Disorders of personality or disorders of relationships? A comprehensive examination of two competing hypotheses on the nature of personality pathology

Contemporary perspectives view personality disorder as a disorder of either all aspects personality (that is, maladaptive ways of ‘thinking, feeling, behaving, and relating’) or mainly some aspects of personality (that is, maladaptive ways of ‘relating’). In this study, we examine these competing hypotheses by quantifying the diagnostic importance of 124 personality symptoms in three large representative samples of the United Kingdom population: namely, the Adult Psychiatric Morbidity Surveys of 2000 (APMS2000; N= 8,580), 2007 (APMS2007; N=7,403), and 2014 (APMS2014; N=7,546). To identify the most important diagnostic predictors, we regress diagnostic outcomes (presence vs. absence of personality disorder) on all DSM-IV personality symptoms: first in random forests and then in logistic regression models. Our results show that the strongest predictors of a ‘personality disorder’ diagnosis are self-harming, social suspiciousness, and intimacy problems (in APMS2000); self-harming, unstable relationships, conduct problems, and fears of abandonment (in APMS2007); social suspiciousness, impulsiveness, difficulty making and keeping friends, and being overly dependent on others (in APMS2014). These results support recent perspectives that view personality disorders as relational disorders by showing that their diagnosis is primarily determined by maladaptive ways of relating to oneself (e.g., self-harming) and others (e.g., intimacy problems).