Borderline Symptom List – 23 (BSL-23)

1. Abstract

The Borderline Symptom List – 23 (BSL-23) is a widely utilized, highly sensitive self-report psychometric instrument developed to assess the subjective severity and core phenomenology of borderline personality disorder (BPD) symptomatology. Derived empirically from the original 95-item Borderline Symptom List (BSL-95) using advanced item response theory and classical test theory principles, the BSL-23 captures essential symptom clusters across affect dysregulation, self-perception disturbances, chronic emptiness, interpersonal mistrust, and autodestructive tendencies. Comprising 23 items scored on a 5-point Likert-type scale ranging from 0 (“not at all”) to 4 (“very much”), the instrument quantifies patient distress over the preceding week, yielding a mean score between 0 and 4 or a sum score between 0 and 92. A single visual analog scale assessing overall psychological well-being is frequently administered concurrently.

Extensive psychometric investigations consistently demonstrate exceptional internal consistency (Cronbach’s alpha typically exceeding 0.93 to 0.97 across inpatient, outpatient, and general population cohorts) and high test-retest reliability ($r_{tt} \approx 0.82$). Exploratory and confirmatory factor analyses robustly substantiate a dominant unidimensional general factor reflecting core borderline pathology, alongside identifiable facet structures. Convergent validity is confirmed through robust positive correlations with the Beck Depression Inventory, the Brief Symptom Inventory, and DSM-IV/DSM-5 diagnostic interviews for BPD, while discriminant validity distinguishes BPD patients from clinical controls diagnosed with major depressive disorder, anxiety disorders, and schizophrenia. The measure exhibits profound sensitivity to week-by-week change, establishing it as an international gold-standard primary outcome instrument in clinical trials evaluating Dialectical Behavior Therapy (DBT), schema therapy, and mentalization-based treatment.

2. Keywords

Borderline Symptom List – 23, BSL-23, borderline personality disorder, psychometrics, affect dysregulation, dialectical behavior therapy, outcome measurement, self-report inventory, self-harm, emotional vulnerability

3. Authors

The Borderline Symptom List – 23 was developed and psychometrically validated by a collaborative team of clinical psychologists and psychiatrists specializing in personality disorders and dialectical behavior therapy:

• Martin Bohus, M.D. — Professor and Chair of Psychosomatic Medicine and Psychotherapy, Central Institute of Mental Health (ZI Mannheim), Heidelberg University, Mannheim, Germany.
• Mona F. Limberger, Ph.D. — Department of Psychosomatic Medicine and Psychotherapy, Central Institute of Mental Health, Mannheim, Germany.
• Ursula Frank, Dipl.-Psych. — Central Institute of Mental Health, Mannheim, Germany.
• Alexander L. Chapman, Ph.D., R.Psych. — Professor of Psychology, Department of Psychology, Simon Fraser University, Burnaby, British Columbia, Canada.
• Thomas Kühler, Ph.D. — Methodology and Statistics Consultant, Mannheim, Germany.
• Rolf-Dieter Stieglitz, Ph.D. — Professor of Clinical Psychology and Psychiatry, Department of Clinical Psychology and Psychiatry, University of Basel, Basel, Switzerland.

4. Purpose

The primary clinical and empirical purpose of the Borderline Symptom List – 23 (BSL-23) is to deliver a time-efficient, psychometrically rigorous, and clinically responsive index of subjective symptom burden in individuals diagnosed with or presenting traits of Borderline Personality Disorder (BPD). Personality pathology has historically posed acute challenges to measurement; traditional diagnostic interviews such as the Structured Clinical Interview for DSM Disorders (SCID) provide categorical diagnostic classifications but are poorly suited for tracking acute therapeutic fluctuations, symptom remission, or week-by-week treatment response. Conversely, general psychopathology scales (such as the SCL-90-R) fail to encapsulate the distinctive interplay of self-loathing, emotional turbulence, stress-induced dissociation, and autodestructive impulses characteristic of BPD.

To overcome the clinical burden of the comprehensive 95-item instrument (BSL-95), which required significant cognitive effort from patients already experiencing severe distress, Bohus and colleagues developed the BSL-23. The tool serves multiple interrelated functions: it acts as a primary outcome measure in randomized controlled clinical trials evaluating manualized psychotherapies; provides ongoing progress monitoring in routine clinical practice; facilitates clinical risk appraisal related to self-injury and suicidal ideation; and enables dimensional phenotyping in clinical research studies. By focusing explicitly on the subjective experiential distress of the patient over the preceding week, the BSL-23 empowers clinicians to detect subtle changes in affect regulation, self-contempt, and interpersonal tension, directly informing safety planning and therapeutic interventions.

