Journal of Consulting and Clinical Psychology 2015, Vol. 83, No. 3, 643– 648
© 2014 American Psychological Association 0022-006X/15/$12.00 http://dx.doi.org/10.1037/ccp0000014
BRIEF REPORT
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Self-Harm and Suicidal Behavior in Borderline Personality Disorder With and Without Bulimia Nervosa Deborah L. Reas
Geir Pedersen
Oslo University Hospital, Oslo, Norway
Oslo University Hospital, Oslo, Norway and University of Oslo, Oslo
Sigmund Karterud
Øyvind Rø
University of Oslo and Oslo University Hospital, Oslo, Norway
Oslo University Hospital, Oslo, Norway and University of Oslo
Objective: Few studies have investigated whether a diagnosis of Bulimia nervosa (BN) confers additional risk of life-threatening behaviors such as self-harm and suicidal behavior in borderline personality disorder (BPD). Method: Participants were 483 treatment-seeking women diagnosed with BPD according to the Structured Clinical Interview for DSM–IV Axis II Personality Disorders (SCID-II; First, Gibbon, Spitzer, Williams, & Benjamin, 1997; Diagnostic and Statistical Manual of Mental Disorders, 4th ed.; APA, 1994) and admitted to the Norwegian Network of Psychotherapeutic Day Hospitals between 1996 and 2009. Of these, 57 (11.8%) women met DSM–IV diagnostic criteria for BN according to the Mini-International Neuropsychiatric Interview (M.I.N.I.; Sheehan et al., 1998) and they were compared with women with BPD and other Axis I disorders. Results: We found that comorbid BN is uniquely and significantly associated with increased risk of suicidal behavior among women being treated for BPD. Findings underscore the importance of routinely screening for BN among women seeking treatment for BPD, as co-occurring bulimia appears to be a significant marker for immediate life-threatening behaviors in this already high-risk population, which is a significant public health issue. A significantly greater proportion of women with BPD-BN reported suicidal ideation at intake (past 7 days), engaged in self-harm behavior during treatment, and attempted suicide during treatment. All bivariate associations remained significant in the logistic regression models after controlling for mood, anxiety, and substance-related disorders. Conclusion: The presence of a concurrent diagnosis of BN among women with BPD is significantly and uniquely associated with recent suicidal ideation, and self-harm behavior and suicide attempts during treatment after controlling for major classes of mental disorders. Co-occurring BN appears to represent a significant marker for immediate life-threatening behaviors in women seeking treatment for BPD. Extra vigilance and careful monitoring of suicidal behavior during treatment is important for these individuals, and routine screening for BN is warranted.
What is the public health significance of this article? This study found that co-occurring bulimia nervosa is uniquely and significantly associated with increased risk of suicidal behavior among women being treated for borderline personality disorder. Findings underscore the importance of routinely screening for bulimia nervosa among women seeking treatment for borderline personality disorder, as co-occurring bulimia appears to be a significant marker for immediate life-threatening behaviors in this already high-risk population.
Keywords: suicide, self-harm, Bulimia nervosa, borderline personality disorder
This article was published Online First December 15, 2014. Deborah L. Reas, Regional Department of Eating Disorders, Division of Mental Health and Addiction, Oslo University Hospital, Oslo, Norway; Geir Pedersen, Department of Personality Psychiatry, Division of Mental Health and Addiction, Oslo University Hospital, Oslo, Norway, and Institute for Clinical Medicine, Faculty of Medicine, University of Oslo; Sigmund Karterud, Institute for Clinical Medicine, Faculty of Medicine, University of Oslo and Department of Research and Development, Division of Mental Health and Addiction, Oslo University Hospital, Oslo,
Norway; Øyvind Rø, Regional Department of Eating Disorders, Division of Mental Health and Addiction, Oslo University Hospital, Oslo, Norway and Institute for Clinical Medicine, Faculty of Medicine, University of Oslo, Oslo, Norway. Correspondence concerning this article should be addressed to Deborah L. Reas, Regional Department of Eating Disorders (RASP), Division of Mental Health and Addiction, Oslo University Hospital–Ullevål Hospital, P.O. Box 4956 Nydalen, N-0424 Oslo, Norway. E-mail: deborah.lynn [email protected] 643
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
644
REAS, PEDERSEN, KARTERUD, AND RØ
Borderline personality disorder (BPD) is characterized by a pervasive pattern of instability in interpersonal relationships, selfimage, and affect (APA, 2013) and is marked by impulsivity and recurrent suicidal and self-mutilating behavior. Suicidal behavior and self-harm also occur among patients with Bulimia nervosa (BN; Corcos et al., 2002; Crow et al., 2009; Franko & Keel, 2006; Preti, Rocchi, Sisti, Camboni, & Miotto, 2011) and a robust and differential pattern of comorbidity has been found between BPD and BN in clinical samples of mixed personality disorder (PD; Reas, Rø, Karterud, Hummelen, & Pedersen, 2013). Significant associations have also been found between purging behavior, personality traits, and self-injury or suicidality in women with eating disorders (ED) (Paul, Schroeter, Dahme, & Nutzinger, 2002) and adolescent psychiatric inpatients (Zaitsoff & Grilo, 2010), consistent with clinical observations linking BN with impulsivity and dysregulation. The extent to which BN in the presence of BPD might confer additional risk of suicidality and self-injury is unclear. Several studies have investigated the predictive validity of comorbidity associated with suicide risk among individuals with PD, yet these have focused largely on risk attributable to major depression, substance abuse, or anxiety disorders (Soloff, Lynch, Kelly, Malone, & Mann, 2000; Wedig et al., 2012; Yen et al., 2003), with comparatively scarce attention on the independent predictive utility of BN. To our knowledge, only two controlled investigations have specifically addressed the incremental prognostic validity of BN in samples of BPD sufferers (Chen, Brown, Harned, & Linehan, 2009; Dulit, Fyer, Leon, Brodsky, & Frances, 1994). An investigation by Dulit et al. (1994) found that BPD inpatients (N ⫽ 124) with comorbid BN were four times more likely to engage in repeated self-injury (ⱖ5 lifetime acts) than BPD inpatients without BN. An investigation of 135 women with BPD observed a significant association between BN and recurrent suicide attempts (ⱖ2 acts), but not other self-injury, after controlling for age and other non-ED Axis I disorders (Chen et al., 2009). A later study by Chen et al. (2011) found equivalent rates of lifetime suicide attempts and self-injury among 166 women and 166 men with and without ED in a diagnostically heterogeneous sample of PD. Evidence on a cross-sectional level supporting the incremental validity of BN in predicting suicidality in this high-risk population would underscore the importance of screening for BN in treatment settings for BPD. Such an approach is also consistent with recommendations that future researchers consider the effects of comorbidity when elucidating predictive effects of mental disorders on suicide attempts (Nock, Hwang, Sampson, & Kessler, 2010). To be of greatest benefit, further investigations should address previously identified methodological limitations, including (a) the merging of highly select samples from trials recruiting or screening specifically for suicidal BPD or substance-dependent BPD, which may limit ecological or clinical representativeness, and (b) small and/or diagnostically heterogeneous samples which may render effect sizes diminished due to low power. The present study aimed to investigate whether comorbidity-independent associations exist between BN and self-harm, suicidal ideation, and suicide attempts among a naturalistic, treatment-seeking sample of women with BPD.
Method Participants Participants included 483 women diagnosed with BPD, aged 18 – 65 years, with an initial admission between 1996 and 2009 to the Norwegian Network of Personality-Focused Treatment Programs. Established in 1992, this is a clinical research network providing mostly long-term, group-based (or concurrent individual– group) treatment (see also Karterud et al., 2003; Reas et al., 2013). All treatment units used uniform and standardized assessment procedures (Pedersen, Karterud, Hummelen, & Wilberg, 2013) following the longitudinal, expert, all-data (LEAD) standard, which is a comprehensive, integrative diagnostic approach using multiple sources of information (e.g., interview data, informants, behavioral observations, and medical records). Data collection is overseen by a central coordinating site responsible for quality assurance, standardization of routines, and screening data for irregularities and missing data. Raters all held professional degrees and skill acquisition, and maintenance included training courses and supervision. Ongoing monitoring of protocol adherence included checklists and periodic site visits (up to 3– 4 times annually) and in addition, site coordinators from all 16 units met every 6 months to discuss and calibrate diagnostic and clinic procedures.
