Differentiating Abnormal, Normal, and Ideal Personality Profiles in Multidimensional Spaces

Abstract

Current dimensional taxonomies of personality disorder (PD) establish that intense traits do not suffice to diagnose a disorder, and additional constructs reflecting dysfunction are required. However, traits appear able to predict maladaptation by themselves, which might avoid duplications and simplify diagnosis. On the other hand, if trait-based diagnoses are feasible, it is the whole personality profile that should be considered, rather than individual traits. This takes us into multidimensional spaces, which have their own particular - but poorly understood - logic. The present study examines how profile-level differences between normal and disordered subjects can be used for diagnosis. The Dimensional Assessment of Personality Pathology - Basic Questionnaire (DAPP-BQ) and the Personality Inventory for DSM-5 (PID-5) were administered to a community and a clinical sample each (total n = 1,925 and 3,543 respectively). Intense traits proved to be common in the general population, so empirically-based thresholds are indispensable not to take as abnormal what is at most unideal. Profile-level parameters such as Euclidean and Mahalanobis distances outperformed individual traits in predicting mental problems and equaled the performance of published measures of dysfunction or severity. Personality profiles can play a more central role in identifying disorders than is currently acknowledged, provided that adequate metrics are used.