A method comparison of a food frequency questionnaire to measure folate, choline, betaine, vitamin C and carotenoids with 24-h dietary recalls in women of reproductive age.
Coathup V., Wheeler S., Smith L.
BACKGROUND/OBJECTIVES: The objective of this study was to conduct a method comparison of a modified food frequency questionnaire (FFQ), designed to estimate usual dietary intake of selected micronutrients and antioxidants including folate, choline, betaine, vitamin C and carotenoids (α-carotene, β-carotene, lutein, lycopene and β-cryptoxanthin) with 24-h dietary recalls (24-HR) in women of reproductive age. SUBJECTS/METHODS: Sixty-four British women of reproductive age (18-40 years) were recruited in Oxford, UK and provided complete dietary data for analysis. METHODS: We compared micronutrient estimates from the FFQ against estimates derived from three multiple-pass, 24-HR interviews, by evaluating Pearson's correlation coefficients and Bland-Altman plots. RESULTS: Median intakes of most nutrients were higher when measured by FFQ compared with 24-HR. Strong correlation coefficients were observed for folate (r=0.80) and choline (r=0.68), whereas moderate correlation coefficients were observed for vitamin C (0.50) and lycopene (0.43). Weak correlation coefficients were observed for betaine (0.39) and other carotenoids (r=0.26-0.38). Bland-Altman plots indicated that there was a large amount of variability in the FFQ estimates of nutrient intakes compared to those using 24-HR, particularly for carotenoids. CONCLUSIONS: The findings indicate that this FFQ estimated higher mean intakes for most nutrients. Pearson's correlation coefficients were comparable with previous research; however, the Bland-Altman plots suggest a high variability in mean nutrient estimates between the FFQ and 24-h. We recommend further investigation of the validity of this FFQ before use.