The efficacy of carbamazepine (CBZ) in treatment of ADHD has been determined by meta-analysis of 10 reports from the international literature reviewed at Columbia University, St Luke’s-Roosevelt Hospital Center, and New York University Medical Center. In 7 open studies involving a total of 189 patients with features of motor overactivity, impulsivity, and distractibility, 70% showed a marked improvement in target symptoms following treatment with CBZ for periods varying from 1 week to 8 years. Outcome was significantly correlated with duration of treatment; the longer the treatment the better the outcome. In 3 placebo-controlled, double-blind studies, 71% of 53 patients treated with CBZ were benefited whereas only 26% of 52 receiving placebo showed similar improvement in attentiveness and behavior. The difference was significant (p=.018). The most frequent side effects were sedation and skin rash occurring in 7.5% and 5.7% of CBZ-treated patients, respectively. 
COMMENT. The authors concluded that carbamazepine may be an effective alternate treatment for ADHD. A response rate of 70% in both open and controlled studies is about the equivalent effectiveness of stimulant medication.
From a neurologist’s perspective, the obvious questions would relate to the incidence of epilepsy and epileptiform EEG’s in these patients selected for treatment with an anticonvulsant medication. Unfortunately, these data were not discussed and were tabulated for the entire sample and not the subsample with ADHD. My own meta-analysis of these data show that abnormal EEGs occurred in 69% of 57 patients in the controlled studies and in 82% of 50 patients in the one open study providing EEG data. Seizures were mentioned in 4 of the studies, affecting 13 plus patients, but the total number of patients affected was not given. The frequency of abnormal EEGs in these patients is considerably higher than that usually reported for ADHD. It seems that CBZ might be indicated for the treatment of ADHD symptoms in some patients with abnormal EEGs and/or a history of seizures. On a negative note, see Progress in Pediatric Neurology II, 1994, ppl88-190, for references to cognitive impairment and impulsivity caused by CBZ treatment of epilepsy.
The cognitive effects of carbamazepine, phenobarbital, and valproate were compared in 73 children with newly diagnosed epilepsy studied at the National Cheng Kung University; Chi Nei Hospital; and Tainan Municipal Hospital, Tainan, Taiwan, ROC . Only children treated with phenobarbital showed increased P300 latencies on auditory event-related potentials, which was inversely related to IQ scores after treatment for 6 to 12 months. WISC-R IQs and Bender-Gestalt scores were not significantly different in any of the groups before or after treatment. P300 latency was a more sensitive indicator of AED effects on cognitive function than the WISC-R and Bender-Gestalt.
Behavioral side effects of Gabapentin are reported in 7 children with base-line ADHD and developmental delays who were followed in the epilepsy program and in the Department of Child Psychiatry, Emory University, Atlanta, GA . Tantrums, aggression, hyperactivity, and defiance were the most troublesome symptoms. The majority (64%) of intensified behaviors were similar to baseline ADHD symptoms; 21% were ODD and 8% were CD symptoms. New behaviors, not exhibited before gabapentin therapy, were ODD or CD. Behavioral side effects resolved after decrease or withdrawal of gabapentin.