Electroencephalographic sleep profiles before and after cognitive behavior therapy of depression.

BACKGROUND Previous studies have not fully resolved the state-dependent vs traitlike behavior of the electroencephalographic sleep abnormalities associated with depression. We therefore examined the sleep profiles of depressed patients before and after 16 weeks of treatment with cognitive behavior therapy to determine the stability or reversibility of selected abnormalities. METHODS Seventy-eight unmedicated patients with major depressive disorder were stratified into abnormal and normal subgroups on the basis of pretreatment sleep study results. Two prospectively defined types of sleep variables were studied: those expected to be traitlike or state independent (type 1) and those predicted to be reversible or state dependent (type 2). RESULTS The type 1 sleep disturbances (reduced rapid eye movement latency, decreased delta sleep ratio, and decreased slow wave sleep [in percentage]) were stable, as predicted, across time. A composite measure of type 2 disturbances (based on rapid eye movement latency, sleep efficiency, and rapid eye movement density) improved significantly, although a minority of patients in remission had persistent abnormalities. CONCLUSIONS The electroencephalographic sleep correlates of depression can be disaggregated into state-independent and partially reversible subgroups. Persistent sleep disturbances in remitted patients may have ominous prognostic implications.