Abstract:The sleep electroencephalogram (EEG) provides biomarkers of depression, which may help with diagnosis, prediction of therapy response, and prognosis in the treatment of depression. In patients with depression, characteristic sleep EEG changes include impaired sleep continuity, disinhibition of rapid-eye-movement (REM) sleep, and impaired non-REM sleep. Most antidepressants suppress REM sleep in depressed patients, healthy volunteers, and in animal models. REM suppression appears to be an important, but not an absolute requirement, for antidepressive effects of a substance. Enhanced REM density, a measure for frequency of REM, characterizes high-risk probands for affective disorders. REM-sleep changes were also found in animal models of depression. Sleep-EEG variables were shown to predict the response to treatment with antidepressants. Furthermore, certain clusters of sleep EEG variables predicted the course of the disorder for several years. Some of the predicted sleep EEG markers appear to be related to hypothalamic-pituitary-adrenal system activity.