Combining novel trait and neurocognitive frameworks to parse heterogeneity in borderline personality disorder

Borderline Personality Disorder (BPD) diagnosis comprises several constellations of trait, neurocognitive, and psychosocial alterations. Dimensional models of psychopathology provide new opportunities to parse heterogeneity and create a stronger interface between individual characterization and psychosocial outcomes. However, dimensional models have focused on either traits or neurocognitive features, lacking integration to capture the multifaceted nature of BPD.

We assessed 100 participants with BPD using a combination of tools stemming from trait (Alternative Model for Personality Disorders) and neurocognitive models (Research Domain Criteria; RDoC) to examine if trait-derived subgroups display distinctive social-processing and psychosocial profiles. We used two complementary analytical approaches: person-centered (k-means clustering) and construct-based (multiple factor analysis).

Our person-centered approach identified four subgroups with separable internalizing, detached, externalizing, and low psychopathology trait profiles. These profiles revealed distinctive patterns of affiliation, emotion recognition and mentalization performance in RDoC tasks, and psychosocial measures of quality of life and social connectedness. RDoC-based measures showed close construct proximity with negative affectivity, disinhibition, and antagonism trait domains, relative to the detachment domain, which had close proximity with self-knowledge.

Altogether, findings support consilience between trait-based and neurobiological frameworks and suggest that trait models are useful to parse BPD heterogeneity leading to unique social functioning profiles.