Materials and Procedure Patients were interviewed with the Structured Clinical Interview for DSM–IV Axis II Personality Disorders (SCID-II; First, Gibbon, Spitzer, Williams, & Benjamin, 1997), a well-established diagnostic tool with demonstrated reliability for the assessment of PD (Lobbestael, Leurgans, & Arntz, 2011). Interrater reliability for the Norwegian version of the SCID-II has been established ( ⫽ .66 for BPD; Kvarstein et al., 2014). The Mini-International Neuropsychiatric Interview Version 4.4 (M.I.N.I.; Sheehan et al., 1998) was used to establish Axis I diagnoses, which has demonstrated reliability and validity for the assessment of Axis I disorders, including BN ( ⫽ .78; Sheehan et al., 1998). The Norwegian version of the M.I.N.I is validated and is considered a timeefficient and feasible alternative to the SCID-P (SCID-I/P; First et al., 2002) and CIDI (CIDI; Kessler & Ustün, 2004) (Mordal, Gundersen, & Bramness, 2010).
Assessment of Suicidal Behavior The assessment of suicidality included clinician- and selfreported data capturing different time epochs. First, self-reported data were systematically collected at intake using a routinely administered sociodemographic questionnaire. Suicidal ideation was assessed dichotomously to capture the past 7 days and past 12 months, that is, “Have you had thoughts about taking your own life?” Self-harm behavior was assessed dichotomously to capture the past 12 months and lifetime, that is, “Have you [ever] physically hurt yourself, for example, cutting, scratching, burning, headbanging, and so forth?” Suicide attempts were assessed dichotomously to capture the past 12 months and lifetime, that is, “Have you ever tried to kill yourself?” The number of lifetime suicide attempts was also rated. Clinician-rated data were systematically collected at discharge using a routine discharge form to assess (a)
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
SELF-HARM AND SUICIDAL BEHAVIOR IN BPD-BN
the occurrence of self-harm behavior during treatment, that is, physically harmful behaviors such as cutting, scratching, burning, banging against hard objects, and so forth; (b) suicidal ideation during treatment, that is, “Did the patient express thoughts about taking one’s life, excluding very fleeting or dramatic expressions regarding the wish to die?” and (c) suicide attempts occurring during treatment, that is, “lethal acts with the intent to die that would have been successful without acute intervention from others.” It should be noted that assessment of self-harm behavior did not specify expectations (e.g., to gain relief from negative feelings, relieve suffering, resolve interpersonal difficulties) as specified in DSM-5 (APA, 2013) for nonsuicidal self-injury; thus, we use the more general term of self-harm. Bulimic behaviors, sometimes conceptualized under the rubric of self-harm, were not covered by assessment. All data were registered in a central, anonymous database administered by Oslo University Hospital. All patients provided written consent and the study was approved by the State Data Inspectorate and the Regional Ethics Committee.
Data Analyses Analyses were conducted using predictive analysis software (PASW) Version 18.0. Patients were grouped according to the presence of BN (BPD-BN) or non-ED Axis I disorder (BPDother), in line with grouping methods by Chen et al. (2011). Cases of anorexia nervosa (AN) (N ⫽ 7) and eating disorders not otherwise specified (EDNOS) (N ⫽ 81) were excluded from BPD-other due to potential confounding of subthreshold BN in the EDNOS group, or lifetime history of BN, because diagnostic fluctuation or crossover is common in DSM-IV ED (Peterson et al., 2011). Chi-square analyses tested differences for categorical variables. Consistent with methods from earlier studies (Bodell, Joiner, & Keel, 2013), logistic regression analyses (ORs and 95% CIs) controlled for mood, anxiety, and substance-use disorders were
645
conducted when significant bivariate associations were detected. Following guidelines detailed in previous research (Chen, Cohen, & Chen, 2010), ORs of 1.68, 3.47, and 6.71 were considered equivalent to small, medium, and large effect sizes (Cohen’s d ⫽ 0.2, 0.5, and 0.8, respectively). Analyses were two-tailed (p ⬍ .05).
Results Sample Characteristics A total of 57 patients (11.8%) received a comorbid diagnosis of BN and BPD and were grouped as BPD-BN. All other patients (N ⫽ 426) were diagnosed with at least one non-ED Axis I diagnosis (BPD-other). Approximately 68% of both groups had mood disorders; 63.2% versus 52.6% (p ⫽ .123) were diagnosed with anxiety disorder; and 28.1% versus 15.1% (p ⫽ .014) had substance-use disorder. Table 1 shows no significant baseline differences for age, mean global assessment of functioning (GAF) at intake, length of treatment, mean frequency of non-ED Axis I or Axis II disorders, and number of SCID-II BPD criteria fulfilled. No differences were detected for marital status, 2(4, 459) ⫽ 2.97, p ⫽ .562, or work situation, 2(5, 421) ⫽ 4.91, p ⫽ .427.
Clinician and Self-Reported Self-Harm, Suicidal Ideation, and Suicide Attempts As shown in Table 1, a significantly greater proportion of women in the BPD-BN group demonstrated self-harm behavior during treatment (p ⬍ .001). Approximately 50% of both groups engaged in self-harm behavior over the past 12 months. Lifetime self-harm did not show a statistically significant difference between BPD-BN and BPD-other (70.9% vs. 58.2%; p ⫽ .079). A significantly higher proportion of BPD-BN reported suicidal ide-
Table 1 Sample Characteristics for BPD-BN Versus BPD-Other (N ⫽ 483) Variable Age in years (M, SD) Length of admission (days), M (SD) Global assessment functioning (GAF), M (SD) # SCID-II criteria met for BPD, M (SD) # Axis II disorders, M (SD) # Non-ED Axis I disorders, M (SD) Mood disorder, yes (%) Anxiety disorder, yes (%) Substance-use disorder, yes (%) Self-harm during treatment, yes (%)a Self-harm, past 12 months, yes (%) Self-harm, lifetime, yes (%) Suicidal ideation during treatment, yes (%)a Suicidal ideation, past 7 days, yes (%) Suicidal ideation, past 12 months, yes (%) Suicide attempts during treatment, yes (%)a Suicide attempts, past 12 months, yes (%) Suicide attempts, lifetime, yes (%) Suicide attempts, recurrent, yes (%)b
BPD-BN (n ⫽ 57) BPD-other (n ⫽ 426) 29.9 (7.42) 136.7 (58.7) 45.1 (4.30) 5.98 (1.09) 1.51 (0.66) 1.95 (1.28) 68.4% 63.2% 28.1% 39.3% 46.7% 70.9% 52.6% 55.6% 88.4% 10.9% 31.8% 64.8% 45.1%
30.2 (7.32) 136.3 (58.4) 45.7 (5.12) 6.06 (1.08) 1.58 (0.76) 1.95 (1.27) 67.9% 52.6.% 15.1% 15.7% 46.4% 58.2% 39.5% 36.0% 77.4% 4.1% 22.8% 58.0% 35.4%
t/2
p value
0.32 .054 1.01 0.47 0.67 ⫺0.09 0.00 2.38 6.10 18.3 .002 3.22 3.59 6.31 2.68 4.79 1.70 0.91 1.84
.750 .957 .317 .641 .503 .993 .940 .123 .014 .001 .969 .079 .058 .012 .101 .029 .192 .340 .175
Note. BPD ⫽ borderline personality disorder; BN ⫽ Bulimia nervosa; no completed suicides occurred during treatment. a Indicates clinician-reported data. b Recurrent is defined as ⱖ 2 acts.
