Background: There is an ongoing research effort to test if depression is a homogeneous clinical syndrome and to identify valid and useful subtypes based on the number and nature of depressive symptoms. This study summarizes the patterns of depressive symptoms evident in a prospective study of the general population and examines the validity of potential subtypes by studying their course and etiologic heterogeneity.
Methods: A general population sample of 1920 adults (aged 18-96) from the Baltimore Epidemiologic Catchment Area (ECA) follow-up study (1981 to 1993/6) were examined. Data on diagnoses, symptoms, course and risk factors were collected using the Diagnostic Interview Schedule (DIS). Latent class analysis was applied to summarize symptom patterns. Course characteristics and risk factor profiles were compared among potential subtypes based on the number of symptom groups or symptom patterns. Logistic regression models were used to examine the etiologic heterogeneity among potential subtypes based on symptoms.
Results: The number of symptom groups gave the most efficient insight into differential etiologic processes. Severe depression (7-9 symptom groups) was associated with female gender, family history of depression but not with stressful life events before the onset of the first episode. Moderate (5-6 symptom groups) and mild depression (3-4 symptom groups) were associated with family history of depression, stressful life events before the onset, but not with female gender. The latent class model generated patterns of depressive psychopathology as follows: anhedonia, suicidal, psychomotor, and severely depressed subtypes. The Anhedonia subtype showed a course and risk factor profile distinct from the others.
Limitations: The measurement of psychopathology was based on self-reported DIS interviews instead of psychiatric assessments. Recall or report bias cannot be excluded in the ascertainment of family history and stressful life events.
Conclusions: Depression is heterogeneous, even below the threshold of syndromal diagnosis. The severity of an episode appears to be more informative than the pattern of symptoms, with the possible exception of a putative anhedonic subtype.