Elsevier

Psychiatry Research

Volume 228, Issue 3, 30 August 2015, Pages 626-632
Psychiatry Research

Trauma and psychosis: The mediating role of self-concept clarity and dissociation

https://doi.org/10.1016/j.psychres.2015.04.053Get rights and content

Highlights

  • Childhood traumas were compared in early episode psychosis patients and controls.

  • The patients experienced more trauma in childhood than controls.

  • The relationship between trauma and psychosis was mediated by self-concept clarity.

  • A coherent conceptualisation of the self may be important in recovery from psychosis.

Abstract

Childhood trauma (CT) and psychosis may be associated. Drawing on the dissociation and social psychological literature, the current study examined the mediating role of structural aspects of self in explaining the relationship between childhood trauma and psychosis. Twenty-nine individuals with psychosis were compared with 31 healthy volunteers regarding childhood trauma, dissociation and self-concept clarity (SCC). High rates of maltreatment were found in the psychosis sample. Additionally, clinical participants reported more dissociation and less self-concept clarity. Mediational analyses were carried out on pooled data from across both clinical and non-clinical samples. These suggested that the influence of physical neglect in increasing the likelihood of experiencing psychosis was explicable through the effects of increased dissociation. Self-concept clarity mediated the relationship between psychosis and total childhood trauma, emotional abuse, physical abuse, emotional and physical neglect. Furthermore, dissociation and self-concept clarity were strongly correlated providing evidence that they may form a unitary underlying concept of ‘self-concept integration’. The study provides further evidence of the link between childhood trauma and psychosis. Self-concept integration may be adversely affected by negative childhood experiences, which increases psychosis risk. Methodological limitations, clinical implications and suggestions for future research are considered.

Introduction

One potentially important aspect of the development of psychosis is childhood trauma (CT), incorporating experiences of sexual, physical and emotional abuse and neglect (Read et al., 2008).Childhood adversity is associated with psychosis risk (Varese et al., 2012b) and such risk may increase as a function of the frequency and/or cumulative types of CT (Janssen et al., 2004). Such ‘dose-response’ effects support the hypothesis that CT may be causally related to psychosis (Read et al., 2008). Plausible mechanisms have been proposed to explain links between CT and psychosis, and there is a growing movement towards investigating these. One influential theory in this area is the traumagenic neurodevelopmental model, which suggests that CT may underpin psychosis vulnerability via its effects on the developing brain (Read et al., 2001).

Dissociation may mediate the relationship between CT and psychosis (Hammersley et al., 2008), with dissociation defined as ‘a disruption in the usually integrated functions of consciousness, memory, identity, or perception of the environment’ (American Psychiatric Association, 2000). There is widespread acceptance of a traumagenic causal model of dissociation, and evidence suggests that dissociation may partially mediate the relationship between trauma and subsequent psychopathology (Gershuny et al., 2004). In psychosis, the pathway to adult hallucinations, delusions and dissociative symptoms may begin with a dissociative response to CT (Read et al., 2001). As dissociation causes impaired reality testing, individuals who cope with trauma via dissociation are more likely to subsequently experience psychosis, particularly hallucinations (Kilcommons and Morrison, 2005). Recent research has found the relationship between various forms of CT and hallucination proneness to be positively mediated by dissociation, particularly regarding sexual abuse (Varese et al., 2012a).

The concept of ‘self’ in psychosis has a long and complex history, with attempts by some early psychiatrists to understand diagnoses such as schizophrenia as essentially ‘disorders of self’ (Berrios and Markovă, 2003). There now exists an array of theories and measures within social psychology regarding the self (Baumeister, 1999), although these have received limited attention in psychosis research. ‘Self’ can be described as ‘…an organised and interactive system of thoughts, feelings, identities, and motives that (1) is born of self-reflexivity and language, (2) people attribute to themselves, and (3) characterise specific human beings’ (Owens, 2003). Self may be important in relation to dissociation and psychosis vulnerability. For example:

‘…dissociative detachment renders individuals vulnerable to psychosis not only because it deprives them of external anchors. In addition, it robs them of internal anchors - the sense of being connected to one׳s body, a sense of self or identity, and one׳s own actions. The result may be not only profoundly impaired reality-testing but also severe confusion, disorganization, and disorientation’ (Allen et al., 1997).

