Antidepressant responses in direct comparisons of melancholic and non-melancholic depression

Background: Efforts to develop less heterogeneous, more clinically useful diagnostic categories for depressive disorders include renewed interest in the concept of melancholia (Mel). However, clinical or biological differentiation of Mel from other (nonMel) episodes of depression has been questioned, and it remains unclear whether pharmacological responses proposed to be characteristic of Mel are supported by available research. Methods: We carried out a systematic review seeking treatment trials reports comparing Mel and nonMel depressed subjects for meta-analyses of their differences in responses (a) to antidepressants overall, (b) to tricyclic (TCAs) or serotonin-enhancing agents (serotonin reuptake inhibitors/serotonin–norepinephrine reuptake inhibitors) and (c) with placebo treatment. Results: We identified 25 trials in 16 reports comparing 2597 Mel with 5016 nonMel subjects. Overall, responses to antidepressant treatment did not differ between Mel (39.4%) and nonMel (42.2%) subjects. However, all subjects responded better to TCAs (50.6%) than SRIs (30.0%; p<0.0001). Mel subjects also responded less well with placebo, but also were significantly more severely depressed at intake. Conclusions: Antidepressant responses were similar in Mel and nonMel depressed patients. Mel subjects responded 25% less with placebo but were more severely depressed initially, and there was preferential response to TCAs in both Mel and nonMel subjects. The findings provide little support for proposed differences in responses to particular treatments among Mel versus nonMel depressed patients, and underscore the need to match for illness severity in making such comparisons.