The validity of trait, symptom and prototype approaches for describing borderline and antisocial personality disorders

Determining the relative validity of trait, symptom and prototype approaches to rating personality disorders (PDs) is a pressing concern given that each of these approaches has been considered for the DSM-5. However, as of yet, there is limited empirical evidence to guide this decision. In this study, we compared the proposed DSM-5 traits, DSM-IV symptoms and diagnostic prototypes that were initially slated for inclusion in the DSM-5 for assessing antisocial and borderline PDs. Trait and symptom count models converged with one another within diagnoses more than either did with the prototypes but also showed more overlap across diagnoses than the prototypes. The traits generally showed superior criterion validity relative to the symptoms, which were superior to the prototypes, with the exception that symptoms were more valid than traits for predicting externalizing problems. This pattern suggests that traits are the single most effective method for describing antisocial and borderline PDs but that combining traits and symptoms, as currently proposed for the DSM-5, may be the best combination of these three approaches to assessing personality pathology. Copyright (c) 2012 John Wiley & Sons, Ltd.