REAS, PEDERSEN, KARTERUD, AND RØ
646
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
ation during the past 7 days prior to intake (p ⫽ .012), but differences in suicidal ideation (past 12 months) were not significant (88.4% vs. 77.4%, p ⫽ .101). A trend was detected for greater suicidal ideation during treatment in BPD-BN (p ⫽ .058). A significantly higher proportion of women with BPD-BN attempted suicide during treatment (p ⫽ .029). No significant differences were found for suicide attempts during the past 12 months (31.8% vs. 22.8%, p ⫽ .192), lifetime, or recurrent suicide attempts (two or more acts). For suicide attempters, mean (SD) number of lifetime attempts was 3.0 (2.24) for BPD-BN versus 2.80 (2.68) for BPD-other, respectively, t(255) ⫽ .401, p ⫽ .688.
Logistic Regression Table 2 shows significant associations between BN and selfharm during treatment OR ⫽ 3.23, 95% CI [1.76 –5.92], suicidal ideation past 7 days, OR ⫽ 2.37, 95% CI [1.23–4.55]; and suicide attempts during treatment OR ⫽ 2.83; 95% CI [1.05–7.64]. An OR of 1.73, 95% CI [.987–3.04], p ⫽ .055 was detected for suicidal ideation during treatment, indicating a small effect.
Discussion These findings indicate that the presence of a comorbid diagnosis of BN in the context of BPD is significantly and uniquely associated with increased risk of recent suicidal ideation at intake and self-harm and suicide attempts during treatment after controlling for mood, anxiety, and substance-related disorders. As such, a concurrent diagnosis of BN among women seeking treatment for BPD appears to represent a strong and significant marker for immediate life-threatening behaviors. Death by suicide occurs in 8 –10% of individuals with BPD, which is among the highest rates of all mental disorders (Pompili, Girardi, Ruberto, & Tatarelli, 2005). Our study has indicated that bulimic episodes signal even greater risk of suicide attempts within this high-risk population. Findings echo those by Bodell et al. (2013), who found comorbidity-independent associations between BN and lifetime suicidality among university women, prompting calls for a standard risk assessment of suicide among women with BN. The sample size and setting were considered study strengths, that is, over 400 consecutively admitted female patients seeking day treatment for BPD. No significant differences were observed in length of admission (i.e., approximately 4.5 months) or length of referral process; thus, these variables were not
considered potential confounds. Data collection was part of routine clinic procedure, and although risk of bias cannot be eliminated, only minimal bias owing to clinician expectations regarding the present study aims were expected to influence results. Several limitations of our study are important to consider. Despite face validity, the reliance on direct, single-item assessments limited the scope of the measurement. Information regarding lethality, intent, or specific methods of self-harm was unavailable, and behaviors such as aborted, interrupted, and low-lethality attempts might have not been captured. Prior research, however, has used single-item, self-report assessments of suicidal ideation and attempt with demonstrated validity (Bodell et al., 2013; Cougle, Keough, Riccardi, & SachsEricsson, 2009). Retrospective self-report data on the occurrence of self-harm and suicidal behavior might be subject to recall bias. This study design was cross-sectional, thereby precluding the ability to infer stability of findings and the longitudinal risk or other clinical outcomes (e.g., suicide completion). Self-harm behavior, suicidal ideation, and past suicide attempts have been documented as important risk factors for future suicide (Klonsky, May, & Glenn, 2013), also in samples with BPD (Wedig et al., 2012). The overall rate of lifetime attempts in our sample was 59%, which is higher than typically observed in BN (25–35%; Franko & Keel, 2006), although within the 40 – 85% range observed for BPD (Oumaya et al., 2008). Our sample included treatment-seeking women admitted for intensive and specialized day-hospital treatment for PD; thus, results may not generalize to individuals under 18 or over 65 years of age, or to community or nonspecialist treatment settings which serve less severe populations. Because the pattern of findings indicated significantly elevated risk of suicidality at intake and during treatment, but not prior to treatment, replication is necessary to rule out potential state effects underlying results. Further investigation is needed to explore whether findings relate to appropriateness of therapeutic setting or treatment approach, for example, or whether suicidality might constitute a particularly salient motivator for treatment-seeking among women with BPD-BN. Several alternative classification schemes for subtyping BN according to patterned heterogeneity in comorbidity (i.e., multi-impulsive BN, borderline–nonborderline BN, undercontrolled– externalizing; see Wildes & Marcus, 2013) may have relevance for the contextualization and conceptualization of findings, or our re-
Table 2 Logistic Regression Models for the Association Between BN and Self-Harm, Suicidal Ideation, and Suicide Attempts Among Women With BPD Controlling for Mood, Anxiety, and Substance-Use Disorders
Illness Mood disorder Anxiety disorder Substance disorder BN
Self-harm (during tx) OR 95% CI 1.24 0.76 1.47 3.23ⴱⴱ
[0.74–2.09] [0.47–1.23] [0.81–2.67] [1.76–5.92]
Suicidal ideation (during tx) OR 95% CI 0.89 0.97 0.94 1.73 (p ⫽ .055)
[0.61–1.33] [0.66–1.42] [.567–1.54] [.987–3.04]
Suicidal ideation (past 7 days) OR 95% CI 1.49 1.37 0.90 2.37ⴱⴱ
[0.91–2.44] [0.86–2.19] [0.51–1.61] [1.23–4.55]
Suicide attempts (during tx) OR 95% CI 1.06 1.21 1.23 2.83ⴱ
Note. BPD ⫽ borderline personality disorder; BN ⫽ Bulimia nervosa; tx ⫽ treatment; OR ⫽ odds ratio; CI ⫽ 95% confidence interval. p ⱕ .05. ⴱⴱ p ⱕ .01.
ⴱ
[0.42–2.65] [0.49–2.96] [0.43–3.47] [1.05–7.64]
SELF-HARM AND SUICIDAL BEHAVIOR IN BPD-BN
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
sults may speak broadly to cross-cutting behavioral and neurobiologically informed constructs such as trait impulsivity or cognitive control (Insel, 2014). This was a controlled investigation providing evidence on a cross-sectional level supporting the incremental validity of BN in predicting suicidality beyond mood, anxiety, and substancerelated disorders in the high-risk BPD population. Findings warrant routine assessment of BN in treatment-seeking samples of women with BPD, and underscore the importance of high vigilance and fastidious monitoring of suicidal behaviors during treatment for these individuals.
References American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. American Psychiatric Association, DSM-5 Task Force. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: American Psychiatric Publishing. Bodell, L. P., Joiner, T. E., & Keel, P. K. (2013). Comorbidity-independent risk for suicidality increases with BN but not with anorexia nervosa. Journal of Psychiatric Research, 47, 617– 621. http://dx.doi.org/ 10.1016/j.jpsychires.2013.01.005 Chen, H., Cohen, P., & Chen, S. (2010). How big is a big odds ratio? Interpreting the magnitudes of odds ratios in epidemiological studies. Communications in Statistics: Simulation and Computation, 39, 860 – 864. http://dx.doi.org/10.1080/03610911003650383 Chen, E. Y., Brown, M. Z., Harned, M. S., & Linehan, M. M. (2009). A comparison of BPD with and without eating disorders. Psychiatry Research, 170, 86 –90. http://dx.doi.org/10.1016/j.psychres.2009.03.006 Chen, E. Y., McCloskey, M. S., Michelson, S., Gordon, K. H., & Coccaro, E. (2011). Characterizing eating disorders in a personality disorders sample. Psychiatry Research, 185, 427– 432. Corcos, M., Taïeb, O., Benoit-Lamy, S., Paterniti, S., Jeammet, P., & Flament, M. F. (2002). Suicide attempts in women with BN: Frequency and characteristics. Acta Psychiatrica Scandinavica, 106, 381–386. http://dx.doi.org/10.1034/j.1600-0447.2002.02318.x Cougle, J. R., Keough, M. E., Riccardi, C. J., & Sachs-Ericsson, N. (2009). Anxiety disorders and suicidality in the National Comorbidity Survey– Replication. Journal of Psychiatric Research, 43, 825– 829. http://dx.doi .org/10.1016/j.jpsychires.2008.12.004 Crow, S. J., Peterson, C. B., Swanson, S. A., Raymond, N. C., Specker, S., Eckert, E. D., & Mitchell, J. E. (2009). Increased mortality in BN and other eating disorders. The American Journal of Psychiatry, 166, 1342– 1346. http://dx.doi.org/10.1176/appi.ajp.2009.09020247 Dulit, R. A., Fyer, M. R., Leon, A. C., Brodsky, B. S., & Frances, A. J. (1994). Clinical correlates of self-mutilation in BPD. The American Journal of Psychiatry, 151, 1305–1311. First, M. B., Gibbon, M., Spitzer, R. L., Williams, J. B. W., & Benjamin, L. S. (1997). Structured Clinical Interview for DSM–IV Axis II Personality Disorders. Washington, DC: American Psychiatric Association. First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. W. (2002). Structured Clinical Interview for DSM-IV-TR Axis I Disorders. Research Version, Patient Edition. (SCID-I/P) New York: Biometrics Research, New York State Psychiatric Institute. Franko, D. L., & Keel, P. K. (2006). Suicidality in eating disorders: Occurrence, correlates, and clinical implications. Clinical Psychology Review, 26, 769 –782. http://dx.doi.org/10.1016/j.cpr.2006.04.001 Insel, T. R. (2014). The NIMH Research Domain Criteria (RDoC) Project: Precision medicine for psychiatry. The American Journal of Psychiatry, 171, 395–397. http://dx.doi.org/10.1176/appi.ajp.2014.14020138 Karterud, S., Pedersen, G., Bjordal, E., Brabrand, J., Friis, S., Haaseth, O., . . . Urnes, O. (2003). Day treatment of patients with personality disorders: Experiences from a Norwegian treatment research network.
