Running head: COMPLEX PTSD DIAGNOSIS IN HOMELESS ADULTS

The 11th revision of the International Classification of Diseases (ICD-11), ratified at the World Health Assembly in May 2019, introduced revised diagnostic guidelines for posttraumatic stress disorder (PTSD) as well as a separate diagnosis of complex PTSD (CPTSD). We aimed to test the new ICD-11 symptom structure for PTSD and CPTSD in a sample of individuals who have experienced homelessness. Experiences of trauma exposure and the associated mental health outcomes have been underresearched in this population. A COMPLEX PTSD DIAGNOSIS IN HOMELESS ADULTS 2 This article is protected by copyright. All rights reserved. sample of adults experiencing homelessness (N = 206) completed structured and semistructured interviews that collected information about trauma exposure and symptoms of PTSD and CPTSD. We conducted a latent class analysis (LCA) using six symptom clusters (three PTSD symptom clusters that are components of CPTSD and three CPTSD symptom clusters). All participants reported trauma exposure, with 88.6% having experienced at least one event before 16 years of age. Four distinct classes of participants emerged in relation to the potential to meet the diagnosis: LCA CPTSD (n = 122, 59.8%), LCA no diagnosis (n = 27: 13.2%), LCA PTSD (n = 33; 16.2%), and LCA disturbance in self-organization (DSO; n = 22; 10.8%). Of note, participants with an ICD-11 CPTSD as well as those with an ICD-11 PTSD diagnosis fell into the LCA CPTSD class. Our findings provide some support for the distinction between CPTSD and PTSD within this population specifically but potentially have broader implications. Clear diagnoses will allow targeted PTSD and CPTSD treatment development. Using Latent Class Analysis to Support the ICD-11 Complex Posttraumatic Stress Disorder Diagnosis in a Sample of Homeless Adults Complex posttraumatic stress disorder (Complex PTSD) is a separate diagnosis recently included in the new 11th revision of the International Classification of Diseases (ICD-11; World Health Organization [WHO], 2018), which was released in June 2018. The ICD-11 offers significant changes from 10th revision (ICD-10) in the classification of traumatic stress–related pathology. For example, posttraumatic stress disorder (PTSD) was previously categorized with neuroticism, somatoform, and adjustment disorders. “Disorders specifically associated with stress” is the new classification that independently constitutes disorders that are of consequence to experiencing stress. Posttraumatic stress disorder, with a revised item structure, as well as the new inclusion of complex PTSD (CPTSD) are diagnoses COMPLEX PTSD DIAGNOSIS IN HOMELESS ADULTS 3 This article is protected by copyright. All rights reserved. that represent the extreme presentations in this category (World Health Organization [WHO], 2018). A CPTSD diagnosis is made when an individual meets the diagnostic requirements for PTSD and disturbances in self-organization (DSO) are present at any point during the course of the disorder. Previously, DSO symptoms were included in the diagnostic guidelines for “enduring personality change after catastrophic experience” (EPCACE) in the ICD-10 (WHO, 1993). This new category in the ICD-11 consolidates trauma-related symptomology and offers the first formalized diagnosis of CPTSD (Cloitre, Garvert, Brewin, Bryant, & Maercker, 2013; Maercker, Brewin, Bryant, Cloitre, Ommeren, et al., 2013; Maercker, Brewin, Bryant, Cloitre, Reed, et al., 2013). The ICD-11 PTSD diagnostic guidelines include three symptom clusters: (a) reexperiencing the traumatic event or events in the present (e.g., vivid memories or flashbacks); (b) avoidance of thoughts and memories of the event or events or avoidance of activities, situations, or people reminiscent of the event or events; and (c) persistent perceptions of heightened current threat (e.g., hypervigilance or an enhanced startle reaction to stimuli such as unexpected noises). Symptoms must persist for at least several weeks and cause significant impairment in personal, family, social, educational, occupational, or other important areas of functioning. For a diagnosis of CPTSD, all PTSD core symptoms must be present at some point during the course of CPTSD and three additional symptom clusters specific to CPTSD must also be present: (a) severe and pervasive problems in affect regulation (i.e., affect dysregulation; AD); (b) persistent beliefs about oneself as diminished, defeated or worthless, accompanied by deep and pervasive feelings of shame, guilt or failure related to the traumatic event (i.e., negative self-concept; NSC); and (c) persistent difficulties in sustaining relationships and feeling close to others (i.e., disturbances in relationships; DR). An individual is considered to have met the diagnostic criteria for a symptom cluster when they endorse at least one of two items in the cluster (World Health Organization, 2018). COMPLEX PTSD DIAGNOSIS IN HOMELESS ADULTS 4 This