May 2010  << Back  

  Can't open the PDF? Click here for help.



Impact on suicidality of the borderline personality traits impulsivity and affective instability

Zoltán Rihmer, MD, PhD, DSc

Department of Clinical and Theoretical Mental Health, Faculty of Medicine, Semmelweis University, Budapest, Hungary

Franco Benazzi, MD, PhD, DTMH†

Hecker Psychiatry Research Center, University of California at San Diego, Collaborating Center at Forli (Italy), Forli, Italy
Department of Psychiatry, National Health Service, Forli, Italy
Department of Psychiatry, University of Szeged, Szeged, Hungary

† Dr. Benazzi died August 22, 2009

BACKGROUND: The aim of this study was to test the impact on suicidality (suicide threats, attempts) of the borderline personality disorder (BPD) traits impulsivity and affective instability in mood disorders.

METHODS: In a general psychiatry private practice (nontertiary care), consecutive remitted, non-substance-abusing outpatients—138 with bipolar II disorder (BP II) and 71 with major depressive disorder (MDD)—self-assessed using the Structured Clinical Interview for DSM-IV Axis II Personality Disorders (SCID-II) Questionnaire.

RESULTS: The frequency (higher in BP II) of suicidality was 14%; impulsivity, 37%; and affective instability, 58%. The suicidality-positive patients (n = 30), when compared with the suicidality-negative patients (n = 179), had more BP II, more impulsivity (odds ratio [OR], 5.5; 95% confidence interval [CI], 2.3 to 13.3), and more affective instability (OR, 2.4; 95% CI, 0.99 to 6.0). Logistic regression of suicidality vs impulsivity and affective instability (controlled for BP II; age; and interactions among BP II, age, impulsivity, and affective instability), showed that impulsivity was a strong independent predictor of suicidality (OR, 4.3; 95% CI, 1.7 to 10.6), and that affective instability was not an independent predictor of suicidality (OR,1.6; 95% CI, 0.6 to 4.1). BP II showed neither confounding nor interactions.

CONCLUSION: Results showed a strong independent impact of impulsivity—but not affective instability—on suicidality in BPD. No confounding by mood and substance disorders supported the BPD nature of these associations.

KEYWORDS: borderline personality disorder, bipolar II disorder, suicide attempts, impulsivity, affective instability



The toll of suicide is high among patients with borderline personality disorder (BPD). Most BPD patients attempt suicide, with nearly 10% eventually committing suicide.1 Identifying the BPD traits that most increase the risk of suicidality (suicide threats, attempts) is a priority in order to better target prevention strategies.

Impulsivity and affective instability are among the most common and persistent traits of BPD, with impulsivity seemingly more episodic and less stable in the long run.2-5 Impulsivity and affective instability—risk factors for suicide in many psychiatric disorders6—appear to be the main traits increasing the risk of suicidality among patients with BPD.1,4,7-17 It is still unclear which of these traits is more strongly related to suicidality in BPD, with some studies reporting affective instability4,9,10,13-15 and others reporting impulsivity1,7,11,17,18 to be the strongest or only predictor of suicidality. Treatment(s) should target BPD traits that most increase the risk of suicidality.

A dimensional model of personality disorders, rather than the current DSM-IV-TR categorical model, has been gaining consensus and research support.3,4,19-27 Studies on BPD have been carried out mainly in tertiary care and BPD care centers. These studies are limited because they assessed the most severe part of the BPD spectrum. general psychiatry outpatient private practice setting could more likely see the less severe part of the BPD spectrum (eg, more compliant individuals with much less or no substance abuse, who present voluntarily for treatment). Studies in this setting could complement previous studies by testing other findings on this neglected population, which is more likely to have BPD traits than full-blown BPD.28-34 Actually, the subsyndromal phenomenology of BPD may be more common than the BPD diagnosis itself.35

Several reviews and papers on the frequency of comorbid mood disorders (especially bipolar disorder [BP]) and BPD have come to opposite conclusions, eg, both high and low comorbidity rates have been report-ed.23,24,28,29,32,33,36-41 A discussion of comorbidity is beyond the scope of this article. However, it is important to note that, among mood disorders, suicidality—a BPD trait—has been reported to occur most often in bipolar II disorder (BP II).42

The goal of this study was to test the impact of impulsivity and affective instability traits on BPD suicidality (suicide threats, attempts) in a sample of mood disorder patients in a nontertiary care setting.


Detailed descriptions of the sample (previously studied for different goals) and the study methods can be found in previously published reports.32,43 Approval was obtained from the local Institutional Review Board, and appropriate consent was obtained from the study participants.

The setting of the study was a general psychiatry outpatient private practice in northern Italy (Emilia-Romagna region), which is the first- or second- (after general practitioners) line of treatment of nonpsychotic and nonsubstance-abusing mood disorder patients instead of tertiary care (eg, a setting for the most common and less severe mental disorders).28-31,33 Most residents of this wealthy region can afford a private psychiatrist (fee-for-service), reducing a possible income bias.


