In the efficacy evaluation of antidepressant treatments the total score of the Hamilton Depression Rating Scale (HAMD) is still regarded as the’gold standard’. Studies suggest that unidimensional subscales of the HAMD, which capture the core depressive symptoms, outperform the full HAMD regarding the detection of antidepressant treatment effects. the present study compared several unidimensional subscales of the HAMD and the Inventory of Depressive Symptomatology (IDS) regarding their sensitivity to changes in depression symptoms in a sample of patients with minor depression (MIND). Biweekly IDS-C28 and HAMD17 data from 287 patients of a 10-week randomised, placebo-controlled trial comparing the effectiveness of sertraline and cognitive-behavioural group therapy in patients with MIND were converted to subscale scores and analysed during the antidepressant treatment course. We investigated sensitivity to depressive change for all scales from assessment-to-assessment, in relation to depression severity level and placebo-verum differences.

The subscales performed similarly during the treatment course, with slight advantages for some subscales in detecting treatment effects depending on the treatment modality and on the items included. Most changes in depressive symptomatology were detected by the IDS short scale, but regarding effect sizes it performed worse than most subscales. Unidimensional subscales are a time- and cost-saving option in judging drug therapy outcomes, especially in antidepressant treatment efficacy studies. However, subscales do not cover all facets of depression, which might be important for comprehensively understanding the nature of the disease depression. Therefore, the cost-to-benefit ratio must be carefully assessed in the decision for using unidimensional subscales.