BACKGROUND: Inhibitors of steroid synthesis have been reported to exert antidepressive effects, according to preliminary findings. OBJECTIVE: To test whether the addition of metyrapone to standard antidepressants induces a more rapid, more efficacious, and sustained treatment response in patients with major depression. DESIGN: Double-blind, randomized, placebo-controlled trial. SETTING: Hospitalized care. PATIENTS: Sixty-three inpatients with a DSM-IV diagnosis of major depression and a baseline score 18 points or higher on the Hamilton Rating Scale for Depression. INTERVENTIONS: Random allocation to 2 treatment groups receiving either placebo or metyrapone (1 g/d) for the first 3 weeks during a 5-week treatment with standard serotonergic antidepressants (nefazodone or fluvoxamine). MAIN OUTCOME MEASURES: Primary outcome criteria were the number of responders and the time to onset of action. Responder rates were considered twice after 3 and 5 weeks with a definition of treatment response as 30% and 50% reduction, respectively, of baseline Hamilton Rating Scale for Depression scores. Onset of action was defined as the time point at which at least a 20% reduction of baseline Hamilton Rating Scale for Depression scores occurred. RESULTS: Using intention-to-treat analysis, we found that a higher proportion of patients receiving metyrapone showed a positive treatment response at day 21 (23 of 33 patients) and at day 35 (19 of 33 patients) compared with placebo patients (day 21: 13 of 30 patients; Fisher exact P = .031; day 35: 10 of 30 patients; Fisher exact P = .047). The clinical course of patients treated with metyrapone showed an earlier onset of action (Kaplan-Meier analysis; log-rank test P