A consecutive sample of 138 BP II and 71 major depressive disorder (MDD) outpatients, previously presented voluntarily for treatment of depression, who are in remission. Remission was measured by a score of >80 for at least 1 month on the Global Assessment of Functioning (GAF) scale from the Structured Clinical Interview for DSM-IV Axis I Disorders (SCID-I). Patients were rediagnosed during follow-up visits in 2005 by a senior clinical and mood disorder research psychiatrist (F.B.). The frequency of BP II was 66.0% (138/209): mean (SD) age, 39.0 (9.8) years; females, 67.3%. The frequency of MDD was 33.9% (71/209): mean (SD) (age, 39.2 (10.6) years; females, 77.1%. BP II was overrepresented because it requires more follow-up visits compared with MDD, due to its higher instability and more complex treatments. However, a BP II:MDD ratio of around 1 recently has been reported in depressed outpatients.44 Interviewing remitted patients should have reduced recall bias and biases related to the prevalent mood of an episode, which impact assessment of personality.45,46 Substance-abusing patients were not included (such patients are uncommon in this study setting, although common in tertiary care and BPD care centers,28,29,33,34) in order to study the impact of predictors on BPD suicidality in a sample not confounded by a high-risk factor for suicidality such as substance abuse.42,47

Patients were assessed by the following instruments: (1) the Structured Clinical Interview for DSM-IV Axis I Disorders–Clinician Version48 (SCID-CV; reported inter-rater reliability, κ = 0.70 to 1.0), as modified and validated by Benazzi and Akiskal,43,44 to improve probing for BP II (the interviewers showed an inter-rater reliability of κ = 0.73 for diagnosing BP II)49; the SCID-CV is partly semistructured (eg, the wording of the questions can be changed to improve and check understanding by the interviewee if necessary) and is based on clinical evaluation; it has been shown that semistructured interviewing by trained clinicians outperforms structured interviewing in correctly diagnosing mood disorders50-53; (2) the Global Assessment of Functioning scale (GAF), from the SCID-CV, to assess symptoms and functional remission; (3) the self-assessed Structured Clinical Interview for DSM-IV Axis II Personality Disorders (II) Questionnaire, relative to BPD54; questions address “the kind of person you are” and “how you usually are;” the trait impulsivity is defined by observable behaviors, eg, have “often done things impulsively,” such as “buying things you couldn’t afford, having sex with people you hardly know, or unsafe sex, drinking too much or taking drugs,” “driving recklessly,” and “uncontrollable eating;” the trait affective instability is defined by “a lot of sudden mood changes,” such as “marked reactivity of mood, dysphoria, irritability, anxiety;” the trait suicidality is defined by the question “Have you tried to hurt or kill yourself or threatened to do so, have you ever cut, burned, or scratched yourself on purpose?” For the present study, suicidality was defined by suicide threats, suicide attempts, and minor aggressive acts toward self, following the SCID-II. Although BPD self-reports can yield a higher endorsement of BPD traits compared with diagnostic interviews, the common assumption that diagnostic interviews are more valid than self-reports was found to be only partly supported, as the disagreement between the 2 methods was not high.55

Patients were interviewed by the following methods: Remitted patients were interviewed by the SCID-CV for rediagnosing BP II and MDD (blind to patients), who soon later self-assessed using the SCID-II Questionnaire in the waiting room. Clinically significant distress or impairment of functioning (the features distinguishing personality disorder and personality traits, according to DSM-IV-TR) are not included in the self-assessed SCID-II, as clinical evaluation would be required. Therefore, although a diagnosis of BPD could not be made, BPD traits could be assessed (eg, all BPD traits, but not the degree of impairment); the clinical significance (eg, levels of severity) of these traits cannot be assessed by the yes/no questions of this questionnaire.

Statistical analyses. Multivariate analyses were used to assess the impact of the BPD traits impulsivity and affective instability on suicidality.

Logistic regression was used for testing associations and interactions, and to control for confounding effects. Principal-component factor analysis was used to assess correlations among BPD traits (varimax rotation, eigenvalue >1, item loading >0.50). Multivariate analyses were carried out on the entire sample, as BPD traits were common in both BP II and MDD. Stata Statistical Software, Release 9.2, was used (StataCorp, College Station, TX, USA; 2006). P values were 2-tailed, and the alpha level was set at 0.05. Trends in ORs were also assessed by inspecting the 95% CIs, following Rothman et al.56,57


In the entire sample, the frequency of BPD traits ranged from 14.3% to 65.8%: suicidality, 14.3%; impulsivity, 36.7%; and affective instability, 57.7%. The frequency of BPD traits in MDD vs BP II is reported in TABLE 1. Suicidality, impulsivity, and affective instability were more common in BP II (at a statistically significant level, or by inspecting the trends in the 95% CIs), and were not uncommon in MDD.

The sample was then split into a suicidality-positive group and suicidality-negative group (TABLE 2). The suicidality-positive group had lower age, more BP II (OR, 2.62; 95% CI, 0.95 to 7.22), and a higher loading of BPD traits, including impulsivity (OR, 5.50; 95% CI, 2.28 to 13.25) and affective instability (OR, 2.44; 95% CI, 0.99 to 6.05).