article is protected by copyright. All rights reserved. The development of the CPTSD construct was initially driven by clinicians who recognized the need to adequately care for individuals who presented with persistent and disabling AD, NSC, and DR, collectively identified as DSO. The momentum for recognition of an associated diagnosis has been building for more than 20 years (Beltran, Silove, & Llewellyn, 2009; Courtois & Ford, 2009; Ford, Stockton, Kaltman, & Green, 2006; Herman, 1992). The ICD-11 Working Group recognized this field of research in developing diagnostic guidelines for CPTSD. In the time since the ICD-11 Working Group released their proposed diagnostic guidelines for CPTSD, there has been a growing body of research focused on the theoretical separation of CPTSD from PTSD (Cloitre et al., 2013; Wolf et al., 2015) across samples exposed to different trauma types, such as (e.g. bereavement, sexual victimization, and physical assault: Elklit, Hyland, & Shevlin, 2014; e.g. childhood institutional abuse: Knefel, Garvert, Cloitre, & Lueger-Schuster, 2015) as well as in non-Western samples (Murphy, Elklit, Dokkedahl, & Shevlin, 2016). This research has predominantly used latent profile analyses (LPA) and latent class analyses (LCA). These analytic methods are used to cluster individuals into groups based on patterns of symptom response on continuous measures for LPA and categorical measures for LCA (Muthén & Muthén, 2000). If CPTSD is a separate construct from PTSD, as is described by the new ICD definition, then separate CPTSD and PTSD classes would be expected to emerge from clustering. If they are part of the same construct, profiles would likely diverge regarding severity scores or endorsement probability across all items rather than along phenomenology lines. Cloitre et al. (2013) ran an LPA and found three distinct diagnostic profiles in a sample of 302 individuals seeking treatment for interpersonal trauma exposure: PTSD (31.8%), CPTSD (36.1%), and a low-symptom profile (32.1%). Using LCA, Elklit et al. (2014) demonstrated further support for CPTSD as a separate diagnosis in a mixed-trauma–exposed COMPLEX PTSD DIAGNOSIS IN HOMELESS ADULTS 5 This article is protected by copyright. All rights reserved. community sample of 1,251participants; this study found a CPTSD class (13.0%), a PTSD class (33.6%), and a class that had a low probability of endorsing either diagnosis (53.4%). In contrast, Wolf et al. (2015) initially replicated the results reported by Cloitre et al. (2013), using LPA with their sample of 323 trauma-exposed, treatment-seeking veterans. However, the authors also investigated the theoretical framework of the clusters, using applied factor mixture modeling (FMM), which is a combination of latent class analysis and standard factor analysis (Lubke & Muthén, 2007) and groups symptoms into classes. Response patterns are then considered in conjunction with the symptom structure of those response patterns (Kramer et al., 2016). An LCA does not include the simultaneous consideration of the latent symptom structure (i.e., the factor-analytic component). In this case, the FMM failed to find differentiation between the PTSD and CPTSD and instead found a model of severity in which the level of symptom endorsement positively increased in line with other symptom severity endorsement across both disorders (Wolf et al., 2015). The authors concluded from their findings that CPTSD was a subtype of PTSD rather than a distinct disorder. More recently, Murphy et al. (2016) compared two samples of young Northern Ugandan adults (N = 314; n = 124 former child soldiers, n = 190 civilians) using LCA to test the ICD-11 diagnostic algorithm. Their results also separated participants into three classes: a CPTSD class (40.3%), a PTSD class (43.1%), and a low-symptom class (16.6%). The former child soldiers were five times more likely to be grouped into the CPTSD class than in the low-symptom class, odds ratio (OR) = 5.43, 95% CI [2.44, 12.09]. Knefel and colleagues (2015) found a variation of the three-class solution in a sample of 229 Austrian adult survivors of childhood institutional abuse, which included physical, sexual, and emotional maltreatment. The authors identified a four-class LPA solution that included a group with elevated symptoms of CPTSD (20.1%), a PTSD group (17.5%), a group with elevated DSO symptoms and some elevated PTSD symptoms (19.2%), and a COMPLEX PTSD DIAGNOSIS IN HOMELESS ADULTS 6 This article is protected by copyright. All rights reserved. group with low levels of symptoms (43.2%). Although the authors demonstrated a similar separation of PTSD and CPTSD as was shown in the other investigations, the third class (i.e., elevated DSO with some elevated PTSD symptoms) represented a substantial minority of the sample. The authors argued that the most likely reason for this deviation from the three-class solutions found in similar studies rested in the characteristics of the sample, as participants had universally