647
Journal of Personality Disorders, 17, 243–262. http://dx.doi.org/ 10.1521/pedi.17.3.243.22151 Kessler, R. C., & Ustün, T. B. (2004). The World Mental Health (WMH) Survey Initiative version of the World Health Organization (WHO) Composite International Diagnostic Interview (CIDI). International Journal of Methods in Psychiatric Research, 13, 93–121. Klonsky, E. D., May, A. M., & Glenn, C. R. (2013). The relationship between nonsuicidal self-injury and attempted suicide: Converging evidence from four samples. Journal of Abnormal Psychology, 122, 231– 237. http://dx.doi.org/10.1037/a0030278 Kvarstein, E. H., Pedersen, G., Urnes, O., Hummelen, B., Wilberg, T., & Karterud, S. (2014). Changing from a traditional psychodynamic treatment programme to mentalization-based treatment for patients with BPD: Does it make a difference? Psychology and Psychotherapy: Theory, Research and Practice. Advance online publication. http://dx.doi .org/10.1111/papt.12036 Lobbestael, J., Leurgans, M., & Arntz, A. (2011). Inter-rater reliability of the Structured Clinical Interview for DSM–IV Axis I Disorders (SCID I) and Axis II Disorders (SCID II). Clinical Psychology & Psychotherapy, 18, 75–79. http://dx.doi.org/10.1002/cpp.693 Mordal, J., Gundersen, O., & Bramness, J. G. (2010). Norwegian version of the Mini-International Neuropsychiatric Interview: Feasibility, acceptability and test-retest reliability in an acute psychiatric ward. European Psychiatry, 25, 172–177. http://dx.doi.org/10.1016/j.eurpsy.2009 .02.004 Nock, M. K., Hwang, I., Sampson, N. A., & Kessler, R. C. (2010). Mental disorders, comorbidity and suicidal behavior: Results from the National Comorbidity Survey Replication. Molecular Psychiatry, 15, 868 – 876. http://dx.doi.org/10.1038/mp.2009.29 Oumaya, M., Friedman, S., Pham, A., Abou Abdallah, T., Guelfi, J.-D., & Rouillon, F. (2008). Personnalité borderline, automutilations et suicide: Revue de la littérature [BPD, self-mutilation and suicide: Literature review]. L’Encéphale, 34, 452– 458. http://dx.doi.org/10.1016/j.encep .2007.10.007 Paul, T., Schroeter, K., Dahme, B., & Nutzinger, D. O. (2002). Selfinjurious behavior in women with eating disorders. The American Journal of Psychiatry, 159, 408 – 411. http://dx.doi.org/10.1176/appi.ajp.159 .3.408 Pedersen, G., Karterud, S., Hummelen, B., & Wilberg, T. (2013). The impact of extended longitudinal observation on the assessment of personality disorders. Personality and Mental Health, 7, 277–287. http:// dx.doi.org/10.1002/pmh.1234 Peterson, C. B., Crow, S. J., Swanson, S. A., Crosby, R. D., Wonderlich, S. A., Mitchell, J. E., . . . Halmi, K. A. (2011). Examining the stability of DSM–IV and empirically derived eating disorder classification: Implications for DSM-5. Journal of Consulting and Clinical Psychology, 79, 777–783. http://dx.doi.org/10.1037/a0025941 Pompili, M., Girardi, P., Ruberto, A., & Tatarelli, R. (2005). Suicide in BPD: A meta-analysis. Nordic Journal of Psychiatry, 59, 319 –324. http://dx.doi.org/10.1080/08039480500320025 Preti, A., Rocchi, M. B., Sisti, D., Camboni, M. V., & Miotto, P. (2011). A comprehensive meta-analysis of the risk of suicide in eating disorders. Acta Psychiatrica Scandinavica, 124, 6 –17. http://dx.doi.org/10.1111/j .1600-0447.2010.01641.x Reas, D. L., Rø, Ø., Karterud, S., Hummelen, B., & Pedersen, G. (2013). Eating disorders in a large clinical sample of men and women with personality disorders. International Journal of Eating Disorders, 46, 801– 809. http://dx.doi.org/10.1002/eat.22176 Sheehan, D. V., Lecrubier, Y., Sheehan, K. H., Amorim, P., Janavs, J., Weiller, E., . . . Dunbar, G. C. (1998). The Mini-International Neuropsychiatric Interview (M. I. N. I.): The development and validation of a structured diagnostic psychiatric interview for DSM–IV and ICD-10. Journal of Clinical Psychiatry, 59, 22–33.
