Maternal treatment with valproate during pregnancy is associated with an increased risk of impaired cognitive function in the child at three years of age.

Women with epilepsy taking carbamazepine, lamotrigine, phenytoin or valproate monotherapy during pregnancy were enrolled in an uncontrolled, prospective, observational study between 1999 and 2004 in the UK and US.

There were 309 live births. A total of 258 of these children underwent testing of cognitive function at two or three years of age or both ages.

Adjustment was made for folate use (which varied between 40 and 63%) and other relevant variables such as maternal and gestational age at delivery, maternal IQ and antiepileptic drug dose.

A significantly lower IQ was found in children exposed to valproate. The mean IQ for these children was 92 compared with 101 for those exposed to lamotrigine, 99 for phenytoin and 98 for carbamazepine. The IQ of those exposed to the other three medications was thus between 6 and 9 points higher.

Sub-analysis of the data showed that the child's IQ was strongly correlated to maternal IQ except for those exposed to valproate in whom it was significantly lower than expected. The adverse effect of valproate on IQ was also shown to be dose dependent.

Maternal epilepsy is related to poorer pregnancy outcome especially if seizure control is not good. Sodium valproate is known to carry a high teratogenic risk and if possible is best avoided in pregnancy.

This study has shown that sodium valproate also carries a significant behavioural teratogenic risk as well.

The importance of education and counselling of all women of childbearing age on antiepileptic medication is vital if optimum pregnancy outcome is to be achieved.

Meador KJ, Baker GA, Browning N et al. Cognitive function at 3 years of age after fetal exposure to antiepileptic drugs. N Engl J Med 2009; 360: 1597-607