Elsevier

Psychiatry Research

Volume 188, Issue 1, 30 June 2011, Pages 102-108
Psychiatry Research

Non-suicidal self-injury in eating disordered patients: A test of a conceptual model

https://doi.org/10.1016/j.psychres.2010.12.023Get rights and content

Abstract

A theoretical model explaining the high co-occurrence of non-suicidal self-injury (NSSI) in eating disordered populations as resulting from childhood traumatic experiences, low self-esteem, psychopathology, dissociation, and body dissatisfaction was previously proposed but not empirically tested. The current study empirically evaluated the fit of this proposed model within a sample of 422 young adult females (mean age = 21.60; S.D. = 6.27) consecutively admitted to an inpatient treatment unit for eating disorders. Participants completed a packet of questionnaires within a week of admission. Structural equation modeling procedures showed the model provided a good fit to the data, accounting for 15% of the variance in NSSI. Childhood trauma appears to have an indirect relationship to NSSI that is likely to be expressed via relationships to low self-esteem, psychopathology, body dissatisfaction, and dissociation. It appears that dissociation and body dissatisfaction may be particularly salient factors to consider in both understanding and treating NSSI within an eating disordered population.

Introduction

Non-suicidal self-injury (NSSI) refers to the deliberate and direct destruction of one's body tissue without suicidal intent (Claes and Vandereycken, 2007a), and is found among eating disorder (ED) patients with prevalence rates ranging between 25.4% and 55.2% (Svirko and Hawton, 2007). Given that both NSSI and eating disorders are body-focused disorders, researchers have begun to investigate potential shared factors in their co-occurrence. Svirko and Hawton (2007) proposed a conceptual model of risk, identifying the factors of childhood trauma, low self-esteem, affective disorders, dissociation, and body dissatisfaction as being key etiological variables for both ED and NSSI. No studies to date have tested a model integrating these factors to explain the occurrence of NSSI within ED populations; which is the primary aim of this study.

Theorists have proposed that experiencing childhood trauma, often in the form of abuse, plays a central role in the etiology of both NSSI and ED (Brodsky and Stanley, 2008, Thompson and Wonderlich, 2004, Yates, 2009). In a recent meta-analysis, Klonsky and Moyer (2008) found that childhood sexual abuse had a small association with NSSI (ф = 0.23), and a collection of studies have documented that past physical abuse may have a particularly strong association to NSSI (Favaro et al., 2007, Muehlenkamp et al., 2010, Tyler et al., 2003, Wachter et al., 2009). Emotional abuse has also been shown to be a robust predictor of NSSI (Glassman et al., 2007, Heath et al., 2009, Low et al., 2000). Similarly, Claes et al., 2003, Claes and Vandereycken, 2007b found that ED patients with NSSI reported higher levels of childhood emotional, physical and sexual abuse than ED patients without NSSI. Other authors (e.g., Brown et al., 1999, Favaro and Santonastaso, 1998, Favaro and Santonastaso, 2000, Paul et al., 2002) have also confirmed higher levels of childhood physical and sexual abuse in self-injurious ED patients compared to their non self-injurious peers. An important question remains as to whether childhood trauma has a direct or indirect effect on NSSI, and which other processes may also be involved in the observed associations.

Recent studies suggest that self-esteem, particularly in the form of self-criticism, may also be an important factor in understanding NSSI. Glassman et al. (2007) reported that self-criticism mediated the relationship between emotional abuse and NSSI, and this finding was conceptually replicated by Wedig and Nock (2007) who demonstrated that self-criticism was particularly salient in predicting NSSI. Others (Fliege et al., 2009) found that self-injurers showed lower self-efficacy and had higher self-blaming coping styles than non self-injurers (Hoff and Muehlenkamp, 2009). Claes et al. (2010) also found that participants who self-injured reported significantly lower self-esteem and exhibited poorer self-concepts relative to non-NSSI peers. Similar findings have been established within an eating disordered population as one study (Claes et al., 2003) found significantly more self-criticism and guilt in self-injurious ED patients compared to non self-injurious ED patients. Lastly, Low et al. (2000) conducted a path analysis that found an indirect path from childhood abuse through low self-esteem to NSSI in a sample of inpatient women. Together, these findings suggest that low self-esteem may be an important mediator of the abuse–NSSI relationship but additional studies are needed.

It has been noted that the presence of childhood traumatic experiences and/or low self-esteem increase the probability of affective symptoms such as anxiety and depression. These affective symptoms may place individuals at risk for NSSI (Klonsky et al., 2003) and several authors have found significant links between a history of depression and anxiety in NSSI (Lofthouse et al., 2009, Nock and Kessler, 2006, Nock et al., 2006). Furthermore, Wildman et al. (2004) showed that depressive symptoms preceded the onset of both ED and NSSI in a sample of adolescents. Finally, several studies in ED patients (Claes et al., 2001, Claes et al., 2003, Favaro and Santonastaso, 1998, Favaro and Santonastaso, 2000, Paul et al., 2002) showed elevated levels of anxiety and depression in ED patients with NSSI compared to ED patients without NSSI. Therefore, the literature is largely consistent in identifying anxiety and depression as being important aspects of psychopathology associated with increased risk for either ED or NSSI.