One fundamental aspect of self is that of ‘self-concept’, defined broadly as ‘the totality of an individual׳s thoughts and feelings with reference to himself as an object’ (Rosenberg, 1979). Self-concept is viewed not as a unitary entity, but rather a multi-dimensional and multi-faceted dynamic system (Baumeister, 1999, Markus and Wurf, 1987). Theory about the self has three aspects: Content, structure and process (Stopa, 2009). Little attention has been given to process and structural aspects of the self, with most literature focussing upon content (Campbell et al., 2003, Rafaeli-Mor and Steinberg, 2002, Stopa, 2009). Considering self purely in terms of content, whilst important, is potentially limiting. Focusing on the structural organisation of the self ‘. will result in a more comprehensive understanding of the role of the self-concept in psychopathology and behavioural change.’ (Stein and Markus, 1994)

The relationship between self-concept structure and psychological well-being and adjustment has been examined from several perspectives. One approach suggests that a ‘divided self’ (i.e. a self-concept which lacks integration) is an important precursor to the development of psychological problems (Lutz and Ross, 2003). It has been suggested that a damaged self-concept may confer risk of psychosis (Bell and Wittkowski, 2009). Indeed, phenomenological approaches have implicated a disturbed basic sense of self in prodromal psychosis (Nelson et al., 2009), and evidence suggests that a lack of unity within the self-concept structure may be related to psychosis-like experiences (Preston, 2008).

One line of research which examines structural integration/unity in the self-concept and its relationship to psychological adjustment is that relating to self-concept clarity (SCC) (Campbell et al., 1996, Campbell et al., 2003). Self-concept clarity has been defined as ‘the extent to which the contents of the self-concept are clearly and confidently defined, internally consistent, and temporally stable’ (Campbell et al., 1996). There is emerging evidence that SCC is important in relation to psychosis. A series of recent studies found that high aberrant salience and low self-concept clarity interacted to predict psychosis-like experiences (Cicero et al., 2013). In another study investigating SCC and psychosis-like experiences in the general population, SCC accounted for significant variance in delusional beliefs, hallucination proneness, impulsive non-conformity and unusual experiences; these relationships were partially mediated via depression and anxiety (Preston, 2008). In summary, there is some tentative evidence that low SCC may be related to psychosis; however, as yet there has been no attempt to examine SCC as a mediator in the relationship between childhood trauma and psychosis.

It is suggested that trauma disrupts the development of an integrated self-concept, leading to increased vulnerability to psychosis. Furthermore, there exists a preliminary theoretical and empirical basis for suggesting that levels of dissociation would be inversely associated with self-concept integration, potentially tapping similar underlying aspects of self-concept structure (Lutz and Ross, 2003, Pollack et al., 2001). However, to date no research has examined this relationship in a psychosis sample. This study aimed to investigate childhood trauma, dissociation and self-concept clarity across a clinical and non-clinical group. A specific aim was to examine the extent to which dissociation and self-concept clarity mediate the relationship between childhood trauma and psychosis.

The hypotheses for this study were that participants in a psychosis sample would have experienced more childhood trauma and dissociation than those in a non-clinical sample and report lower self-concept clarity. It was also hypothesised that childhood trauma would be positively correlated with dissociative experiences, and negatively correlated with self-concept clarity; furthermore, dissociative experiences would be negatively correlated with self-concept clarity. Lastly, a model was tested whereby self-concept clarity and dissociation were examined as mediators in the relationship between childhood trauma and psychosis.

