Abstract 121

The aim was to define 1) the risk of hypoglycaemia (blood glucose conc.(BG) < 1.8 mmol/l) in term infants exposed in utero to valprote and 2) the frequency of withdrawal symptoms. Twenty epileptic women were treated with valproate as monotherapy and two with valproate and carbamazepine. In 1st trimester the daily median dose of valproate was 1.0 g (range 0.3-4.2 g) and in 3rd trimester 1.2 g (range 0.3-4.8 g). 22 infants (59%) became hypoglycaemic and hypoglycaemia was seen mostly in 1st or 2nd hour of life. 1 infant had 8 episodes of hypoglycaemia, 1 had 3 episodes, 2 infants 2 episodes and 9 infants 1 episode each. The lowest value of BG was 1.0 mmol/l. All episodes were asymptomatic. The BG and frequency of hypoglycaemia 1 hour of life were significantly lower than in a reference group of normal term infants (p < 0.01 and < 0.05). Further, the BG and development of hypoglycaemia were significantly negatively correlated to the maternal median plasma conc. of total valproate during 3rd trimester (p < 0.0003 and < 0.0001). There was no basis for the hypoglycaemia was caused by hyperinsulinaemia. 22 infants (45%) had withdrawal symptoms such as irritability, jitteriness, hypertonia and vomiting. Whether or not an infant developed withdrawal symptoms was significantly positively correlated to median dose of valproate during 3rd trimester and to maternel plasma conc. of free fraction of valproate at delivery (p < 0.02). Conclusion: Infants exposed in utero to valproate have a significantly risk of hypoglycaemia, and withdrawal symptoms are very often seen.