Elsevier

Clinical Psychology Review

Volume 57, November 2017, Pages 208-225
Clinical Psychology Review

Review
A systematic review of person-centered approaches to investigating patterns of trauma exposure

https://doi.org/10.1016/j.cpr.2017.08.009Get rights and content

Highlights

  • Seventeen studies used latent class analysis to examine patterns of trauma exposure.

  • Individuals group together across traumatic events to form reliable classes.

  • There was some evidence that risk for psychiatric disorder differed across classes.

  • These findings lead to more accurate models of vulnerability to psychiatric disorder.

Abstract

Recent research has found that exposure to traumatic events may occur in certain patterns, rather than randomly. Person-centered analyses, and specifically latent class analysis, is becoming increasingly popular in examining patterns, or ‘classes’ of trauma exposure. This review aimed to identify whether there are consistent homogeneous subgroups of trauma-exposed individuals, and the relationship between these trauma classes and psychiatric diagnosis. A systematic review of the literature was completed using the databases EMBASE, MEDLINE (PubMed) and PsycINFO. From an initial yield of 189, 17 studies met inclusion criteria. All studies identified a group of individuals who had a higher likelihood of exposure to a wide range of traumas types, and this group consistently exhibited worse psychiatric outcomes than other groups. Studies differed in the nature of the other groups identified although there was often a class with high levels of sexual interpersonal trauma exposure, and a class with high levels of non-sexual interpersonal trauma. There was some evidence that risk for psychiatric disorder differed across these classes. Person-centered approaches to understanding the relationship between trauma exposure and mental health may offer ways to improve our understanding of the role trauma exposure plays in increasing vulnerability to psychiatric disorder.

Introduction

Exposure to traumatic events has been an interest in the psychiatric literature for many decades in recognition that there is a robust relationship between trauma exposure and mental disorders. The definition of what constitutes a traumatic event was introduced into taxonomy in the Diagnostic and Statistical Manual of Mental Disorders 3rd Edition (DSM-III) in 1980 (American Psychiatric Association, 1980), and has undergone significant developments and revisions ever since. Initially, a traumatic event was described as an event outside the range of usual experiences that is distressing to most people, and this definition was maintained in DSM-III-TR (American Psychiatric Association, 1987). The subsequent epidemiological studies which repeatedly showed that trauma exposure was common (Creamer et al., 2001, Kessler et al., 1995), led to the questioning of assumptions made about trauma exposure frequency. In response, DSM-IV (American Psychiatric Association, 1994) provided an alternative definition of a traumatic event, identifying the pathogenesis of trauma on the basis of physical harm. DSM-IV also recognized that a person could also witness or be confronted by a traumatic event, in addition to experiencing it. DSM-IV presented a broad range of events as being traumatic, although the tendency was to focus on single event, rather than cumulative or protracted events. DSM-IV also introduced a subjective component into the definition in recognition that not everyone exposed to a traumatic event necessarily develops psychopathology. Specifically, the diagnostic criteria required that the experience of the traumatic event to be coupled with fear, helplessness, or horror.

DSM-5 (American Psychiatric Association, 2013) expanded the definition of traumatic events further, by elevating sexual violence from the descriptive text into the explicit diagnostic criteria. Specifically, a traumatic event in DSM-5 is defined as exposure to actual or threatened death, serious injury or sexual violence through direct experience or witnessing. Learning about a traumatic event has been elevated from inclusion in the descriptive text to specification among diagnostic criteria, but narrowed to include only learning that a traumatic event happened to a close family member or friend. Further, if the traumatic event learned about involves actual or threatened death, the event must have been violent or accidental. The definition also includes repeated or extreme exposure to aversive details of traumatic events (such as police officers being repeatedly exposed to child abuse) which acknowledges the frequency of trauma exposure. Importantly, the subjective immediate response introduced in DSM-IV was omitted in DSM-5 because of the realization that many people develop psychopathology after trauma exposure without manifesting the specific emotional response of fear, helplessness or horror (Friedman, Resick, Bryant, & Brewin, 2011).

Until relatively recently, the relationship between trauma exposure and psychopathology focused mostly on posttraumatic stress disorder (PTSD). This was mainly due to trauma exposure being etiologically tied to the diagnostic criteria of PTSD (i.e., a traumatic event is a necessary component of the diagnostic criteria). Initially, no other psychiatric disorder had trauma exposure as a necessary diagnostic criterion. The introduction of acute stress disorder (ASD) in DSM-IV linked trauma exposure to a second disorder. Recently, there has been increasing recognition that a wide range of psychopathologic responses can be associated with trauma exposure, including anxiety and depressive disorders (Bryant et al., 2010), substance use disorders (Giordano, Ohlsson, Kendler, Sundquist, & Sundquist, 2014), psychosis (Duhig et al., 2015), suicidality (Jankovic et al., 2013), and borderline personality disorder (Westphal et al., 2013).

