Cluster Analysis and the Classification of Depression

Abstract

During recent years there has been considerable controversy concerning the classification of depressive disorders. Investigators disagree as to whether depressive disorders are a single phenomenon with varying degrees of severity or are a group of two or more discrete illnesses with differing courses, prognoses, responses to treatment, familial backgrounds, aetiologies, and clinical pictures (Eysenck, 1970; Kendell, 1976; Foulds and Bedford, 1976; Andreasen and Winokur, 1979). They also disagree about the boundaries between depressive disorders and anxiety disorders, schizophrenia, personality disorders, and normality (Rothet al, 1972; Kayet al, 1976; Kendell and Gourlay, 1970a and b; Tsuanget al, 1976). There is no consensus as to what constitutes adequate validation of a nosological class; a variety of independent validators have been proposed, including genetic factors, outcome, responses to treatment, and neurochemical or neurophysiological markers (Guze, 1978; Blashfield and Draguns, 1976; Kupferet al, 1975; Reider and Gershon, 1978).