Posttraumatic stress disorder (PTSD) consists of four main components in the Diagnostic and Statistical Manual of Mental Disorders (5th ed.; DSM-5): intrusions (INT), avoidance (AV), negative alterations in cognition and mood (NACM), and arousal and reactivity (AAR); however, studies do not always support this four-factor model. A sample of 348 treatment-seeking adults was assessed for PTSD symptoms at baseline (Time 1) and then 12 weeks later (Time 2). Confirmatory factor analysis (CFA) was used to examine the model fit of the DSM-5 four-factor model of PTSD with and without a general factor at both time points, and structural equation modeling allowed for examination of these associations between time points. The four-factor model did not meet the criteria for excellent model fit, and the bifactor model provided improved model fit. The NACM specific factor did not meet the criteria for unique variance above and beyond the general factor. The bifactor model of PTSD symptoms was reliable over time, and both the general factor and the AAR factor significantly predicted subsequent symptom severity.
Keywords: bifactor; confirmatory factor analysis; factor structure; posttraumatic stress disorder; structural equation modeling.