DSM-III Personality Disorder Clusters in Three Populations

Abstract

DSM-III personality disorder clusters had not previously been tested for their ability to distinguish between different population groups. A major obstacle - different demographic characteristics of different populations - was overcome by using logistic regression. Three out-patient psychiatric referral groups (chronic pain, disability applicants and competency to stand trial) provided subjects; adjustments were made for age, race, sex and education. The groups were significantly different in the adjusted prevalence of DSM-III personality disorder clusters. Clinical association between personality disorder and drug or alcohol abuse was again confirmed.