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
648
REAS, PEDERSEN, KARTERUD, AND RØ
Soloff, P. H., Lynch, K. G., Kelly, T. M., Malone, K. M., & Mann, J. J. (2000). Characteristics of suicide attempts of patients with major depressive episode and BPD: A comparative study. The American Journal of Psychiatry, 157, 601– 608. http://dx.doi.org/10.1176/appi.ajp.157.4 .601 Wedig, M. M., Silverman, M. H., Frankenburg, F. R., Reich, D. B., Fitzmaurice, G., & Zanarini, M. C. (2012). Predictors of suicide attempts in patients with BPD over 16 years of prospective follow-up. Psychological Medicine, 42, 2395–2404. http://dx.doi.org/10.1017/ S0033291712000517 Wildes, J. E., & Marcus, M. D. (2013). Alternative methods of classifying eating disorders: Models incorporating comorbid psychopathology and associated features. Clinical Psychology Review, 33, 383–394. http://dx .doi.org/10.1016/j.cpr.2013.01.006
Yen, S., Shea, M. T., Pagano, M., Sanislow, C. A., Grilo, C. M., McGlashan, T. H., . . . Morey, L. C. (2003). Axis I and axis II disorders as predictors of prospective suicide attempts: Findings from the collaborative longitudinal personality disorders study. Journal of Abnormal Psychology, 112, 375–381. http://dx.doi.org/10.1037/0021-843X.112.3 .375 Zaitsoff, S. L., & Grilo, C. M. (2010). Eating disorder psychopathology as a marker of psychosocial distress and suicide risk in female and male adolescent psychiatric inpatients. Comprehensive Psychiatry, 51, 142– 150. http://dx.doi.org/10.1016/j.comppsych.2009.03.005
Received June 6, 2014 Revision received October 29, 2014 Accepted October 31, 2014 䡲
© 2014 American Psychological Association 0022-006X/15/$12.00 http://dx.doi.org/10.1037/ccp0000014
BRIEF REPORT
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Self-Harm and Suicidal Behavior in Borderline Personality Disorder With and Without Bulimia Nervosa Deborah L. Reas
Geir Pedersen
Oslo University Hospital, Oslo, Norway
Oslo University Hospital, Oslo, Norway and University of Oslo, Oslo
Sigmund Karterud
Øyvind Rø
University of Oslo and Oslo University Hospital, Oslo, Norway
Oslo University Hospital, Oslo, Norway and University of Oslo
Objective: Few studies have investigated whether a diagnosis of Bulimia nervosa (BN) confers additional risk of life-threatening behaviors such as self-harm and suicidal behavior in borderline personality disorder (BPD). Method: Participants were 483 treatment-seeking women diagnosed with BPD according to the Structured Clinical Interview for DSM–IV Axis II Personality Disorders (SCID-II; First, Gibbon, Spitzer, Williams, & Benjamin, 1997; Diagnostic and Statistical Manual of Mental Disorders, 4th ed.; APA, 1994) and admitted to the Norwegian Network of Psychotherapeutic Day Hospitals between 1996 and 2009. Of these, 57 (11.8%) women met DSM–IV diagnostic criteria for BN according to the Mini-International Neuropsychiatric Interview (M.I.N.I.; Sheehan et al., 1998) and they were compared with women with BPD and other Axis I disorders. Results: We found that comorbid BN is uniquely and significantly associated with increased risk of suicidal behavior among women being treated for BPD. Findings underscore the importance of routinely screening for BN among women seeking treatment for BPD, as co-occurring bulimia appears to be a significant marker for immediate life-threatening behaviors in this already high-risk population, which is a significant public health issue. A significantly greater proportion of women with BPD-BN reported suicidal ideation at intake (past 7 days), engaged in self-harm behavior during treatment, and attempted suicide during treatment. All bivariate associations remained significant in the logistic regression models after controlling for mood, anxiety, and substance-related disorders. Conclusion: The presence of a concurrent diagnosis of BN among women with BPD is significantly and uniquely associated with recent suicidal ideation, and self-harm behavior and suicide attempts during treatment after controlling for major classes of mental disorders. Co-occurring BN appears to represent a significant marker for immediate life-threatening behaviors in women seeking treatment for BPD. Extra vigilance and careful monitoring of suicidal behavior during treatment is important for these individuals, and routine screening for BN is warranted.
What is the public health significance of this article? This study found that co-occurring bulimia nervosa is uniquely and significantly associated with increased risk of suicidal behavior among women being treated for borderline personality disorder. Findings underscore the importance of routinely screening for bulimia nervosa among women seeking treatment for borderline personality disorder, as co-occurring bulimia appears to be a significant marker for immediate life-threatening behaviors in this already high-risk population.
Keywords: suicide, self-harm, Bulimia nervosa, borderline personality disorder
This article was published Online First December 15, 2014. Deborah L. Reas, Regional Department of Eating Disorders, Division of Mental Health and Addiction, Oslo University Hospital, Oslo, Norway; Geir Pedersen, Department of Personality Psychiatry, Division of Mental Health and Addiction, Oslo University Hospital, Oslo, Norway, and Institute for Clinical Medicine, Faculty of Medicine, University of Oslo; Sigmund Karterud, Institute for Clinical Medicine, Faculty of Medicine, University of Oslo and Department of Research and Development, Division of Mental Health and Addiction, Oslo University Hospital, Oslo,
Norway; Øyvind Rø, Regional Department of Eating Disorders, Division of Mental Health and Addiction, Oslo University Hospital, Oslo, Norway and Institute for Clinical Medicine, Faculty of Medicine, University of Oslo, Oslo, Norway. Correspondence concerning this article should be addressed to Deborah L. Reas, Regional Department of Eating Disorders (RASP), Division of Mental Health and Addiction, Oslo University Hospital–Ullevål Hospital, P.O. Box 4956 Nydalen, N-0424 Oslo, Norway. E-mail: deborah.lynn [email protected] 643
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
644
REAS, PEDERSEN, KARTERUD, AND RØ
Borderline personality disorder (BPD) is characterized by a pervasive pattern of instability in interpersonal relationships, selfimage, and affect (APA, 2013) and is marked by impulsivity and recurrent suicidal and self-mutilating behavior. Suicidal behavior and self-harm also occur among patients with Bulimia nervosa (BN; Corcos et al., 2002; Crow et al., 2009; Franko & Keel, 2006; Preti, Rocchi, Sisti, Camboni, & Miotto, 2011) and a robust and differential pattern of comorbidity has been found between BPD and BN in clinical samples of mixed personality disorder (PD; Reas, Rø, Karterud, Hummelen, & Pedersen, 2013). Significant associations have also been found between purging behavior, personality traits, and self-injury or suicidality in women with eating disorders (ED) (Paul, Schroeter, Dahme, & Nutzinger, 2002) and adolescent psychiatric inpatients (Zaitsoff & Grilo, 2010), consistent with clinical observations linking BN with impulsivity and dysregulation. The extent to which BN in the presence of BPD might confer additional risk of suicidality and self-injury is unclear. Several studies have investigated the predictive validity of comorbidity associated with suicide risk among individuals with PD, yet these have focused largely on risk attributable to major depression, substance abuse, or anxiety disorders (Soloff, Lynch, Kelly, Malone, & Mann, 2000; Wedig et al., 2012; Yen et al., 2003), with comparatively scarce attention on the independent predictive utility of BN. To our knowledge, only two controlled investigations have specifically addressed the incremental prognostic validity of BN in samples of BPD sufferers (Chen, Brown, Harned, & Linehan, 2009; Dulit, Fyer, Leon, Brodsky, & Frances, 1994). An investigation by Dulit et al. (1994) found that BPD inpatients (N ⫽ 124) with comorbid BN were four times more likely to engage in repeated self-injury (ⱖ5 lifetime acts) than BPD inpatients without BN. An investigation of 135 women with BPD observed a significant association between BN and recurrent suicide attempts (ⱖ2 acts), but not other self-injury, after controlling for age and other non-ED Axis I disorders (Chen et al., 2009). A later study by Chen et al. (2011) found equivalent rates of lifetime suicide attempts and self-injury among 166 women and 166 men with and without ED in a diagnostically heterogeneous sample of PD. Evidence on a cross-sectional level supporting the incremental validity of BN in predicting suicidality in this high-risk population would underscore the importance of screening for BN in treatment settings for BPD. Such an approach is also consistent with recommendations that future researchers consider the effects of comorbidity when elucidating predictive effects of mental disorders on suicide attempts (Nock, Hwang, Sampson, & Kessler, 2010). To be of greatest benefit, further investigations should address previously identified methodological limitations, including (a) the merging of highly select samples from trials recruiting or screening specifically for suicidal BPD or substance-dependent BPD, which may limit ecological or clinical representativeness, and (b) small and/or diagnostically heterogeneous samples which may render effect sizes diminished due to low power. The present study aimed to investigate whether comorbidity-independent associations exist between BN and self-harm, suicidal ideation, and suicide attempts among a naturalistic, treatment-seeking sample of women with BPD.
Method Participants Participants included 483 women diagnosed with BPD, aged 18 – 65 years, with an initial admission between 1996 and 2009 to the Norwegian Network of Personality-Focused Treatment Programs. Established in 1992, this is a clinical research network providing mostly long-term, group-based (or concurrent individual– group) treatment (see also Karterud et al., 2003; Reas et al., 2013). All treatment units used uniform and standardized assessment procedures (Pedersen, Karterud, Hummelen, & Wilberg, 2013) following the longitudinal, expert, all-data (LEAD) standard, which is a comprehensive, integrative diagnostic approach using multiple sources of information (e.g., interview data, informants, behavioral observations, and medical records). Data collection is overseen by a central coordinating site responsible for quality assurance, standardization of routines, and screening data for irregularities and missing data. Raters all held professional degrees and skill acquisition, and maintenance included training courses and supervision. Ongoing monitoring of protocol adherence included checklists and periodic site visits (up to 3– 4 times annually) and in addition, site coordinators from all 16 units met every 6 months to discuss and calibrate diagnostic and clinic procedures.