Because BP II and age are risk factors for suicidality,42 the possible confounding effect (eg, a change of ≥20% in the OR)56 of these covariates was tested. Interactions were tested, and the significant ones should have been included in the final model.56

By multivariable logistic regression of suicidality (dependent variable) vs all BPD traits (TABLE 3), impulsivity (among the BPD traits related to the study goal, eg, suicidality, impulsivity, and affective instability) was the only independent and strong (OR, 3.17; 95% CI, 1.14 to 8.80) predictor of suicidality. BP II and age did not show confounding effects.

Multivariable logistic regression of suicidality (dependent variable) vs impulsivity and affective instability alone showed a higher OR for impulsivity (4.26 vs 3.17) compared with the previous analysis, which included all BPD traits, suggesting confounding effects by some BPD items. BP II and age also did not show confounding effects in this analysis. Interactions were tested. No statistically significant interaction was found among all possible pair combinations of impulsivity, affective instability, BP II, and age (by including interactions one by one in the above model, the P value range was 0.217 to 0.835). The interaction between impulsivity and BP II could not be calculated because of collinearity.

Finally, in order to replicate the above findings using different statistics, factor analysis of all BPD traits was conducted (TABLE 4). Two factors were found: a first factor loading on affective instability and not correlated with suicidality and impulsivity, and a second factor loading on impulsivity and suicidality and not correlated with affective instability. These findings support the regression analyses carried out before.


Borderline personality disorder traits in major depressive disorder and bipolar II disorder*

BPD traits, % MDD BP II OR (95% CI)
Frantic efforts to avoid abandonment 65.7 65.9 1.00 (0.54 to 1.85)
Unstable and intense interpersonal relationships 34.2 47.4 1.73 (0.95 to 3.14)
Unstable self-image 27.5 44.4 2.10 (1.12 to 3.94)a
Impulsivity 27.9 41.1 1.80 (0.96 to 3.39)
Suicidality 7.1 16.7 2.62 (0.95 to 7.22)
Affective instability 47.8 62.7 1.83 (1.02 to 3.30)a
Chronic emptiness 46.3 59.8 1.72 (0.96 to 3.08)
Inappropriate intense anger 37.1 39.4 1.10 (0.60 to 1.99)
Paranoid ideation 50.0 49.6 0.98 (0.55 to 1.75)
BPD: borderline personality disorder; BP II: bipolar II disorder; CI: confidence interval; MDD: major depressive disorder; OR: odds ratio.
*By univariate logistic regression.
aP < .05.


Comparison of suicidality-positive patients and suicidality-negative patients*

Variables S+ (n = 30) S– (n = 179) OR (95% CI)
Age, y, mean (SD) 36.6 (10.8) 39.5 (9.9) 0.67 (0.44 to 1.03)
Women, % 76.6 69.6 1.30 (0.52 to 3.25)
BP II, % 80.0 63.6 2.62 (0.95 to 7.22)
MDD, % 20.0 36.3 0.38 (0.13 to 1.05)
BPD traits, %
  Frantic efforts to avoid abandonment 77.7 64.0 1.96 (0.75 to 5.11)
  Unstable interpersonal relationships 60.7 40.2 2.29 (1.01 to 5.18)a
  Unstable self-image 71.4 33.5 4.95 (2.06 to 11.92)b
  Impulsivity 71.4 31.2 5.50 (2.28 to 13.25)b
  Affective instability 75.0 55.0 2.44 (0.99 to 6.05)
  Chronic emptiness 89.2 50.0 8.33 (2.42 to 28.59)b
  Inappropriate intense anger 64.2 34.6 3.39 (1.47 to 7.80)b
  Paranoid ideation 78.5 45.1 4.44 (1.71 to 11.49)b
No. of BPD traits, mean (SD) 6.8 (1.4) 3.5 (2.1) 2.65 (1.86 to 3.78)b
BPD: borderline personality disorder; BP II: bipolar II disorder; CI: confidence interval; MDD: major depressive disorder; OR: odds ratio; S–: suicidality-negative patients; S+: suicidality-positive patients.
*By univariate logistic regression.
aP < .05
bP < .01


Borderline personality disorder trait suicidality (dependent variable) vs BPD traits, controlled for bipolar-II disorder and age

Variables OR (95% CI)
All BPD traits
  Frantic efforts to avoid abandonment 1.51 (0.49 to 4.62)
  Unstable and intense interpersonal relationships 1.25 (0.45 to 3.44)
  Unstable self-image 1.81 (0.63 to 5.13)
  Impulsivity 3.17 (1.14 to 8.80)a
  Affective instability 0.43 (0.12 to 1.53)
  Chronic emptiness 5.20 (1.19 to 22.61)a
  Inappropriate intense anger 2.20 (0.79 to 6.11)
  Paranoid ideation 2.80 (0.93 to 8.40)
Selected BPD traits
  Impulsivity 4.26 (1.71 to 10.64)b
  Affective instability 1.56 (0.59 to 4.11)
BPD: borderline personality disorder; CI: confidence interval; OR: odds ratio.
*By multivariable logistic regression on the entire sample (n = 209).
aP < .05
bP < .01


Factor analysis of borderline personality disorder traits*

BPD traits F1 F2
Frantic efforts to avoid abandonment –0.15 0.68
Unstable interpersonal relationships 0.66 –0.08
Unstable self-image 0.54 0.39
Impulsivity 0.28 0.52
Suicidality 0.24 0.60
Affective instability 0.75 0.06
Chronic emptiness 0.67 0.23
Inappropriate intense anger 0.58 0.04
Paranoid ideation 0.29 0.42
Eigenvalue 2.76 1.11
Variance 0.30 0.12
BPD: borderline personality disorder; F: factor.
*Varimax rotation, eigenvalue >1, item loading >0.50, on the entire sample (n = 209).


