Positron emission tomography with 18-F deoxyglucose (FDG) will be used to study regional brain metabolic patterns during rapid eye movement (REM) sleep and non-REM sleep in 40 patients with affective disorder and age and sex-matched controls. In Year 1 of this project, we have carried out nighttime PET scans in sleeping normals in REM sleep (n=12), NREM sleep (n=12) and waking (n=12). In this continuation proposal, patients with effective disorder will be studied together with concurrent controls, effective patients would be pre-screened to establish sleep stage timing and identify those with a REM latency between 30 and 60 minutes. Patients ages 21-55, meeting DSM-III criteria for Major Affective Disorder and drawn from the NIMH Clinical Research Center, UCSD, and the Research Unit if the Department of Psychiatry, UC Irvine, will serve as subjects. All patients will be off psychoactive medication a minimum of two weeks, subjects will sleep in a small bedroom with two 10-foot IV lines and polygraph leads extending to a monitoring station outside. We will monitor the polygraphic record and infuse FDG either five minutes after entering Stage 1 or five minutes after entering REM period 2, as in our initial experiments. After 30 minutes of FDG uptake, patients will be scanned, and glucose metabolic rate calculated for cortical and subcortical regions. Twenty patients will be studied in REM sleep and 20 in NREM sleep. controls will be matched to each group. If a second project, patients will be followed and restudied when remitted and again off medication and groups and stages contrasted by repeated measures NOVA. In a third experiment, we will study 20 patients and 20 normals NREM sleep following a night of sleep deprivation. We predict that affective patients may show higher than normal metabolic rates in NREM sleep.