Comparative cognitive effects of carbamazepine and phenytoin in healthy adults.

We investigated neuropsychological effects of carbamazepine and phenytoin in 21 healthy adults using a randomized, double-blind, double-crossover design and treating each subject with each drug for 1 month, separated by a 1-month washout. There were neuropsychological evaluations at baseline, the end of each treatment month, and 1 month after the last treatment phase. Cognitive measures included Symbol Digit Modalities Test, Selective Reminding Test, Complex Figures, Paced Auditory Serial Addition Test, Stroop, Finger Tapping, Grooved Pegboard, Choice Reaction Time, P3 Event-Related Potential, Hopkins Symptom Checklist, and Profile of Mood States (POMS). Compared with nondrug conditions, the anticonvulsants significantly impaired Stroop, Choice Reaction Time, Grooved Pegboard, Hopkins, and POMS. Employing anticonvulsant blood levels as covariates, there were only two significant differences between drugs, one in favor of carbamazepine (ie, Finger Tapping) and one in favor of phenytoin (ie, Stroop). The results suggest that differences in cognitive effects of carbamazepine and phenytoin are not clinically significant.