Application of cluster analysis in the classification of depression. A replication study

Area of Science:

• Psychiatry
• Clinical Psychology
• Psychopathology

Background:

• Patient typologies are crucial for understanding depression heterogeneity.
• Previous cluster analyses identified distinct subgroups of depressed patients.
• Replication studies are essential for validating typological findings.

Purpose of the Study:

• To replicate a previously established four-group cluster-analytic typology of depressed patients.
• To assess the stability and generalizability of depression subtypes.
• To refine patient classification in depressive disorders.

Main Methods:

• Applied hierarchical agglomerative clustering procedures to a patient sample.
• Utilized Ward's error sum of squares method for optimal clustering.
• Employed iterative relocation to ensure cluster stability.

Main Results:

• Successfully replicated three patient depression groups: psychotic, anxious, and young depressives with personality disorders.
• The hostile depressives group did not consistently separate from the young depressives group.
• Ward's method demonstrated robust replication under iterative relocation.

Conclusions:

• The cluster-analytic typology shows partial replicability in depressed patients.
• Psychotic, anxious, and young depressives with personality disorders represent stable subtypes.
• Further investigation is needed to clarify the 'hostile depressives' subtype.