Diagnosis and Classification Subtyping of Depressive Disorders: Comparison of Three Methods.

Objective: Melancholic versus non-melancholic depression dichotomy is perhaps the most widely accepted distinction in categorization of depression. This research aims to compare symptom based, severity based, and biology based categorization of depression. Methods: To achieve this, the cluster analysis was performed on a sample of 78 depressed patients, first by using 14 SCID-I depressive symptoms. Patients were clustered again with regard to post dexhametasone cortisol levels (suppressed vs non-suppressed groups), and lastly according to HDRS (Hamilton Depression Rating Sacale) scores (high vs low severity groups). Biological (thyroid stimulating hormone -TSH, basal and post dexhametasone cortisol levels), clinical (age, age of onset, severity of depression, psychosocial stressors, and personality disorder) and demographic variables of these categories were compared. Results: There was a high degree of accordance between the cluster analytically derived endogenous group and the DSM-IV diagnosis of melancholia. Cluster analytically generated endogenous group were older, more severely depressed, and had higher basal cortisol levels than non-endogenous depressive subtype. Severely depressed group according to HDRS scores had lower TSH levels. Only DST (dexamethasone suppression test) non-suppressive patients had more depressive disorders in their family. Conclusions: The results of our study partly support the hypothesis that endogenous and melancholic depression have distinct clinical and biological features. [ABSTRACT FROM AUTHOR]