AAN: In Utero Exposure to Valproate Lowers IQ

BOSTON, May 3 -- Children whose mothers took valproate (Depakene) during pregnancy for control of epilepsy are at risk for significantly lower IQs than children whose mothers used other anticonvulsants, reported researchers here.

The findings suggest that valproate, which is known to carry a high risk of anatomical birth defects, may also adversely affect neurologic development, and should not be the drug of first choice in women of childbearing age, advised Kimford J. Meador, M.D., of the University of Florida in Gainseville.

Dr. Meador and colleagues had previously reported that major congenital malformations and fetal deaths were significantly more common in pregnancies of women taking valproate compared with other antiepileptic drugs, and that the effect of valproate was dose-dependent. They reported those findings in the Aug. 8, 2006, issue of Neurology.

The new findings emerged from on the same cohort of children, and suggest that valproate, or valproic acid, should be used only in women for whom no other approved anticonvulsant provides adequate epilepsy control, Dr. Meador said at the American Academy of Neurology meeting.

In the current analysis of the NEAD (Neurodevelopment Effects of Antiepileptic Drugs) study, a longitudinal follow-up of 187 two-year-old children whose mothers used one of the four most common epilepsy drugs during pregnancy, 24% of the children of women who used valproate tested in the mental retardation range, compared with 12% or less of children whose mothers took three other anti-epileptic drugs, Dr. Meador said in a briefing.

The effect of valproate on intelligence was not affected by the mother's IQ, one of the most consistently accurate predictors for a child's IQ. In contrast, the IQ of the children was significantly associated with the mother's IQ for women who took either carbemazepine (Tegretol), lamotrigine (Lamictal), or phenytoin (Dilantin) as monotherapy for control of epilepsy.

Dr. Meador noted that findings are of particular concern because in addition to control of epilepsy, valproate, or valproic acid, is frequently used for treatment of other conditions. More than half of all prescriptions for the drug are written for conditions other than epilepsy, he said, including mood disorders, migraine prevention, and several off-label indications.

"We decided to do this study because anticonvulsant drugs are known anatomical teratogens," Dr. Meador said.

Animal studies have also suggested that anticonvulsants can induce behavioral changes at doses lower than those required to produce gross anatomical malformations, he said.

"But when we looked at the human data, it was not clear that this occurred, and the more important question is, if a woman has to take a medication, is the risk higher from one drug to another," Dr. Meador said.

In the NEAD study, outcomes data are controlled for several variables, including seizure frequency, type, maternal IQ, and history of drug, alcohol and tobacco use during pregnancy, and childhood illnesses.

They found that on the Bayley scales of the Mental Development Index, scores for children on valproate were five to 11 points lower than the scores for children on the other drugs, and the differences between valproate and each of the other three agents were statistically significant.

"To put this in perspective, a half standard deviation in IQ score is about seven points, which is clearly clinically significant," Dr. Meador said.

The effects of valproate on intelligence were dose dependent, and were related to blood levels of the drug levels of the blood during pregnancy. Valproate was the only drug to show this type of association he added.

When the investigators analyzed the relationship between drug type used and maternal IQ, they found that for women taking either carbemazepine, lamotrigine, and phenytoin, there was a positive association between the mother's IQ and that of her child, but this effect was not present for valproate, suggesting that it interfered with the association.

"No individual study by itself can prove a causal relationship," Dr. Meador commented, "but when you look at these data in the context of the other data that have come out in just the last few years, which includes at least two retrospective cohorts and one small prospective cohort that also showed IQ differences in the range of approximately 10 IQ points compared with other drugs such as carbemazepine, and if you also take it in the context of six other studies that also showed more major malformation in children exposed to valproate than other antiepileptic drugs, it looks like there's a special risk of valproate compared to any of the other antiepileptic drugs."