Neuropsychological effects of antiepileptic drugs (carbamazepine versus valproate) in adult males with epilepsy.

PURPOSE

To evaluate the effect of antiepileptic drugs (AEDs) on cognition and behavior in adult epileptic males controlled on treatment with conventional antiepileptic medications.

METHODS

Cognitive, mood, behavior and personality traits were assessed in 45 epileptic patients treated with carbamazepine and/or valproate and free of seizures for >/=1 year. Thirty-four newly diagnosed or untreated patients with epilepsy and 58 matched healthy subjects were also included for comparison. A battery of psychometric tests was utilized including Stanford-Binet (4th edition), Beck Inventory for Depression, Aggressive Scale and Eysenck Personality Questionnaire.

RESULTS

Compared to matched control subjects, treated and untreated epileptic patients had poor performance in different cognitive and behavioral functions testing. Treated patients had worse scores in memory for digits forward and backward, total short-term memory, extroversion and psychosis. The duration of AEDs intake was correlated with memory of objects (r = -0.323; P = 0.030), bead memory (r = -0.314; P = 0.036) and total nonverbal short-term memory (r = -0.346; P = 0.020). Treated and untreated epileptic patients had poor performance of similar extent in behavioral functions testing (depression, aggression and neurosis). The dose of AEDs was correlated with testing scores for neurosis (r = 0.307; P = 0.040), verbal aggression (r = 0.483; P = 0.001) and nonverbal aggression (r = 0.526; P = 0.000), and duration of drug intake was correlated with scores for depression (r = 0.384; P = 0.009), psychosis (r = 0.586; P = 0.0001) and nonverbal aggression (r = 0.300; P = 0.045).

CONCLUSIONS

This study provides support for the notion that AEDs can impair performance in cognition, mood and behavior. Duration of drug intake and the number of the utilized AEDs are the main confounding variables. This study did not provide clues on how to exclude the effect of epilepsy itself and psychosocial variables as additional important confounding variables.