As our knowledge of the effects of trauma exposure has increased, so too has the recognition that not everyone who is exposed to a traumatic event will develop psychopathology. As a result, there has been increasing focus on the specific features of trauma exposure that may increase the risk of developing mental illness. For example, experiencing interpersonal violence is often associated with higher rates of PTSD compared to other types of trauma exposures (Forbes et al., 2012) a finding particularly evident for extreme forms of traumatic exposure such as rape and gender-based violence in civilian settings (Rees et al., 2011), and for acts of torture identified in populations subject to mass conflict and organized violence (Steel et al., 2009). Other studies show that the accumulation of multiple adverse and traumatic experiences impacts on mental health (Karam et al., 2014).

Research has increasingly recognized that traumatic events, more often than not, co-occur (Creamer et al., 2001, Steel et al., 1999), with much research showing that exposure to one traumatic event significantly increases risk for further exposure (Finkelhor, Turner, Ormrod, & Hamby, 2009). This suggests that exposure to multiple traumas often reflects an interrelated pattern, and research has emphasized this especially in the case of childhood trauma (Dong et al., 2004). For example, a single parent with a substance use disorder may create a familial environment in which multiple types of trauma are more likely to occur to a child, including witnessing violence, child physical abuse, and neglect. As such, traumatic experiences may not occur in isolation, and, for some people may represent an interrelated pattern of multiple traumatic experiences (Rees et al., 2011, Steel et al., 2009). The influential cumulative burden model or dose response model was based on research that showed the accumulation of traumatic events is often a stronger predictor of outcomes than single event exposure (Mollica et al., 1998, Sameroff et al., 1987). Generally, however, research has taken an additive approach to assessing trauma, where the number of different events to which an individual is exposed are summed to get a quantity of trauma exposure. This approach implicitly assumes that all traumas have equal weight, and may be overly simplistic (Hagan et al., 2016, Houston et al., 2011).

Extending cumulative trauma models to consider patterns of exposure may represent an improvement in our approach to understanding trauma exposure and its relationship to mental health. Some studies that have attempted to investigate patterns of trauma exposure have employed variable centered approaches (e.g., factor analysis) and they have shown that traumatic events may cluster in predictable ways (Bolger and Patterson, 2001, Silove et al., 2010). However, variable centered approaches are based on the assumption that the population of interest is homogenous with respect to how variables influence each other and influence outcomes (Laursen & Hoff, 2006). More recently there has been recognition that trauma-exposed populations may be heterogeneous and therefore, different groups of individuals may differ in their patterns of trauma exposure. As such, person-centered approaches (e.g., latent class analysis and cluster analysis) have become influential in looking at patterns in trauma exposure, and their subsequent impact on mental health. The unit of analyses within these approaches is the individual rather than the variables representing types of traumatic events, and as such person-centered approaches aim to identify homogeneous subgroups while acknowledging the heterogeneity of traumatic events experienced at a population level.

Latent Class Analysis (LCA) is one approach that makes it possible to take into account multiple dimensions of trauma exposure instead of focusing only on single indicators, such as number of different types of trauma exposure. LCA has been increasingly favoured in analyses investigating the impact of trauma exposure on mental health outcomes, and as such is the focal analytical approach central to this systematic review. A strength of LCA, compared with other types of person-centered analyses (e.g., cluster analysis), is that it offers a single solution based on maximum likelihood estimates and generates fit statistics which provide information about the fit between the model and the data. LCA is a statistical method for identifying unmeasured class membership among populations of interest using observed categorical variables. It divides subjects into mutually exclusive and exhaustive latent classes (Collins & Lanza, 2010). The classes are referred to as ‘latent’ because one's class membership is not directly observed or measured. Importantly, the decision about the optimal number of latent classes is based on several statistical fit indices. The likelihood ratio chi-square examines the goodness of fit between models, where a non-significant likelihood ratio chi-square indicates acceptable model fit. The Akaike information criterion (AIC, Akaike, 1987), Bayesian information criterion (BIC, Schwartz, 1978), and sample-size adjusted BIC (ssaBIC, Sclove, 1987), are all goodness of fit measures used to compare competing models with lower observed values indicating better fit. The Bootstrapped Lo-Mendell Rubin's adjusted likelihood ratio test (BLRT, Lo, Mendell, & Rubin, 2001) is used to compare models with differing numbers of latent classes. A non-significant value suggests that a model with k classes does not evidence significantly improved fit over a model with k  1 classes, indicating that the more parsimonious model should be retained. Entropy is a standardized measure of how accurately subjects are classified. Entropy values range between 0 and 1 with higher values (approaching 1) indicating better classification (Ramaswamy, Desarbo, Reibstein, & Robinson, 1993).

This review aimed to identify whether there are consistent homogeneous subgroups of trauma-exposed individuals, and the relationship between these trauma classes and mental disorder. In order to achieve this a systematic review was conducted of studies that have used person-centered analysis in the form of LCA to identify classes of trauma exposure, and the relationship between these classes and mental disorder.