Our study findings complement earlier research carried out in different settings: (1) the setting was not a tertiary care or BPD care center; and (2) substance-abuse comorbidity was not present. The absence of substance-abuse comorbidity (common in other settings)28,33,47 suggests that the present study findings cannot be related to this comorbidity (substance abuse is a strong risk factor for suicidality).42,58

The frequency of BPD traits in the study sample was high. It was higher in BP II vs MDD: eg, suicidality was 16.7% vs 7.1% (OR, 2.62; 95% CI, 0.95 to 7.22); impulsivity was 41.1% vs 27.9% (OR, 1.80; 95% CI, 0.96 to 3.39); and affective instability was 62.7% vs 47.8% (OR, 1.83; 95% CI, 1.02 to 3.30). In another study,59 by using a different BPD self-assessment instrument, an even higher frequency of BPD traits was found in a BP II and MDD sample (higher in BP II but common also in MDD, as in the present study). It has been found that the subsyndromal phenomenology of BPD (eg, BPD traits) is more common than the BPD diagnosis itself.29,35 The high frequency of BPD traits in the study sample may be related to the self-assessed SCID-II Questionnaire, which does not assess the degree of impairment of each trait (as there is no clinical evaluation). The self-assessed SCID-II Questionnaire could have inflated the frequency of BPD traits by including an unknown proportion of traits that were marginally or not clinically significant.55,59 Self-report on personality can yield higher endorsement of BPD traits compared with diagnostic interviews, but the degree of disagreement was not found to be high between the 2 methods.55 According to a dimensional model of BPD, BPD traits should lie along a continuum of severity in the community, ranging from traits that are not clinically significant to traits of increasingly higher clinical severity.3,4,19-21,25-27 It is worth noting that BPD traits were also common in the MDD study sample, supporting the dimensional model of BPD.

Our study’s goal was to test the impact on BPD suicidality of the BPD traits impulsivity and affective instability. Univariate analysis, comparing the suicidality-positive and the suicidality-negative patients, showed that suicidality was strongly associated with impulsivity (OR, 5.50; 95% CI, 2.28 to 13.25) and, to a lower degree, with affective instability (OR, 2.44; 95% CI, 0.99 to 6.05). Among mood disorders, BP II has the highest risk of suicidality,42 so the associations found could have been related to the large proportion of BP II present in the study sample. To control for confounding and interactions of BP II, other BPD traits, and age, multivariate analyses were carried out. Multivariable logistic regression of suicidality vs all BPD traits (TABLE 3) showed that impulsivity was a strong (OR, 3.17; 95% CI, 1.14 to 8.80) independent predictor of suicidality, whereas affective instability was not associated with suicidality. By focusing only on the relationships among BPD suicidality, impulsivity, and affective instability, and including all possible interactions among impulsivity, affective instability, BP II, and age, impulsivity showed a stronger OR (4.26; 95% CI. 1.71 to 10.64), compared with the previous analysis, and was the only independent predictor of suicidality. Affective instability did not show any significant effect on suicidality. Controlling for BP II did not show confounding effects. No significant interaction was found among all possible combinations of impulsivity, affective instability, BP II, and age. Disentangling BP II and BPD has been shown to be possible by semistructured clinical evaluation (by the SCID-CV in the present study) and by structured instruments (the SCID-II in the present study).29,59,60

These findings were replicated by factor analysis, which showed no correlation of affective instability with suicidality and impulsivity, but did show a correlation between suicidality and impulsivity. A priority of impulsivity over affective instability as a predictor of suicidality was shown by the present study, supporting some previous studies1,7,11,17,18 carried out by different methods and in populations with more severe BPD.

The finding that affective instability was not an independent predictor of suicidality in the present study does not dismiss affective instability as a risk factor for suicidality in BPD. In the univariate analysis, affective instability was 2.4 times more likely in the suicidality-positive group. Affective instability, a more persistent BPD trait than impulsivity,3,19,20,21,24,27,41 might be a baseline trait that episodically triggers impulsivity and suicidality by its marked reactivity of mood.28,33,61

Results could be questioned by mood (bipolar) disorder studies, which have reported a high frequency of the traits impulsivity and affective instability, and the negative impacts of these traits on suicidality.28,30-34,40,62-69 The traits impulsivity and affective instability could not be distinguished in BP II and BPD by the SCID-II (if such a clinically significant distinction were possible). However, it has been found that such a problem can be greatly reduced by clinical diagnostic interviewing for mood disorders, and by using the SCID-II or similar instruments for BPD.29,59,60 Domain-focused instruments such as the Barratt Impulsiveness Scale and the Affective Lability Scale could have found some distinguishing trait features70,71 that were undetected by the SCID-II broad indicators of these domains. An advantage of the SCID-II assessment of impulsivity is that it is based on observable behaviors. Our multivariate analyses disentangled the BPD traits impulsivity and affective instability from BP II (the mood disorder showing the highest risk of suicidality),42 by showing that it was not a confounding factor and that it did not significantly interact with impulsivity and affective instability. These findings may support the BPD nature of the traits assessed by the SCID-II in the study sample. Furthermore, the broad indicators of the traits impulsivity and affective instability assessed by the SCID-II replicated the basic findings of studies that used domain-focused instruments in bipolar and BPD samples.34,62,67,70 Clinically significant features distinguishing the traits impulsivity and affective instability of BPD and BP II seem unlikely.36,38,39,72 The few reported different features38,39 of trait impulsivity and affective instability between BPD and BP II have not yet shown an impact on the management of BPD.36