Also commonly studied as a correlate of ED and NSSI, are symptoms of dissociation. Several studies (e.g., Low et al., 2000, Nijman et al., 1999, Vanderlinden et al., 2007) have shown an indirect path from childhood traumatic experiences to NSSI via dissociation. A positive association between dissociation and NSSI within ED patients has also been documented (Claes et al., 2001, Claes et al., 2003, Paul et al., 2002). The potential impact of dissociation on NSSI appears to be particularly robust, as past studies (Brodsky et al., 1995, Mulder et al., 1998) have also demonstrated positive associations between depressive symptoms, dissociative experiences and NSSI even after controlling for traumatic experiences. While some empirical evidence suggests that only a minority report that the NSSI is in response to dissociation (Kemperman et al., 1997, Nock and Prinstein, 2004), the research appears to support the need to consider dissociation within etiological models.

In addition to the bodily disturbance that accompanies dissociation, some researchers acknowledge that a negative attitude toward the body may be a critical component in NSSI (e.g., Muehlenkamp, 2005, Muehlenkamp et al., 2005, Walsh, 2006) because dissatisfaction may promote views of the body as a hated object, making it easier to harm (Orbach, 1996). Supporting this notion are studies that show comparatively high rates of body dissatisfaction in self-injurers versus non self-injurers (e.g., Favaro et al., 2007, Ross et al., 2009). Other studies have documented that body dissatisfaction is significantly associated with NSSI and other self-harm behaviors after controlling for associated variables such as depression (Brausch and Gutierrez, 2009, Brausch and Muehlenkamp, 2007). The increased rates of body dissatisfaction among those who engage in NSSI are even notable within ED patients, who conceptually already suffer from high levels of body dissatisfaction. Studies have shown that ED patients with a history of NSSI endorse higher levels of body dissatisfaction or body disgust, than ED patients without NSSI (e.g., Claes et al., 2003, Solano et al., 2005). Consequently, body dissatisfaction should be represented within etiological models of NSSI, but it has been largely overlooked.

Based on the extant literature, it would seem that that each of the aforementioned risk factors may play an important role in understanding the etiology and risk for NSSI within ED samples. One shortcoming of the current literature, however, is that the majority of these studies have investigated each factor separately or, in some cases, examined relationships between only a couple of these factors. Few studies have sought to integrate these factors into one conceptual model and to empirically test the model. The current study proposes a model of NSSI risk (see Fig. 1) for an eating-disorder sample that integrates research findings regarding the associations between NSSI and childhood trauma, self-esteem, dissociation, affective pathology, and body dissatisfaction.

Consistent with the literature reviewed, childhood trauma is seen as one potential developmental precursor to NSSI (Yates, 2009), but also to the onset of low self-esteem, affective pathology, dissociation, and body dissatisfaction all of which have demonstrated strong relationships to NSSI. It seems plausible that each of these factors could act as possible mediators in the abuse–NSSI link. Since some research has found self-esteem to be a mediator between trauma and NSSI (Low et al., 2000), and low self-esteem is commonly identified as a precipitant for internalizing disorders such as depression, self-esteem was placed as an exogenous factor resulting from childhood trauma within the model. Building from the research suggesting trauma and self-esteem precede internalizing symptoms, affective pathology was placed next in the model. Theorists (Orbach, 1996, Walsh, 2006) have suggested that disregard for the body is a necessary but not sufficient factor for NSSI to occur even in the context of other psychopathology, and studies so far have documented high rates of body dissatisfaction among those reporting NSSI (e.g., Favaro et al., 2007, Ross et al., 2009), as well as dissociative symptoms (e.g., Vanderlinden et al., 2007). From this theoretical perspective, dissociation and body dissatisfaction should partially mediate the relationship between other pathologies and NSSI behavior, so they were placed as the final risk variables for NSSI within the proposed model. It was hypothesized that this model would provide a significant fit to the data relative to a null model and account for a small to moderate portion of variance in NSSI. The hypothesized pathways are presented in Fig. 1.

