A battery of psychometric tests was administered to 102 adult patients
with epilepsy. Forty-eight were receiving long-term treatment with so
dium valproate (VPA) as monotherapy, and 28 took a combination of VPA
with either carbamazepine or phenytoin. An additional 26 patients were
untreated, and the tests were also administered to 24 nonepileptic co
ntrols. The polypharmacy group performed significantly less well than
the other groups in 6 of the 11 tests. Overall, there were no differen
ces in performance between the VPA monotherapy patients and both the d
rug-free groups. However, patients with higher total and free VPA conc
entrations reported more sedation and scored less well on the threshol
d detection task. Patients were divided into those "tolerant" and "int
olerant' to the drug in terms of these two tests. The "intolerant" gro
up had a significantly earlier age of onset of epilepsy and a lower no
nverbal IQ. Monotherapy with VPA is unlikely to impair cognitive funct
ion and may be a suitable choice if cognitive function is an important
consideration.