The present study results strengthen the priority of impulsivity over affective instability as a predictor of suicidality in BPD, which was previously found by some studies in more severe BPD populations,1,7,11,17 by finding the same associations in a less severe population in whom BPD traits are much more common than the BPD diagnosis itself.29 The association between the BPD traits suicidality and impulsivity thus seems independent of the severity of BPD traits, according to a dimensional model of BPD.25

For the pharmacologic prevention of the suicidality of BPD, treatments shown effective for impulsivity of BPD might also be useful to prevent suicidality (this has yet to be demonstated, however), on the basis of the present study findings on the relationship between impulsivity and suicidality. Reviews, meta-analyses, and reports have found that antipsychotic agents seem more effective on the impulsivity of BPD than mood stabilizing agents and antidepressants.73-78 Psychopharmacologic agents often have been shown to be effective only on some domains of BPD, eg, more on impulsivity than on affective instability and vice versa. However, some second-generation antipsychotics seem to be “broad spectrum” (eg, aripiprazole, quetiapine, and olanzapine).76,78,79 Psychological and psychosocial treatments have shown positive effects on BPD and on its suicidality, such as dialectical behavior therapy, mentalization-based programs, and cognitive-behavioral therapy.80-83 A combination of pharmacologic and psychological/psychosocial treatments seems logical and necessary.6


Some limitations of the self-assessment questionnaire for BPD used in this study (eg, the SCID-II Questionnaire) are also reported in the discussion. Self-assessment questionnaires are increasingly used in personality disorder research.61,68,84,85 Interview methods are widely regarded as the standard for the diagnosis of BPD, whereas self-report methods are considered a time-efficient alternative.55 In the present study setting—a busy private practice in which research and clinical work are carried out side by side—time is a strong limiting factor for research. Although BPD self-report can yield higher endorsement of traits compared with diagnostic interviews, the common assumption that diagnostic interviews are more valid than self-reports was only partly supported, as the disagreement between the 2 methods was not high.55

Given these study limitations, the results should be seen as preliminary; also, the clinician version of the SCID-II should be administered by an independent interviewer, and domain-focused instruments for impulsivity and affective instability should also be used. Ideally, a sample of BPD without any Axis I disorder comorbidity should be studied, but BPD and Axis I (especially mood) disorder comorbidity is high in clinical practice.23

Interviews were conducted by the same interviewer. However, the interviewer has been studying BP II for many years, his inter-rater reliability for BP II diagnosis is acceptable,37 and validated instruments were used.43,44,48,54 Diagnosing BP II followed the current best practice, based on semistructured interviews by clinicians trained in the diagnosis and treatment of bipolar disorder.44 Bipolar I disorder patients were not included, but BP II has been found to have similar interepisode traits.28,30,33,34,62,68,69


Results showed a strong independent impact of impulsivity—but not affective instability—on suicidality in BPD. No confounding by mood and substance disorders supported the BPD nature of these associations.

DISCLOSURES: Dr. Rihmer is a consultant to AstraZeneca, Bristol-Myers Squibb, Eli Lilly and Company, Pfizer Inc., Roche, Servier, sanofi-aventis, and Solvay Pharmaceuticals. He receives travel support from AstraZeneca, Lundbeck, Eli Lilly and Company, sanofi-aventis, and Servier. Dr. Benazzi died August 22, 2009.


