Objective: To determine the prevalence of electroencephalogram (EEG) abnormalities at different levels of cognitive impairment and to assess the possible diagnostic usefulness of the test.
Design: Combined prospective assessment of subjects receiving EEGs and retrospective chart review of symptoms and medications.
Setting: Academic geriatric psychiatry service.
Patients: 350 adults age 50 and above; 312 were patients being evaluated for possible organic mental syndrome and 38 were normal controls.
Measurements: All subjects had EEGs and Mini-Mental State Examinations (MMSE) performed at the time of the EEG. EEGs were rated for the presence and type of abnormality, and subjects were stratified according to the severity of impairment. Charts were reviewed by a person blinded to EEG results to determine clinical diagnosis and medications received.
Main results: Abnormal EEGs were significantly more common among all patients (67%) in the study than among controls (11%), and the prevalence of abnormality increased with increasing impairment. Many demented patients with equivocal impairment (42%), and most with mild-to-moderate impairment (65%) had abnormal EEGs. An abnormal EEG was not indicative of dementia even when clear cognitive impairment was present, since patients with depression frequently also had abnormal EEG results.
Conclusions: These findings suggest that the EEG is a moderately sensitive but non-specific indicator of brain dysfunction in the elderly. The significance of abnormalities among patients with equivocal impairment should be more fully assessed by longitudinal follow-up to determine if greater cognitive impairment develops.