Classification of Depressed Patients: A Cluster Analysis Derived Grouping

TLDR: The emergence of these multiple and easily interpretable groups of depressed patients suggests that non-psychotic depressives are diverse and comprise several groups, and suggests the utility of cluster analysis grouping techniques for studying the classification of psychiatric patients.

Abstract: One hundred and sixty-five depressed patients from a variety of treatment settings were rated as to clinical symptoms, previous history, life stress, and premorbid neuroticism, and the ratings were subjected to a special cluster analysis for classifying individuals. The results indicated a hierarchical structure of groups. At the first level there were two groups, one older and more severely ill, one younger and more mildly ill. Subsequent division of each of these produced four approximate groups. One group was characterized by a severe illness, sometimes with delusions, on a background of good premorbid adjustment, and corresponded to psychotic depressives as described in the literature. A second group consisted of moderately depressed patients with a strong admixture of anxiety, a high incidence of previous illnesses and high neuroticism scores. A third group comprised depressives with a considerable element of hostility; and the last group contained young patients whose relatively mild illnesses developed on a background of personality disorder. These groups resemble other groups of depressed patients previously described in the literature. They are also consistent with the psychotic-neurotic dichotomy, but suggest that non-psychotic depressives are diverse and comprise several groups. The emergence of these multiple and easily interpretable groups of depressed patients also suggests the utility of suck cluster analysis grouping techniques for studying the classification of psychiatric patients.