Anticonvulsants, folate levels, and pregnancy outcome: a prospective study.

Folate levels in serum and red cells, as determined by a microbiological assay using Lactobacillus casei, and plasma anticonvulsant concentrations were monitored concurrently in nonpregnant (50 subjects) and pregnant (49 pregnancies in 46 subjects) epileptic women. Twenty-three (46%) nonpregnant women had subnormal serum folate levels and 4 nonpregnant women (8%) showed subnormal red cell folate levels. In pregnant women not taking folate supplements, the incidence of folate deficiency increased as the pregnancy advanced. Pregnant women taking folate supplements achieved normal or supranormal blood folate concentrations. In both nonpregnant and pregnant women, serum and red cell folate levels were inversely correlated with plasma concentrations of phenytoin and of phenobarbital, and with the number of anticonvulsants. In 49 pregnancies, there were 10 abnormal outcomes (20.4%): 4 spontaneous abortions (8.2%) and 6 children with major congenital malformations (12.2%). Blood folate levels were significantly lower in pregnancies with an abnormal outcome than in those with a normal outcome. The results suggest a dose-response relationship among anticonvulsants, folate, and adverse pregnancy outcome.