| Factor analytic work demonstrates that adult psychiatric comorbidity (i.e., the co-occurrence of two or more disorders) is explained by latent variables termed internalizing and externalizing. Posttraumatic stress disorder (PTSD) evidences a particularly diverse pattern of comorbidities across these spectra and across Axes I and II of the fourth edition of the Diagnostic and Statistical Manual (DSM-IV), however, no study has examined if these dimensions might account for PTSD-Axis II comorbidity and no study has broadly examined the applicability of this model to personality disorders (PDs). This study addressed this gap in the literature by evaluating an internalizing/externalizing model of PTSD-Axis II disorder comorbidity. Participants were 245 veterans and non-veterans with PTSD who completed self-report measures of PTSD (PTSD Checklist), trauma history (Traumatic Life Events Questionnaire), personality (Brief Form of the Multidimensional Personality Questionnaire), and PD (Schedule for Nonadaptive and Adaptive Personality; SNAP). Following a series of preliminary analyses, a modified bifactor confirmatory factor analysis was fit to 35 SNAP parcels, created from the 10 DSM-IV SNAP PD scales. A model testing series provided evidence of a hierarchical structure in which nine lower-order common factors, indexing pathology ranging from aggression to dependency, accounted for covariance across subsets of SNAP parcels; the correlations among these factors were due to higher-order Internalizing and Externalizing PD common factors. Further, a general factor, reflecting boundary disturbances, accounted for additional covariance of nearly all the indicators, consistent with theory regarding the centrality of self-other relationship disturbances across Axis II. Structural equation modeling revealed that trait negative emotionality predicted both internalizing and externalizing, while low positive emotionality was specific to the former and low constraint specific to the latter; demographic and trauma-related covariates were also differentially associated with the factors. This work suggests continuity in the underlying structure of psychopathology across Axis I and II and raises questions regarding the organization of the Axis II disorders in DSM-IV. It also provides empirical evidence of a common disturbance in how individuals with PD features relate to others, a disturbance that may reflect the maladaptive poles of core, normal-range, and biological drives for affiliation and agency. |