5. Psychological Construct

The BSL-23 measures the composite psychopathological construct of borderline personality pathology, conceptualized as a pervasive, multi-systemic pattern of emotional dysregulation, self-identity disturbance, interpersonal instability, and behavioral dyscontrol. Although the instrument demonstrates a strong unidimensional general factor of BPD severity, the items thoroughly map onto four clinically salient phenotypic dimensions:

Affective Dysregulation and Heightened Vulnerability

Affect dysregulation represents the operational core of BPD pathology. Patients experience heightened baseline emotional sensitivity, rapid oscillations in affective valence, and intense, protracted emotional reactivity. In the BSL-23, this is captured by items such as “My feelings oscillated back and forth” (Item 15), “My mood was rapidly changing” (Item 16), and “I experienced a painful state of inner tension” (Item 19). The construct incorporates not only dysphoria and affective lability, but also unmodulated aversive emotional states such as terror (“I felt terrified”, Item 20) and sudden, uncontrollable hostility (“I had intense rage”, Item 23).

Self-Image Pathology, Self-Hatred, and Worthlessness

Disturbances in identity and the self-concept are profound sources of disability in BPD. Rather than mere low self-esteem, borderline self-pathology involves severe self-alienation, intense self-disgust, and moral invalidation. This is manifested in items evaluating perceived inadequacy (“I felt like a loser”, Item 6; “I felt worthless”, Item 7), extreme moral guilt and internal shame (“I felt ashamed”, Item 13; “I did not believe that I had any rights”, Item 4), and actively aggressive self-directed contempt (“I was disgusted by myself”, Item 10; “I hated myself”, Item 12).

Autodestructive Cognitions and Suicidality

Self-directed aggression serves both as a maladaptive affect-regulation mechanism and as an expression of internal self-punitive directives. The scale captures these phenomena across cognitive appraisal and motivational drive: conscious urges toward deliberate self-harm (“I thought about hurting myself”, Item 8), internal demands for self-chastisement (“I wanted to punish myself”, Item 9), explicit suicidal ideation (“I thought about suicide”, Item 11), and a morbid cognitive orientation toward death (“The thought of death had a certain attraction for me”, Item 18). These items capture the escalating continuum of self-destructive behavioral impulses that characterize acute borderline crises.

Interpersonal Alienation, Helplessness, and Cognitive Dyscontrol

Interpersonal schemas in BPD are dominated by expectations of betrayal, abandonment, and persecutory intent. The scale operationalizes these through pervasive mistrust (“I mistrusted other people”, Item 22) and deep social disconnection (“I felt lonely”, Item 5). Concurrent with these interpersonal disruptions are cognitive disruptions stemming from severe stress, including cognitive fragmentation (“It was hard for me to concentrate”, Item 1), neurovegetative/somatosensory manifestations (“I felt dizzy”, Item 14), profound demoralization (“Everything seemed hopeless to me”, Item 17; “I felt helpless”, Item 2), and existential fears of mental disintegration (“I was afraid of losing control”, Item 21).

6. Theoretical Framework

The construction and clinical application of the BSL-23 are anchored in contemporary evidence-based paradigms of borderline personality pathology, most notably Marsha Linehan’s Biosocial Model, Jeffrey Young’s Schema Therapy formulation, and John Gunderson’s Interpersonal Hypersensitivity model.

According to Linehan’s biosocial framework, borderline pathology arises from a pervasive transactional dynamic between a biologically determined emotional vulnerability (characterized by high sensitivity, high reactivity, and a slow return to baseline) and an emotionally invalidating developmental environment. In this framework, behaviors such as self-harm, cognitive collapse, and explosive anger function as dysfunctional, learned attempts to regulate unbearable internal tension. The BSL-23 reflects this formulation directly: items do not merely gauge static traits, but rather assess the dynamic, aversive affective surges and subsequent maladaptive regulatory strategies (such as thoughts of self-punishment) that define the biosocial transaction.

Furthermore, from the standpoint of Schema Therapy, the BSL-23 evaluates acute activations of maladaptive early schema modes, specifically the “Vulnerable Child” (experiencing helplessness, terror, and profound loneliness), the “Punitive/Demanding Parent” (manifested in severe self-directed disgust, self-loathing, and feeling stripped of basic human rights), and the “Angry Child” (intense, dysregulated rage). The operational focus of the BSL-23 on the past week allows clinicians to capture real-time mode shifts and internal conflict, making it uniquely compatible with modern process-based and cognitive-behavioral treatment modalities.