Section snippets

Participants and procedure

Data were collected from 422 female patients (mean age = 21.60, S.D. = 6.27) admitted to a specialized inpatient eating disorder unit in Belgium. All patients admitted to the inpatient unit were invited to participate in the study shortly after admission (approximately 5–6 days). Patients willing to participate (100%) provided written informed consent and completed the questionnaires on a PC within a private room in the hospital. Eating disorder diagnoses, as specified in the DSM-IV (APA, 1994),

Results

Of the total sample, 34.60% (n = 146) reported engaging in NSSI, with a mean age of onset of 16.48 years (S.D. = 4.93). The most common methods of NSSI were cutting (n = 114; 78.1%), skin abrating/severe scratching (n = 77; 52.7%), bruising (n = 57; 39.0%), and burning (n = 29; 19.9%). There were no differences between the eating disorder subtypes and the duration of NSSI (χ2 (20) = 29.38, P > 0.05), nor the frequency of NSSI per week (χ2 (36) = 40.15, P > 0.05). However, binge/purging subtypes (AN-P and BN-P)

Discussion

Research has suggested that NSSI and disordered eating behaviors share a variety of similar risk factors, which may help to explain why these two behaviors co-occur so frequently. In this study, an integrated model was tested to evaluate whether it could explain a significant amount of variance in NSSI behavior within an eating-disordered sample. The results partially supported the model. In support of the preliminary hypotheses, significant pathways were found from childhood abuse and low

References (66)

  • J.J. Muehlenkamp et al.

    Emotional states preceding and following acts of non-suicidal self-injury in bulimia nervosa patients

    Behavior Research and Therapy

    (2009)
  • P.E. Mullen et al.

    The long-term impact of the physical, emotional, and sexual abuse of children: a community study

    Child Abuse & Neglect

    (1996)
  • M.K. Nock et al.

    Non-suicidal self-injury among adolescents: diagnostic correlates and relation to suicide attempts

    Psychiatry Research

    (2006)
  • M.M. Wedig et al.

    Parental expressed emotion and adolescent self-injury

    Journal of the American Academy of Child and Adolescent Psychiatry

    (2007)
  • American Psychiatric Association

    Diagnostic and Statistical Manual of Mental Disorders

    (1994)
  • J.C. Anderson et al.

    Structural equation modeling in practice: a review and recommended two-step approach

    Psychological Bulletin

    (1988)
  • W.A. Arrindell et al.

    SCL-90: Handleiding bij een multidimensionele psychopathologie-indicator

    (1986)
  • A.M. Brausch et al.

    The role of body image and disordered eating as risk factors for depression and suicidal ideation in adolescents

    Suicide and Life-Threatening Behavior

    (2009)
  • B. Brodsky et al.

    Relationship of dissociation to self-mutilation and childhood abuse in borderline personality disorder

    The American Journal of Psychiatry

    (1995)
  • J. Brown et al.

    A longitudinal analysis of risk factors for child maltreatment: findings of a 17-year prospective study of officially recorded and self-reported child abuse and neglect

    Child Abuse & Neglect

    (1999)
  • L. Claes et al.

    Is there a link between traumatic experiences and self-injurious behaviours in eating-disordered patients?

    Eating Disorders: the Journal of Treatment and Prevention

    (2007)
  • L. Claes et al.

    Eating-disordered patients with and without self-injurious behaviours: a comparison of psychopathological features

    European Eating Disorders Review

    (2003)
  • P.T. Costa et al.

    Revised NEO Personality Inventory (NEO-PI-R) and the Five Factor Inventory (NEO-FFI): Professional Manual

    (1992)
  • A. Favaro et al.

    Impulsive and compulsive self-injurious behavior in bulimia nervosa: prevalence and psychological correlates

    The Journal of Nervous and Mental Disease

    (1998)
  • A. Favaro et al.

    Self-injurious behavior in anorexia nervosa

    The Journal of Nervous and Mental Disease

    (2000)
  • A. Favaro et al.

    Self-injurious behavior in a community sample of young women: relationship with childhood abuse and other types of self-damaging behaviors

    Journal of Clinical Psychiatry

    (2007)
  • D.M. Garner

    The Eating Disorder Inventory 2

    (1991)
  • N.L. Heath et al.

    An examination of nonsuicidal self-injury among college students

    Journal of Mental Health Counseling

    (2009)
  • H.A. Hoekstra et al.

    NEO-PI-R and NEO FFI Big Five Persoonlijkheidsvragenlijsten

    (1996)
  • E.R. Hoff et al.

    Non-suicidal self-injury in college students: the role of perfectionism and rumination

    Suicide and Life-Threatening Behavior

    (2009)
  • W.R. Holcomb et al.

    Factor structure of the symptom checklist-90 with acute psychiatric inpatients

    Journal of Consulting and Clinical Psychology

    (1983)
  • L. Hu et al.

    Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives

    Structural Equation Modeling

    (1999)
  • K. Jöreskog et al.

    LISRELL 8.71 (Computer Program)

    (2004)
  • Cited by (94)

    View all citing articles on Scopus
    View full text