Section snippets

Methodology

A systematic search of published peer-reviewed studies was conducted using a standardized and replicable process that ensured a comprehensive coverage of the literature. The systematic review search adhered to the guidelines described in the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement (Moher, Liberati, Tetzlaff, & Altman, 2010).

Description of studies

Fig. 1 presents the flow of studies through the study. The original search terms identified a total of 189 studies with the final sample reduced to 17 papers. The samples sizes ranged from large representative population-based samples (n = 7), to convenience samples such as college students or clinic attendees (n = 6) with sample size ranging from 195 to 19,125. The number of traumatic events assessed in each study ranged from seven to 51. Studies were conducted with both adolescent (n = 7) and adult

Discussion

All the studies in this review shared one common outcome - they identified qualitatively different patterns of exposure to traumatic life events. These studies confirm that although traumatic events are heterogeneous in nature, homogenous subgroups of individuals can be identified that experience a similar range of traumatic events. The finding that trauma exposure forms distinct classes confirms earlier reports that trauma exposure is not randomly distributed in the population, despite

Conclusion

This research shows that inviduals group together across types of exposures to traumatic events to form reliable classes. This represents both an exciting direction for trauma research and a challenge to the trauma field to take a more nuanced approach to classifying trauma history. The opportunity for us is that we can develop more accurate models of vulnerability to better describe the burden that trauma places on the individual.

Role of funding sources

Funding for this study was provided by Humboldt Senior Researcher Award 2016 (MOD) and a National Health and Medical Research Council Program Grant (568970). The funding sources had no role in the study design, collection, analysis or interpretation of the data, writing the manuscript or the decision to submit the paper for publication.

Contributors

Author MOD developed the concept for the study. Author TV conducted the literature searches. Authors MOD, TV, and DK screened and provided summaries of the studies. Author MOD wrote the first draft of the manuscript and all authors contributed to and approved the final manuscript.

Conflict of interest

All authors declare no conflict of interest.

Acknowledgements

The authors wish to thank Olivia Metcalf for her assistance in proofreading and administration of the submission process.

References (56)

  • American Psychiatric Association

    Diagnostic and statistical manual of mental disorders

    (1994)
  • American Psychiatric Association

    Diagnostic and statistical manual of mental disorders

    (2013)
  • E.D. Ballard et al.

    Latent classes of childhood trauma exposure predict the development of behavioral health outcomes in adolescence and young adulthood

    Psychological Medicine

    (2015)
  • S.R. Berzenski et al.

    Classes and consequences of multiple maltreatment: A person-centered analysis

    (2011)
  • K.E. Bolger et al.

    Developmental pathways from child maltreatment to peer rejection

    (2001)
  • N. Breslau et al.

    Trauma and posttraumatic stress disorder in the community

    Archives of General Psychiatry

    (1998)
  • R.A. Bryant et al.

    The psychiatric sequelae of traumatic injury

    American Journal of Psychiatry

    (2010)
  • C.E. Cavanaugh et al.

    Mental disorders associated with subpopulations of women affected by violence and abuse

    Journal of Traumatic Stress

    (2013)
  • C.E. Cavanaugh et al.

    Patterns of violence against women: A latent class analysis

    Psychological Trauma-Theory Research Practice and Policy

    (2012)
  • L.M. Collins et al.

    Latent class and latent transition analysis: with applications in the social, behavioral and health sciences

    (2010)
  • M. Creamer et al.

    Post-traumatic stress disorder: Findings from the Australian National Survey of Mental Health and Well-being

    Psychological Medicine

    (2001)
  • M. Duhig et al.

    The prevalence and correlates of childhood trauma in patients with early psychosis

    Australian and New Zealand Journal of Psychiatry

    (2015)
  • D. Finkelhor et al.

    Pathways to poly-victimization

    Child Maltreatment

    (2009)
  • D. Finkelhor et al.

    Violence, abuse, and crime exposure in a national sample of children and youth

    Pediatrics

    (2009)
  • D. Forbes et al.

    Trauma at the hands of another: Longtitudinal study of differences in the posttraumatic stress disorder symptom profile following interpersonal compared with noninterpersonal trauma

    Journal of Clinical Psychiatry

    (2012)
  • D. Forbes et al.

    Trauma at the hands of another part 2: Distinguishing PTSD patterns following intimate and non-intimate interpersonal and non-interpersonal trauma in a nationally representative sample

    Journal of Clinical Psychiatry

    (2014)
  • M.J. Friedman et al.

    Considering PTSD for DSM-5

    Depression and Anxiety

    (2011)
  • G.N. Giordano et al.

    Unexpected adverse childhood experiences and subsequent drug use disorder: A Swedish population study (1995–2011)

    Addiction

    (2014)
  • Cited by (93)

    View all citing articles on Scopus
    View full text