Materials and Procedure Patients were interviewed with the Structured Clinical Interview for DSM–IV Axis II Personality Disorders (SCID-II; First, Gibbon, Spitzer, Williams, & Benjamin, 1997), a well-established diagnostic tool with demonstrated reliability for the assessment of PD (Lobbestael, Leurgans, & Arntz, 2011). Interrater reliability for the Norwegian version of the SCID-II has been established ( ⫽ .66 for BPD; Kvarstein et al., 2014). The Mini-International Neuropsychiatric Interview Version 4.4 (M.I.N.I.; Sheehan et al., 1998) was used to establish Axis I diagnoses, which has demonstrated reliability and validity for the assessment of Axis I disorders, including BN ( ⫽ .78; Sheehan et al., 1998). The Norwegian version of the M.I.N.I is validated and is considered a timeefficient and feasible alternative to the SCID-P (SCID-I/P; First et al., 2002) and CIDI (CIDI; Kessler & Ustün, 2004) (Mordal, Gundersen, & Bramness, 2010).
Assessment of Suicidal Behavior The assessment of suicidality included clinician- and selfreported data capturing different time epochs. First, self-reported data were systematically collected at intake using a routinely administered sociodemographic questionnaire. Suicidal ideation was assessed dichotomously to capture the past 7 days and past 12 months, that is, “Have you had thoughts about taking your own life?” Self-harm behavior was assessed dichotomously to capture the past 12 months and lifetime, that is, “Have you [ever] physically hurt yourself, for example, cutting, scratching, burning, headbanging, and so forth?” Suicide attempts were assessed dichotomously to capture the past 12 months and lifetime, that is, “Have you ever tried to kill yourself?” The number of lifetime suicide attempts was also rated. Clinician-rated data were systematically collected at discharge using a routine discharge form to assess (a)
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
SELF-HARM AND SUICIDAL BEHAVIOR IN BPD-BN
the occurrence of self-harm behavior during treatment, that is, physically harmful behaviors such as cutting, scratching, burning, banging against hard objects, and so forth; (b) suicidal ideation during treatment, that is, “Did the patient express thoughts about taking one’s life, excluding very fleeting or dramatic expressions regarding the wish to die?” and (c) suicide attempts occurring during treatment, that is, “lethal acts with the intent to die that would have been successful without acute intervention from others.” It should be noted that assessment of self-harm behavior did not specify expectations (e.g., to gain relief from negative feelings, relieve suffering, resolve interpersonal difficulties) as specified in DSM-5 (APA, 2013) for nonsuicidal self-injury; thus, we use the more general term of self-harm. Bulimic behaviors, sometimes conceptualized under the rubric of self-harm, were not covered by assessment. All data were registered in a central, anonymous database administered by Oslo University Hospital. All patients provided written consent and the study was approved by the State Data Inspectorate and the Regional Ethics Committee.
Data Analyses Analyses were conducted using predictive analysis software (PASW) Version 18.0. Patients were grouped according to the presence of BN (BPD-BN) or non-ED Axis I disorder (BPDother), in line with grouping methods by Chen et al. (2011). Cases of anorexia nervosa (AN) (N ⫽ 7) and eating disorders not otherwise specified (EDNOS) (N ⫽ 81) were excluded from BPD-other due to potential confounding of subthreshold BN in the EDNOS group, or lifetime history of BN, because diagnostic fluctuation or crossover is common in DSM-IV ED (Peterson et al., 2011). Chi-square analyses tested differences for categorical variables. Consistent with methods from earlier studies (Bodell, Joiner, & Keel, 2013), logistic regression analyses (ORs and 95% CIs) controlled for mood, anxiety, and substance-use disorders were
645
conducted when significant bivariate associations were detected. Following guidelines detailed in previous research (Chen, Cohen, & Chen, 2010), ORs of 1.68, 3.47, and 6.71 were considered equivalent to small, medium, and large effect sizes (Cohen’s d ⫽ 0.2, 0.5, and 0.8, respectively). Analyses were two-tailed (p ⬍ .05).
Results Sample Characteristics A total of 57 patients (11.8%) received a comorbid diagnosis of BN and BPD and were grouped as BPD-BN. All other patients (N ⫽ 426) were diagnosed with at least one non-ED Axis I diagnosis (BPD-other). Approximately 68% of both groups had mood disorders; 63.2% versus 52.6% (p ⫽ .123) were diagnosed with anxiety disorder; and 28.1% versus 15.1% (p ⫽ .014) had substance-use disorder. Table 1 shows no significant baseline differences for age, mean global assessment of functioning (GAF) at intake, length of treatment, mean frequency of non-ED Axis I or Axis II disorders, and number of SCID-II BPD criteria fulfilled. No differences were detected for marital status, 2(4, 459) ⫽ 2.97, p ⫽ .562, or work situation, 2(5, 421) ⫽ 4.91, p ⫽ .427.
Clinician and Self-Reported Self-Harm, Suicidal Ideation, and Suicide Attempts As shown in Table 1, a significantly greater proportion of women in the BPD-BN group demonstrated self-harm behavior during treatment (p ⬍ .001). Approximately 50% of both groups engaged in self-harm behavior over the past 12 months. Lifetime self-harm did not show a statistically significant difference between BPD-BN and BPD-other (70.9% vs. 58.2%; p ⫽ .079). A significantly higher proportion of BPD-BN reported suicidal ide-
Table 1 Sample Characteristics for BPD-BN Versus BPD-Other (N ⫽ 483) Variable Age in years (M, SD) Length of admission (days), M (SD) Global assessment functioning (GAF), M (SD) # SCID-II criteria met for BPD, M (SD) # Axis II disorders, M (SD) # Non-ED Axis I disorders, M (SD) Mood disorder, yes (%) Anxiety disorder, yes (%) Substance-use disorder, yes (%) Self-harm during treatment, yes (%)a Self-harm, past 12 months, yes (%) Self-harm, lifetime, yes (%) Suicidal ideation during treatment, yes (%)a Suicidal ideation, past 7 days, yes (%) Suicidal ideation, past 12 months, yes (%) Suicide attempts during treatment, yes (%)a Suicide attempts, past 12 months, yes (%) Suicide attempts, lifetime, yes (%) Suicide attempts, recurrent, yes (%)b
BPD-BN (n ⫽ 57) BPD-other (n ⫽ 426) 29.9 (7.42) 136.7 (58.7) 45.1 (4.30) 5.98 (1.09) 1.51 (0.66) 1.95 (1.28) 68.4% 63.2% 28.1% 39.3% 46.7% 70.9% 52.6% 55.6% 88.4% 10.9% 31.8% 64.8% 45.1%
30.2 (7.32) 136.3 (58.4) 45.7 (5.12) 6.06 (1.08) 1.58 (0.76) 1.95 (1.27) 67.9% 52.6.% 15.1% 15.7% 46.4% 58.2% 39.5% 36.0% 77.4% 4.1% 22.8% 58.0% 35.4%
t/2
p value
0.32 .054 1.01 0.47 0.67 ⫺0.09 0.00 2.38 6.10 18.3 .002 3.22 3.59 6.31 2.68 4.79 1.70 0.91 1.84
.750 .957 .317 .641 .503 .993 .940 .123 .014 .001 .969 .079 .058 .012 .101 .029 .192 .340 .175
Note. BPD ⫽ borderline personality disorder; BN ⫽ Bulimia nervosa; no completed suicides occurred during treatment. a Indicates clinician-reported data. b Recurrent is defined as ⱖ 2 acts.
