Objective: To assess the effects carbamazepine vs valproate sodium on
cognitive functioning in patients with epilepsy compared with normal c
ontrol subjects,Design: Patients with recently diagnosed, previously u
nmedicated seizures participated in a prospective randomized double-bl
ind Department of Veterans Affairs multicenter study of the efficacy a
nd toxicity of carbamazepine vs valproate. Main Outcome Measure: A beh
aviorial toxicity battery was administered prior to treatment and agai
n 6 and 12 months after the initiation of antiepileptic medication. Re
sults: There were no significant differences in the effect of carbamaz
epine vs valproate on motor speed and coordination, memory, or concent
ration and mental flexibility, and there was no significant decline in
neuropsychological performance from pretreatment baseline levels for
either drug. No significant differences in performance were found betw
een patients with low (mean, 52.8 mu g/mL) vs high (mean, 94.4 mu g/mL
) serum valproate levels within the therapeutic range. Patients treate
d with either carbamazepine or valproate did not show practice effects
experienced by normal controls, a finding that may reflect a subtle c
ompromise in cognitive functioning. Conclusion: The impact of carbamaz
epine and valproate monotherapy on cognitive functioning is similar: b
oth drugs produce minimal negative effects compared with pretreatment
baseline performance.