  1. Black DW, Blum N, Pfohl B, et al. Suicidal behavior in borderline personality disorder: prevalence, risk factors, prediction, and prevention. J Pers Disord. 2004;18:226–239.
  2. Grilo CM, Sanislow CA, Skodol AE, et al. Longitudinal diagnostic efficiency of DSM-IV criteria for borderline personality disorder: a 2-year prospective study. Can J Psychiatry. 2007;52:357–362.
  3. McGlashan TH, Grilo CM, Sanislow CA, et al. Two-year prevalence and stability of individual DSM-IV criteria for schizotypal, borderline, avoidant, and obsessive-compulsive personality disorders: toward a hybrid model of axis II disorders. Am J Psychiatry. 2005;162:883–889.
  4. Yen S, Shea MT, Sanislow CA, et al. Borderline personality disorder criteria associated with prospectively observed suicidal behavior. Am J Psychiatry 2004;161:1296–1298.
  5. Tragesser SL, Solhan M, Schwartz-Mette R, et al. The role of affective instability and impulsivity in predicting future BPD features. J Pers Disord. 2007;21:603–614.
  6. Swartz HA, Pilkonis PA, Frank E, et al. Acute treatment outcomes in patients with bipolar I disorder and co-morbid borderline personality disorder receiving medication and psychotherapy. Bipolar Disord. 2005;7:192–197.
  7. Brodsky BS, Malone KM, Ellis SP, et al. Characteristics of borderline personality disorder associated with suicidal behavior. Am J Psychiatry. 1997;154:1715–1719.
  8. Hollander E, Swann AC, Coccaro EF, et al. Impact of trait impulsivity and state aggression on divalproex versus placebo response in borderline personality disorder. Am J Psychiatry. 2005;162:621–624.
  9. Koenigsberg HW, Harvey PD, Mitropoulou V, et al. Are the interpersonal and identity disturbances in the borderline personality disorder criteria linked to the traits of affective instability and impulsivity? J Pers Disord. 2001;15:358–370.
  10. Koenigsberg HW, Harvey PD, Mitropoulou V, et al. Characterizing affective instability in borderline personality disorder. Am J Psychiatry. 2002;59:784–788.
  11. Links PS, Heslegrave R, van Reekum R. Impulsivity: core aspect of borderline personality disorder. J Pers Disord. 1999;13:1–9.
  12. Links PS, Heslegrave RJ. Prospective studies of outcome. Understanding mechanisms of change in patients with borderline personality disorder. Psychiatr Clin North Am. 2000;23:137–150.
  13. Links PS, Boggild A, Sarin N. Modeling the relationship between affective lability, impulsivity, and suicidal behavior in patients with borderline personality disorder. J Psychiatr Pract. 2000;6:247–255.
  14. Links PS, Eynan R, Heisel MJ, et al. Affective instability and suicidal ideation and behavior in patients with borderline personality disorder. J Pers Disord. 2007;21:72–86.
  15. Links PS, Eynan R, Heisel MJ, et al. Elements of affective instability associated with suicidal behaviour in patients with borderline personality disorder. Can J Psychiatry. 2008;53:112–116.
  16. McGirr A, Seguin M, Renaud J, et al. Gender and risk factors for suicide: evidence for heterogeneity in predisposing mechanisms in a psychological autopsy study. J Clin Psychiatry. 2006;67:1612–1617.
  17. McGirr A, Paris J, Lesage A, et al. Risk factors for suicide completion in borderline personality disorder: a case-control study of cluster B comorbidity and impulsive aggression. J Clin Psychiatry. 2007;68:721–729.
  18. Christodoulou T, Lewis M, Ploubidis GB, et al. The relationship of impulsivity to response inhibition and decision-making in remitted patients with bipolar disorder. Eur Psychiatry. 2006;21:270–273.
  19. First MB, Bell CC, Cuthbert B, et al. Personality disorders and relational disorders. In: Kupfer DJ, First MB, Regier DA, eds. A research agenda for DSM-V. Arlington, VA: American Psychiatric Publishing, Inc.; 2002:123–199.
  20. Jablensky A. The classification of personality disorders: critical review and need for rethinking. Psychopathology. 2002;35:112–116.
  21. Lynam DR, Widiger TA. Using a general model of personality to identify the basic elements of psychopathy. J Pers Disord. 2007;21:160–178.
  22. Skodol AE, Siever LJ, Livesley WJ, et al. The borderline diagnosis II: biology, genetics, and clinical course. Biol Psychiatry. 2002;51:951–963.
  23. Skodol AE, Gunderson JG, Pfohl B, et al. The borderline diagnosis I: psychopathology, comorbidity, and personality structure. Biol Psychiatry. 2002;51:936–950.
  24. Skodol AE, Gunderson JG, Shea MT, et al. The Collaborative Longitudinal Personality Disorders Study (CLPS): overview and implications. J Pers Disord. 2005;19:487–504.
  25. Skodol AE, Oldham JM, Bender DS, et al. Dimensional representations of DSM-IV personality disorders: relationships to functional impairment. Am J Psychiatry. 2005;162:1919–1925.
  26. Widiger TA. Dimensional models of personality disorder. World Psychiatry. 2007;6:79–83.
  27. Walters GD, Gray NS, Jackson RL, et al. A taxometric analysis of the psychopathy checklist: screening version (PCL:SV): further evidence of dimensionality. Psychol Assess. 2007;19:330–339.