REAS, PEDERSEN, KARTERUD, AND RØ
646
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
ation during the past 7 days prior to intake (p ⫽ .012), but differences in suicidal ideation (past 12 months) were not significant (88.4% vs. 77.4%, p ⫽ .101). A trend was detected for greater suicidal ideation during treatment in BPD-BN (p ⫽ .058). A significantly higher proportion of women with BPD-BN attempted suicide during treatment (p ⫽ .029). No significant differences were found for suicide attempts during the past 12 months (31.8% vs. 22.8%, p ⫽ .192), lifetime, or recurrent suicide attempts (two or more acts). For suicide attempters, mean (SD) number of lifetime attempts was 3.0 (2.24) for BPD-BN versus 2.80 (2.68) for BPD-other, respectively, t(255) ⫽ .401, p ⫽ .688.
Logistic Regression Table 2 shows significant associations between BN and selfharm during treatment OR ⫽ 3.23, 95% CI [1.76 –5.92], suicidal ideation past 7 days, OR ⫽ 2.37, 95% CI [1.23–4.55]; and suicide attempts during treatment OR ⫽ 2.83; 95% CI [1.05–7.64]. An OR of 1.73, 95% CI [.987–3.04], p ⫽ .055 was detected for suicidal ideation during treatment, indicating a small effect.
Discussion These findings indicate that the presence of a comorbid diagnosis of BN in the context of BPD is significantly and uniquely associated with increased risk of recent suicidal ideation at intake and self-harm and suicide attempts during treatment after controlling for mood, anxiety, and substance-related disorders. As such, a concurrent diagnosis of BN among women seeking treatment for BPD appears to represent a strong and significant marker for immediate life-threatening behaviors. Death by suicide occurs in 8 –10% of individuals with BPD, which is among the highest rates of all mental disorders (Pompili, Girardi, Ruberto, & Tatarelli, 2005). Our study has indicated that bulimic episodes signal even greater risk of suicide attempts within this high-risk population. Findings echo those by Bodell et al. (2013), who found comorbidity-independent associations between BN and lifetime suicidality among university women, prompting calls for a standard risk assessment of suicide among women with BN. The sample size and setting were considered study strengths, that is, over 400 consecutively admitted female patients seeking day treatment for BPD. No significant differences were observed in length of admission (i.e., approximately 4.5 months) or length of referral process; thus, these variables were not
considered potential confounds. Data collection was part of routine clinic procedure, and although risk of bias cannot be eliminated, only minimal bias owing to clinician expectations regarding the present study aims were expected to influence results. Several limitations of our study are important to consider. Despite face validity, the reliance on direct, single-item assessments limited the scope of the measurement. Information regarding lethality, intent, or specific methods of self-harm was unavailable, and behaviors such as aborted, interrupted, and low-lethality attempts might have not been captured. Prior research, however, has used single-item, self-report assessments of suicidal ideation and attempt with demonstrated validity (Bodell et al., 2013; Cougle, Keough, Riccardi, & SachsEricsson, 2009). Retrospective self-report data on the occurrence of self-harm and suicidal behavior might be subject to recall bias. This study design was cross-sectional, thereby precluding the ability to infer stability of findings and the longitudinal risk or other clinical outcomes (e.g., suicide completion). Self-harm behavior, suicidal ideation, and past suicide attempts have been documented as important risk factors for future suicide (Klonsky, May, & Glenn, 2013), also in samples with BPD (Wedig et al., 2012). The overall rate of lifetime attempts in our sample was 59%, which is higher than typically observed in BN (25–35%; Franko & Keel, 2006), although within the 40 – 85% range observed for BPD (Oumaya et al., 2008). Our sample included treatment-seeking women admitted for intensive and specialized day-hospital treatment for PD; thus, results may not generalize to individuals under 18 or over 65 years of age, or to community or nonspecialist treatment settings which serve less severe populations. Because the pattern of findings indicated significantly elevated risk of suicidality at intake and during treatment, but not prior to treatment, replication is necessary to rule out potential state effects underlying results. Further investigation is needed to explore whether findings relate to appropriateness of therapeutic setting or treatment approach, for example, or whether suicidality might constitute a particularly salient motivator for treatment-seeking among women with BPD-BN. Several alternative classification schemes for subtyping BN according to patterned heterogeneity in comorbidity (i.e., multi-impulsive BN, borderline–nonborderline BN, undercontrolled– externalizing; see Wildes & Marcus, 2013) may have relevance for the contextualization and conceptualization of findings, or our re-
Table 2 Logistic Regression Models for the Association Between BN and Self-Harm, Suicidal Ideation, and Suicide Attempts Among Women With BPD Controlling for Mood, Anxiety, and Substance-Use Disorders
Illness Mood disorder Anxiety disorder Substance disorder BN
Self-harm (during tx) OR 95% CI 1.24 0.76 1.47 3.23ⴱⴱ
[0.74–2.09] [0.47–1.23] [0.81–2.67] [1.76–5.92]
Suicidal ideation (during tx) OR 95% CI 0.89 0.97 0.94 1.73 (p ⫽ .055)
[0.61–1.33] [0.66–1.42] [.567–1.54] [.987–3.04]
Suicidal ideation (past 7 days) OR 95% CI 1.49 1.37 0.90 2.37ⴱⴱ
[0.91–2.44] [0.86–2.19] [0.51–1.61] [1.23–4.55]
Suicide attempts (during tx) OR 95% CI 1.06 1.21 1.23 2.83ⴱ
Note. BPD ⫽ borderline personality disorder; BN ⫽ Bulimia nervosa; tx ⫽ treatment; OR ⫽ odds ratio; CI ⫽ 95% confidence interval. p ⱕ .05. ⴱⴱ p ⱕ .01.
ⴱ
[0.42–2.65] [0.49–2.96] [0.43–3.47] [1.05–7.64]
SELF-HARM AND SUICIDAL BEHAVIOR IN BPD-BN
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
sults may speak broadly to cross-cutting behavioral and neurobiologically informed constructs such as trait impulsivity or cognitive control (Insel, 2014). This was a controlled investigation providing evidence on a cross-sectional level supporting the incremental validity of BN in predicting suicidality beyond mood, anxiety, and substancerelated disorders in the high-risk BPD population. Findings warrant routine assessment of BN in treatment-seeking samples of women with BPD, and underscore the importance of high vigilance and fastidious monitoring of suicidal behaviors during treatment for these individuals.
References American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th ed.). Washington, DC: Author. American Psychiatric Association, DSM-5 Task Force. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Arlington, VA: American Psychiatric Publishing. Bodell, L. P., Joiner, T. E., & Keel, P. K. (2013). Comorbidity-independent risk for suicidality increases with BN but not with anorexia nervosa. Journal of Psychiatric Research, 47, 617– 621. http://dx.doi.org/ 10.1016/j.jpsychires.2013.01.005 Chen, H., Cohen, P., & Chen, S. (2010). How big is a big odds ratio? Interpreting the magnitudes of odds ratios in epidemiological studies. Communications in Statistics: Simulation and Computation, 39, 860 – 864. http://dx.doi.org/10.1080/03610911003650383 Chen, E. Y., Brown, M. Z., Harned, M. S., & Linehan, M. M. (2009). A comparison of BPD with and without eating disorders. Psychiatry Research, 170, 86 –90. http://dx.doi.org/10.1016/j.psychres.2009.03.006 Chen, E. Y., McCloskey, M. S., Michelson, S., Gordon, K. H., & Coccaro, E. (2011). Characterizing eating disorders in a personality disorders sample. Psychiatry Research, 185, 427– 432. Corcos, M., Taïeb, O., Benoit-Lamy, S., Paterniti, S., Jeammet, P., & Flament, M. F. (2002). Suicide attempts in women with BN: Frequency and characteristics. Acta Psychiatrica Scandinavica, 106, 381–386. http://dx.doi.org/10.1034/j.1600-0447.2002.02318.x Cougle, J. R., Keough, M. E., Riccardi, C. J., & Sachs-Ericsson, N. (2009). Anxiety disorders and suicidality in the National Comorbidity Survey– Replication. Journal of Psychiatric Research, 43, 825– 829. http://dx.doi .org/10.1016/j.jpsychires.2008.12.004 Crow, S. J., Peterson, C. B., Swanson, S. A., Raymond, N. C., Specker, S., Eckert, E. D., & Mitchell, J. E. (2009). Increased mortality in BN and other eating disorders. The American Journal of Psychiatry, 166, 1342– 1346. http://dx.doi.org/10.1176/appi.ajp.2009.09020247 Dulit, R. A., Fyer, M. R., Leon, A. C., Brodsky, B. S., & Frances, A. J. (1994). Clinical correlates of self-mutilation in BPD. The American Journal of Psychiatry, 151, 1305–1311. First, M. B., Gibbon, M., Spitzer, R. L., Williams, J. B. W., & Benjamin, L. S. (1997). Structured Clinical Interview for DSM–IV Axis II Personality Disorders. Washington, DC: American Psychiatric Association. First, M. B., Spitzer, R. L., Gibbon, M., & Williams, J. B. W. (2002). Structured Clinical Interview for DSM-IV-TR Axis I Disorders. Research Version, Patient Edition. (SCID-I/P) New York: Biometrics Research, New York State Psychiatric Institute. Franko, D. L., & Keel, P. K. (2006). Suicidality in eating disorders: Occurrence, correlates, and clinical implications. Clinical Psychology Review, 26, 769 –782. http://dx.doi.org/10.1016/j.cpr.2006.04.001 Insel, T. R. (2014). The NIMH Research Domain Criteria (RDoC) Project: Precision medicine for psychiatry. The American Journal of Psychiatry, 171, 395–397. http://dx.doi.org/10.1176/appi.ajp.2014.14020138 Karterud, S., Pedersen, G., Bjordal, E., Brabrand, J., Friis, S., Haaseth, O., . . . Urnes, O. (2003). Day treatment of patients with personality disorders: Experiences from a Norwegian treatment research network.