Section snippets

Clinical participants

Clinical participants were recruited from five Early Intervention in Psychosis (EIP) teams. All participants were assessed by the service as having transitioned to first episode psychosis, were within 3 years of the first treated episode and aged 18–38. Eligibility to take part was based on experiencing psychosis within the last 3 years; aged 18 or over; having no identified organic pathology (e.g. traumatic brain injury); ability to consent (based upon the opinion of the responsible

Matching

Educational attainment (U=445.00, z=−0.07, p=0.94), socio-economic status (U=431.00, z=−0.30, p=0.772), gender (χ2 (1, N=60)=0.12, p=0.734), marital status (χ2 (1, N=60)=1.35, p=0.337), ethnic origin (white British vs. ‘other’; Fisher׳s exact p=0.417) and family history of psychosis (χ2 (1, N=60)=0.22, p=0.638) did not differ significantly between the two groups. However, clinical participants were significantly older than non-clinical participants (U=268.50, z=−2.69, p=0.007) and more of the

Summary of results

As predicted, the clinical group reported significantly higher scores on the CTQ across all subscales. These results are consistent with previous research identifying high levels of childhood trauma within psychosis groups (Bendall et al., 2008, Varese et al., 2012b). Although in the current study the higher scores on the CTQ sexual abuse subscale were no longer significant when using more conservative comparisons, this may be due to lower statistical power and the increased likelihood of a

Funding

This work was supported by the UK National Health Service as partial fulfilment of GJE’s Doctorate in Clinical Psychology.

Contributors

Dr Evans conducted the data collection and analyses and drafted the paper. Dr Reid provided supervisory input and developed the protocol with Drs Evans, Preston and Sellwood. Drs Preston and Palmier-Claus also guided the development of the statistical approach taken. Dr Sellwood developed the idea for the project with the assistance of Drs Preston and Evans as well as having overall supervision of the project. He also assisted the drafting of the paper. All authors contributed to the design.

Conflict of interest

None.

Acknowledgements

The authors are thankful to the staff and service users of the Early Intervention Teams and Adult Education Colleges for their participation in this study. Additionally, helpful advice regarding mediational analysis was provided by Dr Filippo Varese.

References (43)

  • D.P. Bernstein et al.

    Childhood Trauma Questionnaire: A retrospective self-report manual

    (1998)
  • G. Berrios et al.

    The self and psychiatry: a conceptual history

  • J. Campbell et al.

    The structure of the self-concept and its relation to psychological adjustment

    Journal of Personality

    (2003)
  • J. Campbell et al.

    Self-concept clarity: measurement, personality correlates, and cultural boundaries

    Journal of Personality and Social Psychology

    (1996)
  • E.B. Carlson et al.

    An update on the dissociative experiences scale

    Dissociation: Progress in the Dissociative Disorders

    (1993)
  • D. Cicero et al.

    The role of aberrant salience and self-concept clarity in psychotic-like experiences

    Personality Disorders: Theory, Research and Treatment

    (2013)
  • J. Cohen

    Statistical Power Analysis for the Behavioral Sciences

    (1988)
  • M.S. Fritz et al.

    Required sample size to detect the mediated effect

    Psychological Science

    (2007)
  • B. Gershuny et al.

    Relation between trauma and psychopathology: mediating roles of dissociation and fears about death and control

    Journal of Trauma & Dissociation

    (2004)
  • P. Hammersley et al.

    Childhood trauma and psychosis: The genie is out of the bottle

    Journal of Psychological Trauma

    (2008)
  • A. Hayes

    Beyond Baron & Kenny: statistical mediation in the new millennium

    Communication Monographs

    (2009)
  • Cited by (51)

    • Childhood trauma and adolescent psychotic experiences in a community-based cohort: The potential role of positive attributes as a protective factor

      2019, Schizophrenia Research
      Citation Excerpt :

      Mediation analyses investigate the processes by which an independent variable (e.g. childhood trauma) affects an outcome (e.g. PE) (Hayes, 2013). Previous studies found that psychological factors such as low self-esteem, poor emotional regulation, and non-psychotic psychopathology mediate this relationship (Evans et al., 2015; Fisher et al., 2013b; Hardy et al., 2016; Lincoln et al., 2017; Murphy et al., 2015; Pearce et al., 2017; Yamasaki et al., 2016). However, studies that have specifically investigated protective factors as possible explanatory pathways connecting childhood trauma to PE are scarce (Crush et al., 2018).

    View all citing articles on Scopus
    View full text