Children exposed to valproate in utero--population based evaluation of risks and confounding factors for long-term neurocognitive development.

PURPOSE

To evaluate neurological and cognitive functioning of school-aged (> or =6 years) children exposed to valproate monotherapy in utero in a population based, evaluator-blinded, controlled study.

METHODS

Studied children (N=39, aged 6.6-13.4 years) and their mothers were identified through a population based pregnancy registry. Mothers with carbamazepine monotherapy and mothers with epilepsy but without antiepileptic drug (AED) treatment during pregnancy and their age and gender matched children served as controls. Hospital records were reviewed and neurological examination (Touwens test), intelligent quotients (IQ) of mothers (WAIS), and children (WISC-III) and neuropsychological assessment of children (NEPSY) were performed evaluator-blinded.

RESULTS

The prevalence of low intelligence (FIQ<80) was 19% (4/21) and the prevalence of exceptionally low intelligence (FIQ<70) 10% (2/21) in valproate (VPA) monotherapy exposed children. Children exposed to carbamazepine (CBZ) and children of women with epilepsy but without AED exposure during pregnancy had all at least low average intelligence. The mothers using valproate scored significantly lower (p<0.05) in FIQ, VIQ and PIQ tests and had also significantly lower (p=0.035) educational level. Altogether 21% (8/39) of the children had minor neurological dysfunctions.

CONCLUSIONS

In a population based setting inheritance and cumulating environmental factors may partly explain the increased prevalence of neurocognitive symptoms in children exposed to valproate in utero although concern about the possible long-term effects of intrauterine valproate exposure does exist.