7. Validity

The psychometric validity of the BSL-23 has been validated across diverse multinational samples, confirming robust construct, convergent, discriminant, and predictive validity:

• Construct and Convergent Validity: In the original validation study by Bohus et al. (2009), the BSL-23 demonstrated an exceptional correlation with its parent instrument, the 95-item BSL ($r = 0.96$), proving that the abbreviated scale retains the comprehensive measurement coverage of the full inventory. Convergent validity is evidenced by substantial, statistically significant correlations with established clinical indices of emotional distress, including the Beck Depression Inventory (BDI; $r = 0.77$ to $0.85$), the Brief Symptom Inventory (BSI Global Severity Index; $r = 0.76$ to $0.84$), and the state scale of the State-Trait Anxiety Inventory (STAI; $r = 0.68$).
• Discriminant Validity: The BSL-23 effectively differentiates patients with BPD from other clinical populations and healthy individuals. In comparative clinical trials, patients meeting full DSM criteria for BPD scored significantly higher on the BSL-23 (mean scores typically $2.0 – 2.5$) than psychiatric control patients diagnosed with major depressive disorder, social anxiety disorder, or schizophrenia (mean scores typically $0.8 – 1.3$), and far above non-clinical community controls (mean scores typically $0.15 – 0.35$). Area Under the Curve (AUC) analyses in receiver operating characteristic (ROC) evaluations consistently exceed 0.90, demonstrating exceptional diagnostic sensitivity and specificity.
• Sensitivity to Change and Predictive Validity: The BSL-23 was specifically constructed to reflect therapeutic progress. In prospective intervention trials, standardized response means (SRMs) and Cohen’s $d$ effect sizes for pre-to-post DBT and schema therapy interventions regularly range from 0.80 to 1.40, indicating powerful responsiveness to longitudinal clinical recovery.

8. Reliability

The reliability parameters of the BSL-23 consistently meet or exceed the rigorous standards mandated for diagnostic and outcome assessment instruments:

• Internal Consistency: In the seminal development and validation sample of 659 patients with BPD (Bohus et al., 2009), the BSL-23 exhibited an internal consistency of Cronbach’s $\alpha = 0.97$. Subsequent independent cross-cultural validation trials—including validation in Spanish (Soler et al., 2013; $\alpha = 0.94$), French (Nicastro et al., 2016; $\alpha = 0.96$), Turkish (Güleç et al., 2014; $\alpha = 0.93$), Swedish (Kjaersdam Telléus et al., 2016; $\alpha = 0.95$), and Arabic (Al-Hadi et al., 2019; $\alpha = 0.94$)—have replicated these high internal consistency metrics across inpatient, outpatient, and community populations. McDonald’s omega ($\omega$) coefficients consistently mirror alpha values, remaining above 0.95.
• Test-Retest Reliability: Stability evaluations conducted over brief time intervals (one to two weeks) in clinically stable BPD cohorts yield test-retest reliability coefficients ranging between $r_{tt} = 0.82$ and $r_{tt} = 0.89$. Because the instrument is intentionally designed to be sensitive to dynamic clinical fluctuations, the observed test-retest values represent an optimal balance between measurement stability and clinical responsiveness.

9. Factor Analysis

During the initial development of the scale, exploratory factor analysis (EFA) utilizing principal component analysis with scree test evaluations revealed a distinct and predominant first factor accounting for approximately 48% to 54% of the total variance, demonstrating that the 23 items coalesce into a dominant primary construct representing borderline symptom severity.

Subsequent confirmatory factor analyses (CFA) conducted across several independent clinical cohorts have evaluated both unidimensional and multi-factorial models:

• Unidimensional Model: A single-factor model exhibits acceptable to strong fit indices when correlated residuals between conceptually paired items (e.g., Item 15 “oscillating feelings” and Item 16 “rapidly changing mood”; Item 11 “suicide” and Item 18 “attraction of death”) are accommodated: Comparative Fit Index (CFI) $ge 0.92$, Tucker-Lewis Index (TLI) $ge 0.91$, Root Mean Square Error of Approximation (RMSEA) $\approx 0.055 – 0.068$, and Standardized Root Mean Square Residual (SRMR) $le 0.045$.
• Bifactor Model: A bifactor structure incorporating one general borderline pathology factor ($G$-factor) alongside four specific orthogonal group factors (Affective Lability, Self-Contempt, Autodestructive Impulses, and Alienation/Helplessness) achieves superior statistical fit (CFI $> 0.96$, TLI $> 0.95$, RMSEA $< 0.048$). The general factor reliably accounts for more than 85% of the common variance (Explained Common Variance; ECV$> 0.85$), strongly validating the scoring convention of deriving a single global borderline severity index.
• Item Loadings: Standardized factor loadings across all 23 items are uniformly high, ranging from $0.62$ to $0.88$, confirming that each individual indicator contributes robust informational value to the composite latent construct.

