The objectives of this study were ®rst to compare the responses to moclobemide and sertraline in melancholic and nonmelancholic major depressive patients and secondly to compare the responses of melancholic and nonmelancholic patients in general. Sixty-three patients, with diagnosis of major depression according to the DSM-III-R criteria were included in the study. In this single blind, comparative, randomized study, 29 patients were treated with moclobemide and 34 patients were treated with sertraline for 13 weeks. A 50 per cent decrease of the HDRS (Hamilton Depression Rating Scale) is de®ned as response.
In intent-to-treat analysis the response rates were 69 per cent in the melancholic patients and 59 . 3 per cent in the nonmelancholic group. The dierence is statistically insigni®cant. According to the intent-to-treat analysis in the nonmelancholic group the response rate of the moclobemide-treated patients was 73 . 3 per cent, and 41 . 7 per cent in the sertraline-treated patients. In the melancholic group the response rate was 82 . 4 per cent in the sertraline group and 50 per cent in the moclobemide group. Moclobemide was more eective in the nonmelancholic group whilst sertraline was more eective in melancholic group; but the dierences were not statistically signi®cant. Due to the small size our ®ndings are tentative and need con®rmation using more patients.