Gabapentin-Associated Aggression

Case 1. A 38-year-old man was admitted for the third time to our facility with a diagnosis of bipolar disorder, manic phase with psychotic symptoms. There was no history of rapid cycling, comorbid substance abuse, or physical aggression. His symptoms were partially responsive to a combination of lithium carbonate 1,650 mg/day (level of 0.7 mEq/l), olanzapine 25 mg/day, and lorazepam 2 mg bid. Gabapentin was started in a dose of 300 mg tid and increased to 1,200 mg/day by day 3. By day 4, the patient became very irritable and physically aggressive. He attacked a nurse, and mechanical restraints were used to control his behavior. He did not receive any additional medication at this time. The patient later recognized this behavior as out of character for him. Gabapentin was stopped while all other medications were continued. Within a day, his irritability decreased and aggressive behavior stopped completely. No further incidents of aggression were noted throughout the rest of his hospital stay.

Case 2. A 51-year-old man with a diagnosis of bipolar disorder was admitted for the fourth time to our facility in a manic phase with mixed affective features. He had no history of psychotic symptoms, substance abuse, or aggression. On a combination of sodium valproate 2 g/day and risperidone 4 mg/day (to which the patient had responded well on previous admission) he became very withdrawn, spending 12 hours a day in bed, but denied depressive symptoms. When a reduction in sodium valproate to 1 g did not improve his condition, a decision was made to switch mood stabilizers. Gabapentin was started in a dose of 300 mg bid, and sodium valproate was stopped. Within 48 hours, the patient developed new symptoms of hostility, verbal aggression, and command auditory hallucinations and destroyed property by breaking the commode in his room. The patient clearly attributed these new symptoms to gabapentin. Gabapentin was stopped and sodium valproate reinstated in a dose of 1g per day. Within 48 hours, hostility, aggressive behavior, and hallucinations stopped. No further aggressive behavior was noted throughout his hospital stay.

Aggression in the above two cases could be an aspect of mania, or it could be gabapentin-related. The absence of prior aggression, the temporal relation to instituting gabapentin, the cessation of aggression soon after stopping gabapentin, and the ego-dystonic nature of the aggressive behavior indicate an association between gabapentin and aggression. To our knowledge, there are no published case reports of aggression with gabapentin in adult patients. More than 100 cases of hostility and aggression have been reported to the manufacturer in the past 5 years (personal communication, Parke-Davis, 2000). With increasing use of gabapentin in psychiatry, this possible adverse effect should be kept in mind.