. The adapted FFQ is a reliable tool to estimate the dietary intake of energy, macronutrients, folate and vitamin B12 in women of reproductive age. Therefore, this FFQ is suitable for the investigation of nutrient-disease associations in future. Funding was provided by the Netherlands Heart Foundation (Grant 2002.B027).