  28. Akiskal HS, Akiskal KK, Perugi G, et al. Bipolar II and anxious reactive “comorbidity”: toward better phenotypic characterization suitable for genotyping. J Affect Disord. 2006;96:239–247.
  29. Benazzi F. Borderline personality disorder and bipolar II disorder in private practice depressed outpatients. Compr Psychiatry. 2000;41:106–110.
  30. Benazzi F. Validating angst’s “ups & downs” personality trait as a new marker of bipolar II disorder. Eur Arch Psychiatry Clin Neurosci. 2004;254:48–54.
  31. Benazzi F. Does temperamental instability support a continuity between bipolar II disorder and major depressive disorder? Eur Psychiatry. 2006;21:274–279.
  32. Benazzi F. A relationship between bipolar II disorder and borderline personality disorder? Prog Neuropsychopharmacol Biol Psychiatry. 2008;15:1022–1019.
  33. Perugi G, Toni C, Travierso MC, et al. The role of cyclothymia in atypical depression: toward a data-based reconceptualization of the borderline-bipolar II connection. J Affect Disord. 2003;73:87–98.
  34. Swann AC, Dougherty DM, Pazzaglia PJ, et al. Increased impulsivity associated with severity of suicide attempt history in patients with bipolar disorder. Am J Psychiatry. 2005;162:1680–1687.
  35. Zanarini MC, Frankenburg FR, Yong L, et al. Borderline psychopathology in the first-degree relatives of borderline and axis II comparison probands. J Pers Disord. 2004;18:439–447.
  36. Akiskal HS. Demystifying borderline personality: critique of the concept and unorthodox reflections on its natural kinship with the bipolar spectrum. Acta Psychiatr Scand. 2004;110:401–407.
  37. Paris J, Gunderson J, Weinberg I. The interface between borderline personality disorder and bipolar spectrum disorders. Compr Psychiatry. 2007;48:145–154.
  38. Gunderson J, Zanarini MC, Kisiel CL. Borderline personality disorder. In: Widiger TA, Frances AJ, Pincus HL, et al, eds. DSM-IV sourcebook, vol 2. Arlington, VA: American Psychiatric Publishing, Inc.: 1996:717–731.
  39. Gunderson JG, Weinberg I, Daversa MT, et al. Descriptive and longitudinal observations on the relationship of borderline personality disorder and bipolar disorder. Am J Psychiatry. 2006;163:1173–1178.
  40. Smith DJ, Muir WJ, Blackwood DH. Is borderline personality disorder part of the bipolar spectrum? Harv Rev Psychiatry. 2004;12:133–139.
  41. Zanarini MC, Frankenburg FR, Reich DB, et al. The subsyndromal phenomenology of borderline personality disorder: a 10-year follow-up study. Am J Psychiatry. 2007;164:929–935.
  42. Rihmer Z. Suicide risk in mood disorders. Curr Opin Psychiatry. 2007;20:17–22.
  43. Akiskal HS, Benazzi F. Optimizing the detection of bipolar II disorder in outpatient private practice: toward a systematization of clinical diagnostic wisdom. J Clin Psychiatry. 2005;66:914–921.
  44. Benazzi F. Bipolar disorder—focus on bipolar II disorder and mixed depression. Lancet. 2007;369:935–945.
  45.  Diagnostic and statistical manual of mental disorders, 4th ed, text rev. Washington, DC: American Psychiatric Association; 2000.
  46. Peselow ED, Sanfilipo MP, Fieve RR. Relationship between hypomania and personality disorders before and after successful treatment. Am J Psychiatry. 1995;152:232–238.
  47. Links PS, Heslegrave RJ, Mitton JE, et al. Borderline personality disorder and substance abuse: consequences of comorbidity. Can J Psychiatry. 1995;40:9–14.
  48. First MB, Spitzer RL, Gibbon M, et al. Structured clinical interview for DSM-IV axis I disorders-clinician version (SCID-CV). Washington, DC: American Psychiatric Press: 1997.
  49. Benazzi F. Depression with racing thoughts. Psychiatry Res. 2003;120:273–282.
  50. Bader CD, Dunner DL. Bipolar disorder not otherwise specified in relation to the bipolar spectrum. Bipolar Disord. 2007;9:860–867.
  51. Benazzi F. Diagnosis of bipolar II disorder: a comparison of structured versus semistructured interviews. Prog Neuropsychopharmacol Biol Psychiatry. 2003;27:985–991.
  52. Kessler RC, Akiskal HS, Angst J, et al. Validity of the assessment of bipolar spectrum disorders in the WHO CIDI 3.0. J Affect Disord. 2006;69:259–269.
  53. Simpson SG, McMahon FJ, McInnis MG, et al. Diagnostic reliability of bipolar II diagnosis. Arch Gen Psychiatry. 2002;59:736–740.
  54. First MB, Gibbon M, Spitzer RL, et al. SCID-II Personality questionnaire. Washington, DC: American Psychiatric Press; 1997.
  55. Hopwood CJ, Morey LC, Edelen MO, et al. A comparison of interview and self-report methods for the assessment of borderline personality disorder criteria. Psychol Assess. 2008;20:81–85.
  56. Hosmer DW, Lemeshow S. Applied logistic regression. New York, NY: John Wiley & Sons, Inc.; 2000.
  57. Rothman KJ, Greenland S. Modern epidemiology, 2nd ed. Philadelphia, PA: Lippincott Williams & Wilkins; 1998.
  58. Wilson ST, Fertuck EA, Kwitel A, et al. Impulsivity, suicidality and alcohol use disorders in adolescents and young adults with borderline personality disorder. Int J Adolesc Med Health. 2006;18:189–196.
