Validity of food frequency questionnaire estimated intakes of folate and other B vitamins in a region without folic acid fortification

Abstract

B vitamins have been implicated in major chronic diseases but results have been inconsistent. This study evaluated the accuracy of dietary intakes of folate, vitamin B12, riboflavin and vitamin B6 as measured by the Northern Sweden Food Frequency Questionnaire (FFQ) against repeated 24-h recalls (24HR) and plasma levels, taking into consideration the MTHFR 677C>T polymorphism. B vitamin intakes assessed by a semi-quantitative FFQ designed to measure the intake over the previous year were compared with those from 10 24HR, as well as to plasma levels of folate and vitamin B12, in randomly selected men (n=96) and women (n=99) aged 30-60 years. FFQ-based B-vitamin intakes were also compared with plasma levels of B-vitamins and with MTHFR 677C4T genotype in 878 men, aged 40-61 years. Intakes of vitamins B12 and riboflavin were similar, whereas folate and B6 intakes were 16-27% higher, as estimated by FFQ versus 24HR. Spearman correlation coefficients between the two methods ranged from 0.31 to 0.63 (all P<or=0.002), and were lowest for vitamin B12. Intakes estimated by FFQ were correlated with plasma levels, but coefficients were lower (range: 0.13-0.33), particularly for vitamin B12 in men (0.15-0.18). Folate intake was not correlated with plasma levels in subjects with the MTHFR 677 T/T genotype. The validity of the Northern Sweden FFQ for assessing B vitamin intake is similar to that of many other FFQs used in large-scale studies. The FFQ is suitable for ranking individuals by intake of folate, riboflavin, vitamin B6 and to a lesser extent vitamin B12.