A literature review of articles published between 1941 and 1994 was used to determine the utility of the electroencephalogram in the evaluation of patients with headache and the results are reported from the Departments of Neurology, Lackland AFB, TX, and Andrews AFB, MD. EEG findings that occurred more frequently in headache patients than in controls were as follows: generalized or focal slowing, hyperventilation-induced high-voltage slowing, excessive fast activity, epileptiform discharges, prominent photic driving, and differences in symmetry and frequencies of alpha rhythm. Two studies showed an increased prevalence of 14 & 6 positive spikes in the EEGs of children with migraine. One author suggested that headaches of children whose EEGs demonstrated epileptiform activity would respond best to anticonvulsants. In well controlled studies, a prominent photic driving response at high flash rates (the “H-response”) was the only abnormality distinguishing headache patients from those without headache and migraine from other headache types. Clinical criteria were of greater diagnostic value than the EEG in identifying headache subtypes, including those with structural lesions. 
COMMENT. The authors conclude that the EEG is not indicated in routine evaluations of patients presenting with headache. However, they do recommend an EEG in patients with headache and symptoms suggestive of seizures and in those with atypical migrainous auras or episodic loss of consciousness. In one article not cited by the authors, 77 percent of 30 children with migraine benefited from treatment with the anticonvulsant, phenytoin. Response to phenytoin was not correlated with an abnormal EEG In 13 with abnormal and 17 with normal EEGs, the beneficial response rates were 61% and 88%, respectively.