Electroencephalographic correlates of compulsive eating

Fifty-nine compulsive eaters were interviewed, had a complete waking and sleep EEG tracing, and were offered a trial with phenytoin. Forty-seven patients had an adequate pharmacological intervention. Analyses of the EEGs showed a disproportionately higher percentage of abnormal EEGs and of paroxysmal EEGs in particular among compulsive eaters as contrasted with unselected psychiatric patients and normal controls. Patients with abnormal EEGs responded to phenytoin treatment statistically more frequently than those with normal EEGs--even after conservatively classifying "uncertain" patients as non-responders. Patients with 14 & 6/sec. positive spikes alone showed a phenytoin improvement significantly greater than that seen with either normal EEG subjects or other EEG abnormalities. There are suggestions of a weight by EEG interaction in that the combination of weight deviance (viz: emaciated or obese) plus abnormal EEG seems especially highly predictive of good phenytoin response. The EEG results combined with phenytoin response suggest that neurophysiological etiological hypotheses may have merit for some compulsive eating disorders. Suggestions for additional research are made and the need for controlled replication of these data results is stressed.