Abstracts

Rationale: We previously reported that fetal exposure to valproate impairs IQ (Meador et al. NEJM 2009). In this follow up investigation, we examined the dose-related effects of fetal antiepileptic drug (AED) exposure on verbal and nonverbal cognitive measures. Methods: The NEAD Study is an ongoing prospective observational multicenter study in the USA and UK, which has enrolled pregnant women with epilepsy on AED monotherapy from 1999 to 2004. The purpose of the investigation is to determine if differential long-term neurodevelopmental effects exist across four commonly used AEDs (carbamazepine, lamotrigine, phenytoin, or valproate). This report compares the verbal vs. nonverbal cognitive outcomes in 234 children at 3 years/old. Index scores for the following cognitive domains were calculated (using subscores of Differential Ability Scales and Preschool Language Scale-4): Expressive Language, Receptive Language, Visuomotor-Construction, and Non-verbal Intellectual Ability. These analyses used children who completed testing at age 3. Average AED dose during pregnancy was standardized relative to ranges observed within each group, i.e., 100 × (observed dose - minimum dose) ÷ range of doses. Partial correlations for dose vs. child score on each domain were calculated for each AED controlling for maternal IQ. Results: Child performance was negatively associated with valproate dose for all four domains: expressive language (r = -.40; p<.007), receptive language (r = -.45; p<.002), visuomotor-constructive (r = -.39; p<.007) and non-verbal (r = -.34; p<.02) domains. Child performance was also negatively associated with carbamazepine dose for expressive language (r = -.36; p<.005) and receptive language (r = -.32; p<.01). No dose effects were seen for lamotrigine and phenytoin. Conclusions: Higher valproate dose was associated with lower scores in verbal and non-verbal cognitive domains, consistent with our previous finding of lower overall IQ. Carbamazepine dose was associated with lower language scores, raising concerns that higher doses during pregnancy may adversely affect a child’s language ability. Additional research is needed to confirm this finding.