Approximately 60% of depressed patients experience an antidepressant response to sleep deprivation (SD). However, the use of SD as a treatment for depression has been limited by the fact that most patients relapse after a night of recovery sleep. This study tests the efficacy of two possible clinical applications of SD. First, SD is used to hasten the onset of action of fluoxetine (Protocol I). Second, SD is added to the ongoing pharmacotherapy of patients who have had a partial response to antidepressant medication in an attempt to potentiate the effect of the medication (Protocol II). Levels of TSH, free T-3, prolactin, and cortisol are measured before and after SD, in order to determine their potential rerevence to SD's antidepressant effect. Since it is difficult to design an adequate control condition for SD, the SD literature has been largely uncontrolled. SD in the first half of the night (early sleep deprivation, ESD) has been shown to be less effective than SD in the second half of the night (late sleep deprivation, LSD). Therefore, we use ESD as a control condition for LSD. Patients must meet DSM-III-R criteria for major depressive disorder, or for bipolar disorder, depressed. Patients are randomly assigned to ESD or LSD. Patients in Protocol I are started on fluoxetine four days before their first night of SD. Patients in Protocol II, who have all been on a stable regimen of medication for at least six weeks, are continued on that regimen throughout the study. After a baseline day of multiple mood ratings and blood sampling for hormonal levels, patients are sleep deprived for two nights in a row. One week later, the course of two nights of sleep deprivation is repeated. Patients are then followed on a weekly basis for three weeks with mood ratings and repeated blood samples.
