This was a retrospective study in which 50 hospitalized depressed women were rated on 25 items of symptomatology and these items were then subjected to factor analysis. Five factors were extracted and attention was focused on the first factor. Items which loaded heavily on this factor included insomnia, especially in the middle of the night and early morning, severity of depressed mood and global severity of illness, retardation, guilt and self-reproach, weight loss, delusional symptoms, visceral symptoms, agitation, and loss of interest. The symptoms and signs which loaded heavily thus fit the classical pattern of endogenous depression. The patients were scored on this factor and these factor scores were then correlated with the patients' ratings on premorbid personality, personal history, and evidence of apparent precipitant, items which were independent of the factor analysis. The factor scores were found to correlate negatively with presence of apparent precipitants and negatively with hysterical character. The distribution curve of factor scores suggested a distinct group of patients in a hump at the end of the curve. Corresponding factors in other studies were shown to approximately reproduce the items loading on this factor. When applied to our patients these factors produced factor scores which correlated highly with our factor scores and selected the patients in substantially the same distribution. We are currently engaged in investigating the characteristics of our endogenous group of patients to determine how they differ from the remainder of our population. We expect the current studies to lead to a prospective study in which we will take up problems of rater reliability and study a larger patient sample.
Typus melancholicus is not the premorbid personality trait of unipolar (endogenous) depression