10. Instrument / Measurement Tool

• Instrument Name: Borderline Symptom List – 23 (BSL-23)
• Original Authors: Martin Bohus, Mona F. Limberger, Ursula Frank, Alexander L. Chapman, Thomas Kühler, and Rolf-Dieter Stieglitz (2009)
• Primary Target Population: Adolescents and adults (aged 14 and older) presenting with core borderline personality pathology or undergoing specialized personality disorder interventions.
• Test Format: 23-item self-administered psychological rating scale (paper-and-pencil or digital administration).
• Administration Time: Approximately 3 to 5 minutes.
• Recall Period: The preceding week (“Please assess each statement with respect to the past week”).
• Response Format: 5-point Likert-type scale: 0 = not at all, 1 = a little, 2 = rather, 3 = much, 4 = very much.
• Scoring Rules:
  • Overall Mean Score: Sum of all answered items divided by 23 (range: 0.00 to 4.00). Missing items are handled by calculating the mean of the completed items, provided at least 21 items are answered.
  • Sum Score: Direct arithmetic summation of all item ratings (range: 0 to 92).
  • Reverse-Scored Items: None. All items are scored in the direct pathological direction.
• Supplemental Scale: A visual analog scale (VAS) measuring current global state/well-being from 0% (“completely bad/miserable”) to 100% (“excellent/very well”) is included at the conclusion of the instrument.

11. Permissions & Fee and Test Year

The Borderline Symptom List – 23 (BSL-23) was officially published in 2009 by Bohus and colleagues in the peer-reviewed journal Psychopathology. The authors and the Central Institute of Mental Health (ZI Mannheim) provide open, fee-free access to the BSL-23 for academic, scientific research, and routine non-commercial clinical practices. It is widely translated into more than 20 languages and can be obtained via academic repositories, clinical training networks, or the developers’ institutional affiliations. Commercial utilization, integration into proprietary electronic health record systems, or inclusion in sponsored pharmaceutical trials typically requires formal permission from the copyright holders.

12. References

Al-Hadi, A. N., Al-Subaie, A. S., Al-Ghamdi, S. A., & Bohus, M. (2019). Psychometric validation of the Arabic version of the Borderline Symptom List-23 (BSL-23). Comprehensive Psychiatry, 90, 100–106. https://doi.org/10.1016/j.comppsych.2019.01.006

Bohus, M., Kleindienst, N., Limberger, M. F., Stieglitz, R. D., Domsalla, M., Chapman, A. L., Steil, R., Philipsen, A., & Wolf, M. (2007). The short version of the Borderline Symptom List (BSL-23): Development and initial validation. Psychopathology, 40(6), 453–462. https://doi.org/10.1159/000106468

Bohus, M., Limberger, M. F., Frank, U., Chapman, A. L., Kühler, T., & Stieglitz, R. D. (2009). Psychometric properties of the Borderline Symptom List (BSL-23). Psychopathology, 42(6), 406–412. https://doi.org/10.1159/000236908

Güleç, H., Gültekin, B. K., Akvardar, Y., & Bohus, M. (2014). Reliability and validity of the Turkish version of the Borderline Symptom List-23 (BSL-23). Klinik Psikofarmakoloji Bulteni – Bulletin of Clinical Psychopharmacology, 24(2), 143–151. https://doi.org/10.5455/bcp.20130925021235

Kjaersdam Telléus, G., Fagerlund, B., Jørgensen, M. S., & Bohus, M. (2016). Psychometric properties of the Danish version of the Borderline Symptom List-23. Nordic Journal of Psychiatry, 70(6), 427–432. https://doi.org/10.3109/08039488.2016.1155233

Linehan, M. M. (1993). Cognitive-behavioral treatment of borderline personality disorder. Guilford Press.

Nicastro, R., Prada, P., Kung, K. S., Hasler, R., Dayer, A., Aubry, J. M., & Perroud, N. (2016). Psychometric properties of the French version of the Borderline Symptom List-23. Journal of Personality Disorders, 30(4), 543–554. https://doi.org/10.1521/pedi_2015_29_211

Soler, J., Vega, D., Feliu-Soler, A., Trujols, J., Soto, A., Elices, M., Soriano, J., Álvarez, E., & Pérez, V. (2013). Validation of the Spanish version of the Borderline Symptom List-23 (BSL-23). Revista de Psiquiatría y Salud Mental, 6(4), 156–162. https://doi.org/10.1016/j.rpsm.2013.03.001

Young, J. E., Klosko, J. S., & Weishaar, M. E. (2003). Schema therapy: A practitioner’s guide. Guilford Press.

13. Items of the Scale