647
Journal of Personality Disorders, 17, 243–262. http://dx.doi.org/ 10.1521/pedi.17.3.243.22151 Kessler, R. C., & Ustün, T. B. (2004). The World Mental Health (WMH) Survey Initiative version of the World Health Organization (WHO) Composite International Diagnostic Interview (CIDI). International Journal of Methods in Psychiatric Research, 13, 93–121. Klonsky, E. D., May, A. M., & Glenn, C. R. (2013). The relationship between nonsuicidal self-injury and attempted suicide: Converging evidence from four samples. Journal of Abnormal Psychology, 122, 231– 237. http://dx.doi.org/10.1037/a0030278 Kvarstein, E. H., Pedersen, G., Urnes, O., Hummelen, B., Wilberg, T., & Karterud, S. (2014). Changing from a traditional psychodynamic treatment programme to mentalization-based treatment for patients with BPD: Does it make a difference? Psychology and Psychotherapy: Theory, Research and Practice. Advance online publication. http://dx.doi .org/10.1111/papt.12036 Lobbestael, J., Leurgans, M., & Arntz, A. (2011). Inter-rater reliability of the Structured Clinical Interview for DSM–IV Axis I Disorders (SCID I) and Axis II Disorders (SCID II). Clinical Psychology & Psychotherapy, 18, 75–79. http://dx.doi.org/10.1002/cpp.693 Mordal, J., Gundersen, O., & Bramness, J. G. (2010). Norwegian version of the Mini-International Neuropsychiatric Interview: Feasibility, acceptability and test-retest reliability in an acute psychiatric ward. European Psychiatry, 25, 172–177. http://dx.doi.org/10.1016/j.eurpsy.2009 .02.004 Nock, M. K., Hwang, I., Sampson, N. A., & Kessler, R. C. (2010). Mental disorders, comorbidity and suicidal behavior: Results from the National Comorbidity Survey Replication. Molecular Psychiatry, 15, 868 – 876. http://dx.doi.org/10.1038/mp.2009.29 Oumaya, M., Friedman, S., Pham, A., Abou Abdallah, T., Guelfi, J.-D., & Rouillon, F. (2008). Personnalité borderline, automutilations et suicide: Revue de la littérature [BPD, self-mutilation and suicide: Literature review]. L’Encéphale, 34, 452– 458. http://dx.doi.org/10.1016/j.encep .2007.10.007 Paul, T., Schroeter, K., Dahme, B., & Nutzinger, D. O. (2002). Selfinjurious behavior in women with eating disorders. The American Journal of Psychiatry, 159, 408 – 411. http://dx.doi.org/10.1176/appi.ajp.159 .3.408 Pedersen, G., Karterud, S., Hummelen, B., & Wilberg, T. (2013). The impact of extended longitudinal observation on the assessment of personality disorders. Personality and Mental Health, 7, 277–287. http:// dx.doi.org/10.1002/pmh.1234 Peterson, C. B., Crow, S. J., Swanson, S. A., Crosby, R. D., Wonderlich, S. A., Mitchell, J. E., . . . Halmi, K. A. (2011). Examining the stability of DSM–IV and empirically derived eating disorder classification: Implications for DSM-5. Journal of Consulting and Clinical Psychology, 79, 777–783. http://dx.doi.org/10.1037/a0025941 Pompili, M., Girardi, P., Ruberto, A., & Tatarelli, R. (2005). Suicide in BPD: A meta-analysis. Nordic Journal of Psychiatry, 59, 319 –324. http://dx.doi.org/10.1080/08039480500320025 Preti, A., Rocchi, M. B., Sisti, D., Camboni, M. V., & Miotto, P. (2011). A comprehensive meta-analysis of the risk of suicide in eating disorders. Acta Psychiatrica Scandinavica, 124, 6 –17. http://dx.doi.org/10.1111/j .1600-0447.2010.01641.x Reas, D. L., Rø, Ø., Karterud, S., Hummelen, B., & Pedersen, G. (2013). Eating disorders in a large clinical sample of men and women with personality disorders. International Journal of Eating Disorders, 46, 801– 809. http://dx.doi.org/10.1002/eat.22176 Sheehan, D. V., Lecrubier, Y., Sheehan, K. H., Amorim, P., Janavs, J., Weiller, E., . . . Dunbar, G. C. (1998). The Mini-International Neuropsychiatric Interview (M. I. N. I.): The development and validation of a structured diagnostic psychiatric interview for DSM–IV and ICD-10. Journal of Clinical Psychiatry, 59, 22–33.
This document is copyrighted by the American Psychological Association or one of its allied publishers. This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
648
REAS, PEDERSEN, KARTERUD, AND RØ
Soloff, P. H., Lynch, K. G., Kelly, T. M., Malone, K. M., & Mann, J. J. (2000). Characteristics of suicide attempts of patients with major depressive episode and BPD: A comparative study. The American Journal of Psychiatry, 157, 601– 608. http://dx.doi.org/10.1176/appi.ajp.157.4 .601 Wedig, M. M., Silverman, M. H., Frankenburg, F. R., Reich, D. B., Fitzmaurice, G., & Zanarini, M. C. (2012). Predictors of suicide attempts in patients with BPD over 16 years of prospective follow-up. Psychological Medicine, 42, 2395–2404. http://dx.doi.org/10.1017/ S0033291712000517 Wildes, J. E., & Marcus, M. D. (2013). Alternative methods of classifying eating disorders: Models incorporating comorbid psychopathology and associated features. Clinical Psychology Review, 33, 383–394. http://dx .doi.org/10.1016/j.cpr.2013.01.006
Yen, S., Shea, M. T., Pagano, M., Sanislow, C. A., Grilo, C. M., McGlashan, T. H., . . . Morey, L. C. (2003). Axis I and axis II disorders as predictors of prospective suicide attempts: Findings from the collaborative longitudinal personality disorders study. Journal of Abnormal Psychology, 112, 375–381. http://dx.doi.org/10.1037/0021-843X.112.3 .375 Zaitsoff, S. L., & Grilo, C. M. (2010). Eating disorder psychopathology as a marker of psychosocial distress and suicide risk in female and male adolescent psychiatric inpatients. Comprehensive Psychiatry, 51, 142– 150. http://dx.doi.org/10.1016/j.comppsych.2009.03.005
Received June 6, 2014 Revision received October 29, 2014 Accepted October 31, 2014 䡲
