We recently completed a case–control study of 28 women with a previous pregnancy resulting in a neural tube defect and 38 matched controls with a normal pregnancy outcome. All mothers were ascertained to be screen positive by an elevated maternal serum a-fetoprotein level between 1983 and 1999. We used a semiquantitative food frequency survey to measure dietary and supplemental intake of folate and vitamins B₁₂ and B₆. The dietary survey was later validated using biochemical results from 25 and 32 of the case and control mothers respectively. Linear regression analysis showed significant correlation between reported intake of folate and serum folate (p = 0.018) and red blood cell folate levels (p = 0.002), but an inverse correlation with serum homocysteine levels (p = 0.029). Analysis indicated consistent underreporting of actual vitamin intake (common in food frequency surveys). We found no difference between case and control subjects in terms of intake or eating patterns.

Even after correcting for underreporting, we found that in our case and control groups combined, 58% of women were not consuming enough folate, 46% were not consuming enough vitamin B₆ and 28% were not consuming enough vitamin B₁₂. Only 12% reported preconceptional supplementation whereas 82% reported supplementing after they became pregnant. There was no difference in preconceptional supplementation patterns after 1994, when preconceptional supplementation with folic acid was recommended.⁴

Although none of our mothers were folate deficient according to current reference values, 34% had serum homocysteine values in excess of 10 µmol/L and 10% had levels higher than 13 µmol/L, the current reference value. Our results, albeit in a much smaller and differently selected group, support the results of the Newfoundland study and indicate that vitamin B₆ intake may also be suboptimal in many women of child-bearing age.