  59. Smith DJ, Muir WJ, Blackwood DH. Borderline personality disorder characteristics in young adults with recurrent mood disorders: a comparison of bipolar and unipolar depression. J Affect Disord. 2005;87:17–23.
  60. Nakagawa A, Grunebaum MF, Sullivan GM, et al. Comorbid anxiety in bipolar disorder: does it have an independent effect on suicidality? Bipolar Disord. 2008;10:530–538.
  61. Akiskal HS, Akiskal KK, Haykal RF, et al. TEMPS-A: progress towards validation of a self-rated clinical of the temperament evaluation of the Memphis, Pisa, Paris and San Diego autoquestionnaire. J Affect Disord. 2005;85:3–16.
  62. Akiskal HS, Hantouche EG, Allilaire JF. Bipolar II with and without cyclothymic temperament: “dark” and “sunny” expressions of soft bipolarity. J Affect Disord. 2003;73:49–57.
  63. Angst J. The bipolar spectrum. Br J Psychiatry. 2007;190:189–191.
  64. Isometsä ET, Henriksson MM, Heikkinen ME, et al. Suicide among subjects with personality disorders. Am J Psychiatry. 1996;153:667–673.
  65. Kochman FJ, Hantouche EG, Ferrai P, et al. Cyclothymic temperament as a prospective predictor of bipolarity and suicidality in children and adolescents with major depressive disorder. J Affect Disord. 2005;85:181–189.
  66. Maser JD, Aksikal HS, Schettler P, et al. Can temperament identify affectively ill patients who engage in lethal or non-lethal suicidal behavior? A 14-year prospective study. Suicide Life Threat Behav. 2002;32:10–32.
  67. Najt P, Perez J, Sanches M, et al. Impulsivity and bipolar disorder. Eur Neuropsychopharmacol. 2007;17:313–320.
  68. Swann AC, Gerard Moeller F, Steinberg JL, et al. Manic symptoms and impulsivity during bipolar depressive episodes. Bipolar Disord. 2007;9:206–212.
  69. Swann AC, Steinberg JL Lijffijt M, et al. Impulsivity: differential relationship to depression and mania in bipolar disorder. J Affect Disord. 2008;106:241–248.
  70. Henry C, Mitropoulou V, New AS, et al. Affective instability and impulsivity in borderline personality and bipolar II disorders: similarities and differences. J Psychiatr Res. 2001;35:307–312.
  71. Wilson ST, Stanley B, Oquendo MA, et al. Comparing impulsiveness, hostility, and depression in borderline personality disorder and bipolar II disorder. J Clin Psychiatry. 2007;68:1533–1539.
  72. Gunderson JG. Disturbed relationships as a phenotype for borderline personality disorder. Am J Psychiatry. 2007;164:1637–1640.
  73. Abraham PF, Calabrese JR. Evidenced-based pharmacologic treatment of borderline personality disorder: a shift from SSRIs to anticonvulsants and atypical antipsychotics? J Affect Disord. 2008 Mar 4. [Epub ahead of print]
  74. Cardish RJ. Psychopharmacologic management of suicidality in personality disorders. Can J Psychiatry. 2007;52(6 suppl 1):115S–127S.
  75. Grootens KP, Verkes RJ. Emerging evidence for the use of atypical antipsychotics in borderline personality disorder. Pharmacopsychiatry. 2005;38:20–23.
  76. Nickel MK, Muehlbacher M, Nickel C, et al. Aripiprazole in the treatment of patients with borderline personality disorder: a double-blind, placebo-controlled study. Am J Psychiatry. 2006;163:833–838.
  77. Nosè M, Cipriani A, Biancosino B, et al. Efficacy of pharmacotherapy against core traits of borderline personality disorder: meta-analysis of randomized controlled trials. Int Clin Psychopharmacol. 2006;21:345–353.
  78. Van den Eynde F, Senturk V, Naudts K, et al. Efficacy of quetiapine for impulsivity and affective symptoms in borderline personality disorder. J Clin Psychopharmacol. 2008;28:147–155.
  79. Zanarini MC, Frankenburg FR, Parachini EA. A preliminary, randomized trial of fluoxetine, olanzapine, and the olanzapine-fluoxetine combination in women with borderline personality disorder. J Clin Psychiatry. 2004;65:903–907.
  80. Bateman A, Fonagy P. 8-year follow-up of patients treated for borderline personality disorder: mentalization-based treatment versus treatment as usual. J Psychiatry. 2008;165:631–638.
  81. Clarkin JF, Levy KN, Lenzenweger MF, et al. Evaluating three treatments for borderline personality disorder: a multiwave study. Am J Psychiatry. 2007;164:922–928.
  82. Davidson K, Tyrer P, Gumley A, et al. A randomized controlled trial of cognitive behavior therapy for borderline personality disorder: rationale for trial, method, and description of sample. J Pers Disord. 2006;20:431–449.
  83. Linehan MM, McDavid JD, Brown MZ, et al. Olanzapine plus dialectical behavior therapy for women with high irritability who meet criteria for borderline personality disorder: a double-blind, placebo-controlled pilot study. J Clin Psychiatry. 2008;69:999–1005.
  84. Kwapil TR, Miller MB, Zinser MC, et al. A longitudinal study of high scorers on the hypomanic personality scale. J Abnorm Psychol. 2000;109:222–226.
  85. Peluso MA, Hatch JP, Glahn DC, et al. Trait impulsivity in patients with mood disorders. J Affect Disord. 2007;100:227–231.

CORRESPONDENCE: Zoltán Rihmer, MD, PhD, DSc, Kútvölgyi út 4, 1125 Budapest, Hungary E-MAIL: