Journal of Affective Disorders 170 (2015) 39–45

Contents lists available at ScienceDirect

Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad

Research report

Significance of borderline personality-spectrum symptoms among adolescents with bipolar disorder Trehani M. Fonseka, Brenda Swampillai, Vanessa Timmins, Antonette Scavone, Rachel Mitchell, Katelyn A. Collinger, Benjamin I. Goldstein n Department of Psychiatry, Sunnybrook Health Sciences Centre, University of Toronto, Toronto, ON, Canada

art ic l e i nf o

a b s t r a c t

Article history: Received 8 July 2014 Received in revised form 27 July 2014 Accepted 12 August 2014 Available online 3 September 2014

Background: Little is known regarding correlates of borderline personality-spectrum symptoms (BPSS) among adolescents with bipolar disorder (BP). Methods: Participants were 90 adolescents, 13–19 years of age, who fulfilled DSM-IV-TR criteria for BP using semi-structured diagnostic interviews. BPSS status was ascertained using the Life Problems Inventory which assessed identity confusion, interpersonal problems, impulsivity, and emotional lability. Analyses compared adolescents with “high” versus “low” BPSS based on a median split. Results: Participants with high, relative to low, BPSS were younger, and had greater current and past depressive episode severity, greater current hypo/manic episode severity, younger age of depression onset, and reduced global functioning. High BPSS participants were more likely to have BP-II, and had higher rates of social phobia, generalized anxiety disorder, conduct disorder, oppositional defiant disorder, homicidal ideation, assault of others, non-suicidal self-injury, suicidal ideation, and physical abuse. Despite greater illness burden, high BPSS participants reported lower rates of lithium use. The most robust independent predictors of high BPSS, identified in multivariate analyses, included lifetime social phobia, non-suicidal self-injury, reduced global functioning, and conduct and/or oppositional defiant disorder. Limitations: The study design is cross-sectional and cannot determine causality. Conclusions: High BPSS were associated with greater mood symptom burden and functional impairment. Presence of high BPSS among BP adolescents may suggest the need to modify clinical monitoring and treatment practices. Future prospective studies are needed to examine the direction of observed associations, the effect of treatment on BPSS, and the effect of BPSS as a moderator or predictor of treatment response. & 2014 Elsevier B.V. All rights reserved.

Keywords: Bipolar Borderline Personality disorder Adolescent

1. Introduction The association between borderline personality disorder (BPD) and bipolar disorder (BP), particularly BP-II, has been well documented (Benazzi, 2000, 2006; Zimmerman and Morgan, 2013). BPD is conceptualized as a chronic and persistent personality disorder whereas BP is conceptualized as an episodic mood disorder (American Psychiatric Association, 2013). However, there is increasing evidence that symptoms of BPD wax and wane and that there are often substantial inter-episode symptoms in BP (Akiskal et al., 1989; Morriss, 2002; Zanarini et al., 2005). Some studies argue that BPD is best reframed as part of the BP spectrum due to symptomatic and familial genetic overlap between these diagnostic phenotypes. n Correspondence to: Centre for Youth Bipolar Disorder, Sunnybrook Health Sciences Centre, 2075 Bayview Avenue, Room FG53, Toronto, ON, M4N 3M5, Canada. Tel.: þ1 416 480 6100x5328; fax: þ1 416 480 6878. E-mail address: [email protected] (B.I. Goldstein).

http://dx.doi.org/10.1016/j.jad.2014.08.046 0165-0327/& 2014 Elsevier B.V. All rights reserved.

(Akiskal, 2004; Akiskal et al., 1985; Perugi et al., 2003; Smith et al., 2004). For example, both BPD and BP are associated with emotional lability, impulsivity, irritability and anger, unstable interpersonal relationships, feelings of emptiness, and suicidality (Akiskal, 2004; Bowden and Maier, 2003; Henry et al., 2001; Perugi and Akiskal, 2002). The diagnosis of comorbid BPD in BP is also sensitive to the cross-sectional presentation of any active mood symptoms at the time of assessment. For instance, studies have shown that BPD assessments made during episodes of BP illness lead to a 30% increase in BPD prevalence rates compared to if the diagnosis is made during periods of euthymia (Smith et al., 2004). However, others view BPD as a distinct diagnostic entity (Gunderson, 2009) because, even within areas of shared symptomology, there are significant differences in the phenomenology of BPD as compared to BP (Feliu-Soler et al., 2013; Zimmerman and Morgan, 2013). For example, while both BPD and BP patients experience affective lability, the severity and direction of affective shifts differ between groups (Henry et al., 2001; Nilsson et al., 2010). Such observations

40

T.M. Fonseka et al. / Journal of Affective Disorders 170 (2015) 39–45

challenge the conceptualization of BPD as part of the BP spectrum. However, most agree that patients with the symptoms of BPD are highly stigmatized and often neglected. For a recent review on the differential nature of BPD relative to BP, refer to (Bayes et al., 2014). In many cases, it is not a question of BP “or” BPD, but rather BP “and” BPD, and in such cases there appears to be greater complexity and symptom burden. Rates of comorbid BPD in adult BP range from 12% to 30% (Barbato and Hafner, 1998; Benazzi, 2000; Rossi et al., 2001; Vieta et al., 1999), with BPD occurring in approximately 10% of BP-I and 23% of BP-II patients (Zimmerman and Morgan, 2013). Some studies report BPD as the most common personality disorder among adult BP patients (O’Connell et al., 1991; Peselow et al., 1995; Vieta et al., 1999). Adult BP patients with comorbid personality disorders have less favorable outcomes including longer and more frequent hospitalizations (Barbato and Hafner, 1998; Dunayevich et al., 2000), increased suicidal ideation and attempts (Carpiniello et al., 2011; Vieta et al., 1999), greater symptom severity and functional impairment (Barbato and Hafner, 1998; Carpenter et al., 1995; George et al., 2003), earlier age of mood symptom onset (Vieta et al., 1999), greater unemployment (Kay et al., 2002), higher rates of axis I comorbidity (Kay et al., 2002; Preston et al., 2004), and worsened long-term outcomes of symptomatic and functional recovery (Bieling et al., 2003; Dunayevich et al., 2000) compared to those without personality disorders. This comorbidity has been further associated with poor pharmacotherapy outcomes as evidenced by reduced compliance (Colom et al., 2000) and response to treatment (Barbato and Hafner, 1998), and necessity for polypharmacy (Kay et al., 2002). Only one study to our knowledge has focused on personality disorders in BP adolescents. (Kutcher et al., 1990) found that 15% of BP adolescents had comorbid BPD, and among the total sample, personality disturbance was associated with greater use of antipsychotics and decreased lithium response. No studies have specifically investigated the effects of BPD in BP youth. This adolescent BP-BPD relationship is important to investigate given that maladaptive personality traits often first appear during adolescence or early adulthood (Bowden and Maier, 2003), and can negatively impact long-term patient outcomes (Winograd et al., 2008). Given the propensity of comorbid BPD to yield greater symptomatic burden and impairment in BP adults, coupled with a paucity of data on this topic in adolescent samples, we sought to examine the demographic and clinical correlates of self-reported borderline personality-spectrum symptoms (BPSS) among BP adolescents. Given previous findings, we hypothesized that high levels of BPSS would be associated with worsened BP outcomes, as determined by earlier age of mood symptom onset, and increased symptom

severity, functional impairment, axis I comorbidity, psychiatric hospitalization, and suicidality.

2. Methodology 2.1. Participants Ninety adolescent participants, 13–19 years of age, with a Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Revised (DSM-IV-TR) diagnosis of BP-I, BP-II or operationalized BP Not Otherwise Specified (NOS) were included in the study. Participants were recruited from a tertiary sub-specialty outpatient clinic in an academic health sciences center. Operationalized BP-NOS was defined according to the Course and Outcome of Bipolar Youth (COBY) study criteria (for details see (Birmaher et al., 2006)). Participants and their parent(s)/guardian(s) provided written informed consent after reviewing study parameters with research staff. This study was approved by the local research ethics board. 2.2. Assessment Demographic information was collected for all participants including age, sex, race, and family composition. Psychiatric diagnoses were determined using the Schedule for Affective Disorders and Schizophrenia for School-Aged Children, Present and Lifetime Version (KSADS-PL) (Kaufman et al., 1997), a semistructured diagnostic interview. Bachelor's or Master's-level interviewers completed extensive training under the supervision of the senior author, who also provided diagnostic consensus on all cases (B.G). Mood symptom severity was determined with the KSADS Depression Rating Scale (DRS) (Chambers et al., 1985), and KSADS Mania Rating Scale (MRS) (Axelson et al., 2003). Age of depressive and hypo/manic symptom onset was defined as the age when symptoms first impaired functioning. Substance use disorders were defined as alcohol and/or drug abuse and/or dependence. BPSS were self-reported using the 60-item Life Problems Inventory (LPI) which assessed symptom severity across four BPD-related subscales: identity confusion, interpersonal problems, impulsivity, and emotional lability (Rathus and Miller, 1995) (refer to Table 1). Participants characterized BPSS over the past six months using a 5-point likert scale (1 ¼ “not at all like me” to 5¼ “extremely like me”). According to Rathus, Wagner, and Miller's 2005 validation study of the LPI (as cited in (Muehlenkamp et al., 2011)), the LPI was developed and validated using psychiatric outpatient and community-based adolescent samples. Preliminary

Table 1 Life Problems Inventory (LPI) subscale examples (Rathus and Miller, 1995). LPI subscale

Item examples

Confusion about self

“I’m not sure I know who I am or what I want in my life” “Other kids my age seem more sure than I am of who they are and what they want” “I am so different at different times I sometimes don’t know who I really am”

Interpersonal chaos

“Relationships with people I care about have a lot of ups and downs” “Many of my relationships have been full of intense arguments” “I have had a lot of break-ups with people I’ve been close to”

Impulsivity

“I usually act quickly, without thinking” “If I want to do something, I just do it without thinking of what might happen” “I’ve spent money on things I didn’t need or couldn’t afford”

Emotional dysregulation

“I sometimes get so upset that I want to hurt myself seriously” “When I don’t get my way, I quickly lose my temper” “Once I get upset, it takes me a long time to calm down”

T.M. Fonseka et al. / Journal of Affective Disorders 170 (2015) 39–45

data support the LPI's internal consistency reliability (alpha coefficients:0.89–0.92; subscale alpha coefficients:0.82–0.90; testretest correlations:0.88–0.91), and criterion validity with the BPD subscale of the Structured Clinical Interview for DSM-IV Axis II

41

Personality Disorders (SCID-II) (Muehlenkamp et al., 2011; Rathus and Miller, 2002). Finally, LPI discriminant validity differentiates BPD adolescents from depressed patients and healthy controls (Muehlenkamp et al., 2011).

Table 2 Univariate analyses examining demographic and clinical characteristics associated with BPSS.

Demographics Age Sex (%female) Race (%Caucasian) Living with both natural parents

High BPSS (n¼ 44)

Low BPSS (n ¼46)

15.86 7 1.19 33 (75.0%) 37 (84.1%) 23 (52.3%)

16.48 71.68 28 (60.9%) 41 (89.1%) 30 (65.2%)

χ2/t value

2.01 2.06 0.49 1.56

p-value

pcorrecteda

0.048* 0.152 0.482 0.212

0.134 0.322 0.637 0.362

0.011* 0.036* 0.887 o 0.001* 0.025* 0.022* 0.493 0.258 0.001* 0.139 o 0.001* 0.806

0.064 0.119 0.922 0.004* 0.106 0.106 0.637 0.407 0.018* 0.322 0.004* 0.872

Symptom severity and clinical characterization (Domain 1) BP-I 6 (13.6%) BP-II 24 (54.5%) BP-NOS 14 (31.8%) Depression severity (current episode) 23.747 11.73 Depression severity (past episode) 32.42 76.42 Hypo/mania severity (current episode) 18.45 7 10.56 Hypo/mania severity (past episode) 26.16 79.34 Psychosis 8 (18.2%) Age of onset of depressive symptoms 11.19 7 3.66 Age of onset of hypo/manic symptoms 13.39 7 2.40 CGAS Score (Current Episode) 48.43 7 7.00 CGAS score (most severe episode) 41.18 75.52

17 (37.0%) 15 (32.6%) 14 (30.4%) 13.91 711.49 28.34 79.89 13.077 11.38 27.63 7 10.82 13 (28.3%) 14.007 2.58 14.317 2.49 56.047 10.32 40.80 78.69

Lifetime Psychiatric Comorbidities (Domain 2) Panic disorder Social phobia Generalized anxiety disorder Any anxiety Anorexia nervosa Bulimia nervosa Conduct disorder (CD) Oppositional Defiant disorder (ODD) CD and/or ODD Attention deficit hyperactivity disorder Substance use disorder

11 (25.0%) 18 (40.9%) 27 (61.4%) 36 (81.8%) 2 (4.5%) 6 (13.6%) 8 (18.2%) 22 (50.0%) 23 (52.3%) 19 (43.2%) 18 (40.9%)

6 (13.0%) 7 (15.6%) 18 (39.1%) 31 (67.4%) 0 (0.0%) 3 (6.5%) 0 (0.0%) 10 (21.7%) 10 (21.7%) 17 (37.0%) 12 (26.1%)

2.10 7.08 4.45 2.46 2.14 1.27 9.18 7.84 9.03 0.36 2.22

0.147 0.008* 0.035* 0.117 0.144 0.261 0.002* 0.005* 0.003* 0.547 0.136

0.322 0.053 0.119 0.310 0.322 0.407 0.027* 0.044* 0.032* 0.647 0.322

Lifetime Safety (Domain 3) Police contact/arrest Homicidal ideation Death threats Assault of others Sexual activity Non-suicidal self injury Suicidal ideation Suicide attempt Physical abuse Sexual abuse

20 (45.5%) 9 (22.0%) 8 (19.5%) 17 (41.5%) 22 (53.7%) 27 (61.4%) 32 (72.7%) 14 (31.8%) 6 (13.6%) 6 (13.6%)

17 (37.0%) 1 (2.4%) 4 (9.5%) 8 (19.0%) 20 (48.8%) 16 (34.8%) 24 (52.2%) 9 (19.6%) 1 (2.2%) 4 (8.7%)

0.67 7.50 1.67 4.95 0.20 6.37 4.04 1.78 4.12 0.56

0.413 0.006* 0.196 0.026* 0.659 0.012* 0.044* 0.183 0.042* 0.456

0.600 0.045* 0.359 0.106 0.743 0.064 0.130 0.359 0.130 0.620

Lifetime Treatment History Psychiatric hospitalization Any psychotropic medication Second generation antipsychotic Stimulant Lamotrigine SSRI antidepressant Non-SSRI antidepressant Lithium Antimanic anticonvulsantb

20 (45.5%) 30 (68.2%) 21 (47.7%) 5 (11.4%) 2 (4.5%) 16 (36.4%) 6 (13.6%) 4 (9.1%) 5 (11.4%)

24 (52.2%) 34 (73.9%) 26 (56.5%) 10 (21.7%) 4 (8.7%) 11 (23.9%) 7 (15.2%) 12 (26.1%) 5 (10.9%)

0.41 0.36 0.70 1.74 0.62 1.66 0.05 4.44 0.01

0.524 0.549 0.404 0.187 0.430 0.198 0.831 0.035* 0.941

0.646 0.647 0.600 0.359 0.600 0.359 0.881 0.119 0.959

Family Psychiatric History (1st and 2nd Generation) Major depressive episode 36 (81.8%) Hypo/mania 21 (47.7%) Conduct disorder 2 (4.5%) Anxiety 24 (54.5%) Substance use disorder 21 (47.7%) Suicidal ideation 15 (34.1%) Suicide attempt 17 (38.6%)

35 (76.1%) 26 (60.5%) 4 (8.7%) 25 (54.3%) 19 (44.2%) 17 (39.5%) 12 (26.1%)

0.44 1.42 0.62 0.00 0.11 0.28 1.62

0.505 0.233 0.430 0.985 0.740 0.599 0.203

0.637 0.386 0.600 0.985 0.817 0.690 0.359

6.43 4.41 0.02  3.95  2.29  2.33 0.69 1.28 3.59 1.50 4.11  0.25

Values represent number (percent) or mean 7 SD. n

Value significant at p o 0.05. Value corrected for multiple comparisons using false discovery rate. b Antimanic anticonvulsant:valproic acid, divalproex, carbamazepine; BPSS:borderline personality-spectrum symptoms; BP-I:bipolar I disorder; BP-II:bipolar II disorder; BP-NOS:bipolar disorder not otherwise specified; CGAS:Children's Global Assessment Scale; SSRI:selective serotonin reuptake inhibitor. a

42

T.M. Fonseka et al. / Journal of Affective Disorders 170 (2015) 39–45

History of lifetime aggression, suicidality, and sexual activity were obtained from a Safety Assessment Form. History of lifetime physical and/or sexual abuse was ascertained using a Medical History Questionnaire and the KSADS-PL Post-Traumatic Stress Disorder Screen. Measures of global functioning were clinicianrated using the Children's Global Assessment Scale (CGAS) (Shaffer et al., 1983). Family psychiatric history was reported for first and second generation relatives using the Family History Screen (Weissman et al., 2000). 2.3. Statistical analysis Since the LPI does not have established norms or cut-offs, BPSS status was determined using dichotomized total LPI scores with subjects at or above the 50th percentile (“high BPSS”, n ¼44) compared against the remainder of the sample (“low BPSS”, n ¼46). Based on the normal distribution pattern of LPI scores in our sample, a median cut-point was selected over other alternatives to increase statistical power through an almost equal size of observation between groups. A median cut-point, which has been used by other studies (Davis and Brekke, 2013; Li et al., 2014; Sullivan et al., 2012), also prevents any unusual or outlier data points from biasing statistical results (DeCoster et al., 2009). Between-group differences were compared using Pearson χ2 tests for categorical variables, and Student t-tests for continuous variables. False discovery rate was used to correct for multiple comparisons within univariate analyses. Variables significant from univariate analyses with an unadjusted p-value of o 0.05 were analyzed within a logistic regression model performed separately for each domain (refer to Table 2 for domain and variable classification), using BPSS status (high vs. low) for total LPI as the dependent variable. Variables significant at the p o0.05 level across all domain analyses were simultaneously analyzed in an omnibus logistic regression, controlling for participant age. Pearson correlation coefficients examined associations between dimensional scores across total LPI and current CGAS (CGAS-C). Statistical analyses were performed using Statistical Package for the Social Sciences (SPSS), version 20.0.

3. Results 3.1. Univariate Analyses Participants were 90 English-speaking males (n¼ 29) and females (n¼ 61), 13–19 years of age (mean age: 16.18 71.5 years), of various ethnicities, with a DSM-IV-TR diagnosis of BP-I (n ¼23), BP-II (n ¼39), or BP-NOS (n ¼ 28). Results from univariate analyses are summarized in Table 2. High BPSS participants were younger (t(81) ¼2.01, p ¼0.048), and more likely to have a diagnosis of BP-II (54.5% vs. 32.6%, p ¼0.036), in addition to a reduced likelihood of BP-I (13.6% vs. 37.0%, p ¼0.011), as compared to low BPSS participants. High BPSS participants showed higher rates of lifetime psychiatric comorbidity across all examined diagnoses, with significant differences emerging for social phobia (40.9% vs. 15.6%, p ¼0.008), generalized anxiety disorder (GAD) (61.4% vs. 39.1%, p ¼0.035), conduct disorder (CD) (18.2% vs. 0.0%, p ¼0.002), and oppositional defiant disorder (ODD) (50.0% vs. 21.7%, p ¼0.005). All lifetime safety-related variables were elevated in the high BPSS group, with significant differences emerging for homicidal ideation (22.0% vs. 2.4%, p ¼0.006), assault of others (41.5% vs. 19.0%, p ¼0.026), non-suicidal self injury (NSSI) (61.4% vs. 34.8%, p ¼0.012), suicidal ideation (72.7% vs. 52.2%, p ¼0.044), and physical abuse (13.6% vs. 2.2%, p ¼0.042). High BPSS were associated with worsened clinical course, as evidenced by greater current (t(85) ¼  3.95, p o0.001) and past

(t(74) ¼ 2.29, p¼ 0.025) depressive episode severity, current hypo/manic episode severity (t(88) ¼  2.33, p ¼0.022), younger age of depressive symptom onset (t(62) ¼3.59, p¼ 0.001), and reduced current global functioning (t(79) ¼4.11, p o0.001). Despite these worsened outcomes, high BPSS participants were less likely to report lifetime use of lithium (9.1% vs. 26.1%, p ¼0.035). After applying a correction for multiple comparisons, the following outcomes remained significant: CD (p¼ 0.027), ODD (p ¼0.044), homicidal ideation (p ¼0.045), current depression severity (p ¼ 0.004), age of depression onset (p¼ 0.018), and current global functioning (p ¼0.004). No between-group differences were observed for family psychiatric history. There was a significant negative correlation between total LPI and CGAS-C scores (r ¼  0.39, p o0.001) indicating a decline in global functioning with increasing BPSS. Exploratory analyses across each LPI subscale (identity confusion, interpersonal problems, impulsivity, emotional lability) revealed a similar overall pattern of findings as observed within total LPI univariate analyses (data not shown). In particular, analysis of the identity confusion subscale, which isolates a symptomatic construct specific to BPD over BP, revealed that BP participants who endorse high levels of ‘confusion about the self’ have the same general outcomes as presented here for total LPI analyses, in addition to a greater likelihood of panic disorder, and reduced likelihood of lifetime psychosis. These results suggest that overall study findings are indeed informative of a true synergy between BP and BPSS on worsened clinical trajectories in adolescent patients, rather than being a mere artifact of more severe BP symptomology. 3.2. Multivariate analyses Variables significant from total LPI univariate analyses at the unadjusted p o0.05 level were analyzed within logistic regression models performed separately for each domain: (Domain 1) symptom severity and clinical characterization, (Domain 2) lifetime psychiatric comorbidities, and (Domain 3) lifetime safety-related variables. Demographics, lifetime treatment history, and family psychiatric history domains were not analyzed due to a predominance of non-significant findings. However, lithium use was included in Domain 1 as the only medication that was associated with BPSS. Due to a zero frequency distribution in the low BPSS group for CD, CD and ODD were analyzed together as a single variable (CD and/or ODD). To minimize redundancy, only the BP-I variable was entered into the regression while the BP-II variable was excluded. Variables significant at the p o0.05 level from each domain analysis (refer to Table 3) were simultaneously analyzed in an omnibus logistic regression, where participant age was a forced

Table 3 Domain-specific multivariable predictors of high BPSS. OR

95% CI

Symptom Severity and Clinical Characterization BP-I 0.09 0.01–0.58 Age of Onset of Depressive Symptoms 0.73 0.53–1.00 CGAS Score (Current Episode) 0.78 0.69–0.90

Wald

p-value

6.40 3.89 12.91

0.011 0.048 o 0.001

Lifetime Psychiatric Comorbidities Social Phobia CD and/or ODD

3.57 4.99

1.19–10.74 1.82–13.67

5.12 9.74

0.023 0.002

Lifetime Safety Non-Suicidal Self Injury

2.79

1.05–7.47

4.19

0.041

BP-I:bipolar I disorder; CGAS:Children's Global Assessment Scale; CD:conduct disorder; ODD:oppositional defiant disorder, OR:odds ratio, CI:confidence interval.

T.M. Fonseka et al. / Journal of Affective Disorders 170 (2015) 39–45

covariate. Six variables were included: age of depression onset, BP-I, current CGAS, social phobia, CD and/or ODD, and NSSI. Of the six variables entered into the model, the following four remained significant: lifetime social phobia (OR ¼18.26, CI: 2.15154.94, p ¼0.008), NSSI (OR¼ 7.24, CI: 1.12-0.46.95, p¼ 0.038), CGAS-C (OR ¼0.82, CI: 0.74-0.92, p ¼0.001), and CD and/or ODD (OR ¼9.35, CI: 1.49-58.58, p ¼0.017).

4. Discussion In this sample of adolescents with BP, univariate analyses revealed that high BPSS were associated with younger age, earlier age of depression onset, greater mood symptom severity, increased rates of axis I comorbidity in the areas of anxiety and disruptive behavior disorders, greater risk-related behaviors in the areas of aggression, suicidality and abuse, and greater functional impairment. Despite this clinical profile, high BPSS participants did not report greater exposure to psychotropic medication, and in fact had less exposure to lithium. However, after applying a correction for multiple comparisons, only those associations involving disruptive behavior disorders, homicidal ideation, depression severity and age of onset, and global functioning remained significant. Exploratory analyses across each LPI subscale revealed a similar overall pattern of findings, including the identity confusion subscale which assessed a BPD-specific construct that had minimal symptomatic overlap with BP. Multivariate analyses identified lifetime social phobia, lifetime CD and/or ODD, NSSI, and reduced global functioning as the most robust independent predictors of high BPSS. However, this study design is not without limitations. First, the use of cross-sectional and retrospective clinical assessments is subject to potential recall bias. In the absence of longitudinal data, inferences of causality for significant associations cannot be determined. For example, mood symptoms may have impacted self-reported BPSS or the converse may be true. Second, since the LPI is based on self-report, incorporating a parent-report and/or diagnostic interview for BPD may have yielded different findings. Third, dichotomization of total LPI scores may have resulted in a loss of analytic power and a reduction in effect size estimates (DeCoster et al., 2009). Despite being one of the larger samples of BP adolescents, this study was not powered to detect differences with small effect sizes, particularly in multivariate analyses. Despite these limitations, our findings contribute to the nascent literature on the significance of BPSS among BP adolescents, a topic of significant clinical and scientific importance. As documented in other studies, high BPSS in our sample were associated with BP-II (Benazzi, 2000, 2006; Skodol et al., 1999), and younger participant age (Grant et al., 2008). Preston and colleagues (Preston et al., 2004) reported BPD to be significantly more common among BP females compared to males. We found a numerically, but not statistically significantly, higher frequency of females in the high BPSS group. High BPSS in our sample were associated with increased psychiatric comorbidity in the areas of anxiety and disruptive behavior disorders, and greater mood symptom severity. Similar clinical correlates have been observed among BP adults with personality disorders (Garno et al., 2005; George et al., 2003), and independent BPD samples (Lenzenweger et al., 2007; Lewinsohn et al., 1997; Skodol et al., 1999; Zanarini et al., 1998; Zimmerman and Mattia, 1999), with additional work suggesting that rates of axis I comorbidity decline in response to BPD remission (Zanarini et al., 2004). Despite these findings, high BPSS participants had lower rates of lithium use. Reasons for this are uncertain but may be due to concerns about lethality of lithium in overdose or treatment biases driven by the salience of BPSS. For example, since BPD patients are highly stigmatized as exceedingly disruptive and unresponsive to treatment (Nehls, 1998), clinicians may negatively perceive and be

43

reluctant to treat BPD patients (Aviram et al., 2006; Sansone and Sansone, 2013). BPD patients are at an elevated risk of suicidality and aggression. Along with increased rates of CD and ODD in our sample, high BPSS were positively associated with assault of others and homicidal ideation. Similar findings have been documented in BPD studies, with reports of higher rates of BPD in prison populations, often for perpetration of violent offenses (Latalova and Prasko, 2010; Sansone and Sansone, 2009). BPD has also been associated with self-injury and suicidality (Dulit et al., 1994; Gross et al., 2002), with approximately 10% of BPD patients completing suicide (Paris, 2002). (Zimmerman et al., 2014) found that BPD-BP comorbid samples were more likely to attempt suicide than BP patients alone. In our sample, high BPSS was associated with greater NSSI and suicidal ideation, yet suicide attempt rates, although elevated among high BPSS participants, did not reach statistical significance. In conclusion, the results of our study suggest that BP adolescents with high BPSS have less favorable clinical and prognostic outcomes in the areas of increased axis I comorbidity, elevated risk of suicidality and aggression, younger age of mood symptom onset, greater mood symptom severity, and greater functional impairment. These findings are supported by several studies that have concluded that comorbid personality disorders in BP negatively affect patient outcomes (Barbato and Hafner, 1998; Bieling et al., 2003; Carpenter et al., 1995; Carpiniello et al., 2011; Colom et al., 2000; Dunayevich et al., 2000; George et al., 2003; Kay et al., 2002; Kutcher et al., 1990; Preston et al., 2004; Vieta et al., 1999; Winograd et al., 2008). Presence of high BPSS among BP adolescents may suggest the need to modify clinical monitoring and treatment practices. Psychosocial (Clarkin et al., 2007; Giesen-Bloo et al., 2006) and/or psychotropic (Ripoll, 2013) treatment may offer beneficial effects for BPSS. In particular, dialectical behavior therapy (DBT), which is routinely used to treat BPD (Lynch et al., 2007), may be efficacious in comorbid cases as preliminary findings suggest beneficial effects among BP adolescents (Goldstein et al., 2007). Additional work in neuroimaging has characterized BPD by dysfunction in various brain regions, particularly fronto-limbic circuits (Ruocco et al., 2013, 2010; Schmahl and Bremner, 2006), and identified both similar and differential patterns of fronto-limbic activation between BPD and BP patients (Malhi et al., 2013). Whether present findings would extend to full-threshold BPD is uncertain and this question warrants further investigation. Although the LPI can neither be used to diagnose BPD or BP, nor to distinguish between these overlapping disorders, present findings suggest that this self-report may have clinically relevant heuristic value in identifying BP adolescents with high-risk presentations. Further research in this area is required to more clearly elucidate the effects of BPSS among BP adolescents. In particular, these compelling preliminary findings warrant replication in larger samples using prospective methodology and employing biological markers. Role of funding source The authors would like to thank the grant support from Canadian Institutes of Health Research, Heart & Stroke Foundation of Ontario, Ontario Mental Health Foundation, and anonymous philanthropic donations. The content of this paper is solely the responsibility of the authors and does not necessarily represent the official views of these organizations. These funding sources were not directly involved in the study design; collection, analysis and interpretation of data; writing of the report; or the decision to submit the article for publication.

Conflict of interest We confirm this manuscript describes original work that has not been published elsewhere and is not under consideration by another journal. We have no conflicts of interest to disclose.

Acknowledgments The authors would like to thank the participants of this study. We have no acknowledgments of assistance to disclose.

44

T.M. Fonseka et al. / Journal of Affective Disorders 170 (2015) 39–45

References Akiskal, H.S., 2004. Demystifying borderline personality: critique of the concept and unorthodox reflections on its natural kinship with the bipolar spectrum. Acta Psychiatr. Scand. 110, 401–407. Akiskal, H.S., Cassano, G.B., Musetti, L., Perugi, G., Tundo, A., Mignani, V., 1989. Psychopathology, temperament, and past course in primary major depressions 1. Review of evidence for a bipolar spectrum. Psychopathology 22, 268–277. Akiskal, H.S., Chen, S.E., Davis, G.C., Puzantian, V.R., Kashgarian, M., Bolinger, J.M., 1985. Borderline: an adjective in search of a noun. J. Clin. Psychiatry 46, 41–48. American Psychiatric Association, 2013. Diagnostic and Statistical Manual of Mental Disorders, 5th Ed. American Psychiatric Publishing, Arlington, VA. Aviram, R.B., Brodsky, B.S., Stanley, B., 2006. Borderline personality disorder, stigma, and treatment implications. Harv. Rev. Psychiatry 14, 249–256. Axelson, D., Birmaher, B.J., Brent, D., Wassick, S., Hoover, C., Bridge, J., Ryan, N., 2003. A preliminary study of the kiddie schedule for affective disorders and schizophrenia for school-age children mania rating scale for children and adolescents. J. Child Adolesc. Psychopharmacol. 13, 463–470. Barbato, N., Hafner, R.J., 1998. Comorbidity of bipolar and personality disorder. Aust. NZ J. Psychiatry 32, 276–280. Bayes, A., Parker, G., Fletcher, K., 2014. Clinical differentiation of bipolar II disorder from borderline personality disorder. Curr. Opin. Psychiatry 27, 14–20. Benazzi, F., 2000. Borderline personality disorder and bipolar II disorder in private practice depressed outpatients. Compr. Psychiatry 41, 106–110. Benazzi, F., 2006. Borderline personality-bipolar spectrum relationship. Prog. Neuro-Psychopharmacol. Biol. Psychiatry 30, 68–74. Bieling, P.J., MacQueen, G.M., Marriot, M.J., Robb, J.C., Begin, H., Joffe, R.T., Young, L.T., 2003. Longitudinal outcome in patients with bipolar disorder assessed by lifecharting is influenced by DSM-IV personality disorder symptoms. Bipolar Disord. 5, 14–21. Birmaher, B., Axelson, D., Strober, M., Gill, M.K., Valeri, S., Chiappetta, L., Ryan, N., Leonard, H., Hunt, J., Iyengar, S., Keller, M., 2006. Clinical course of children and adolescents with bipolar spectrum disorders. Arch. Gen. Psychiatry 63, 175–183. Bowden, C., Maier, W., 2003. Bipolar disorder and personality disorder. Eur. Psychiatry: J. Assoc. Eur. Psychiatrists 18 (Suppl 1), S9–S12. Carpenter, D., Clarkin, J.F., Glick, I.D., Wilner, P.J., 1995. Personality pathology among married adults with bipolar disorder. J. Affect. Disord. 34, 269–274. Carpiniello, B., Lai, L., Pirarba, S., Sardu, C., Pinna, F., 2011. Impulsivity and aggressiveness in bipolar disorder with co-morbid borderline personality disorder. Psychiatry Res. 188, 40–44. Chambers, W.J., Puig-Antich, J., Hirsch, M., Paez, P., Ambrosini, P.J., Tabrizi, M.A., Davies, M., 1985. The assessment of affective disorders in children and adolescents by semistructured interview. Test-retest reliability of the schedule for affective disorders and schizophrenia for school-age children, present episode version. Arch. Gen. Psychiatry 42, 696–702. Clarkin, J.F., Levy, K.N., Lenzenweger, M.F., Kernberg, O.F., 2007. Evaluating three treatments for borderline personality disorder: a multiwave study. Am. J. Psychiatry 164, 922–928. Colom, F., Vieta, E., Martinez-Aran, A., Reinares, M., Benabarre, A., Gasto, C., 2000. Clinical factors associated with treatment noncompliance in euthymic bipolar patients. The J. Clin. Psychiatry 61, 549–555. Davis, L., Brekke, J.S., 2013. Social networks and arrest among persons with severe mental illness: an exploratory analysis. Psychiatr. Serv. 64, 1274–1277. DeCoster, J., Iselin, A.M., Gallucci, M., 2009. A conceptual and empirical examination of justifications for dichotomization. Psychol. Methods 14, 349–366. Dulit, R.A., Fyer, M.R., Leon, A.C., Brodsky, B.S., Frances, A.J., 1994. Clinical correlates of self-mutilation in borderline personality disorder. Am. J. Psychiatry 151, 1305–1311. Dunayevich, E., Sax, K.W., Keck Jr., P.E., McElroy, S.L., Sorter, M.T., McConville, B.J., Strakowski, S.M., 2000. Twelve-month outcome in bipolar patients with and without personality disorders. J. Clin. Psychiatry 61, 134–139. Feliu-Soler, A., Soler, J., Elices, M., Pascual, J.C., Perez, J., Martin-Blanco, A., Santos, A., Crespo, I., Perez, V., Portella, M.J., 2013. Differences in attention and impulsivity between borderline personality disorder and bipolar disorder. Psychiatry Res. 210, 1307–1309. Garno, J.L., Goldberg, J.F., Ramirez, P.M., Ritzler, B.A., 2005. Bipolar disorder with comorbid cluster B personality disorder features: impact on suicidality. J. Clin. Psychiatry 66, 339–345. George, E.L., Miklowitz, D.J., Richards, J.A., Simoneau, T.L., Taylor, D.O., 2003. The comorbidity of bipolar disorder and axis II personality disorders: prevalence and clinical correlates. Bipolar Disord. 5, 115–122. Giesen-Bloo, J., van Dyck, R., Spinhoven, P., van Tilburg, W., Dirksen, C., van Asselt, T., Kremers, I., Nadort, M., Arntz, A., 2006. Outpatient psychotherapy for borderline personality disorder: randomized trial of schema-focused therapy vs transference-focused psychotherapy. Arch. Gen. Psychiatry 63, 649–658. Goldstein, T.R., Axelson, D.A., Birmaher, B., Brent, D.A., 2007. Dialectical behavior therapy for adolescents with bipolar disorder: a 1-year open trial. J. Am. Acad. Child Adolesc. Psychiatry 46, 820–830. Grant, B.F., Chou, S.P., Goldstein, R.B., Huang, B., Stinson, F.S., Saha, T.D., Smith, S.M., Dawson, D.A., Pulay, A.J., Pickering, R.P., Ruan, W.J., 2008. Prevalence, correlates, disability, and comorbidity of DSM-IV borderline personality disorder: results from the wave 2 national epidemiologic survey on alcohol and related conditions. J. Clin. Psychiatry 69, 533–545.

Gross, R., Olfson, M., Gameroff, M., Shea, S., Feder, A., Fuentes, M., Lantigua, R., Weissman, M.M., 2002. Borderline personality disorder in primary care. Arch. Intern. Med. 162, 53–60. Gunderson, J.G., 2009. Borderline personality disorder: ontogeny of a diagnosis. Am. J. Psychiatry 166, 530–539. Henry, C., Mitropoulou, V., New, A.S., Koenigsberg, H.W., Silverman, J., Siever, L.J., 2001. Affective instability and impulsivity in borderline personality and bipolar II disorders: similarities and differences. J. Psychiatr. Res. 35, 307–312. Kaufman, J., Birmaher, B., Brent, D., Rao, U., Flynn, C., Moreci, P., Williamson, D., Ryan, N., 1997. Schedule for affective disorders and schizophrenia for schoolage children-present and lifetime version (K-SADS-PL): initial reliability and validity data. J. Am. Acad. Child Adolesc. Psychiatry 36, 980–988. Kay, J.H., Altshuler, L.L., Ventura, J., Mintz, J., 2002. Impact of axis II comorbidity on the course of bipolar illness in men: a retrospective chart review. Bipolar Disord. 4, 237–242. Kutcher, S.P., Marton, P., Korenblum, M., 1990. Adolescent bipolar illness and personality disorder. J. Am. Acad. Child Adolesc. Psychiatry 29, 355–358. Latalova, K., Prasko, J., 2010. Aggression in borderline personality disorder. Psychiatr. Q. 81, 239–251. Lenzenweger, M.F., Lane, M.C., Loranger, A.W., Kessler, R.C., 2007. DSM-IV personality disorders in the National Comorbidity Survey Replication. Biol. Psychiatry 62, 553–564. Lewinsohn, P.M., Rohde, P., Seeley, J.R., Klein, D.N., 1997. Axis II psychopathology as a function of axis I disorders in childhood and adolescence. J. Am. Acad. Child Adolesc. Psychiatry 36, 1752–1759. Li, Z., Chen, K., Jiang, P., Zhang, X., Li, X., 2014. CD44v/CD44s expression patterns are associated with the survival of pancreatic carcinoma patients. Diagn. Pathol. 9, 79. Lynch, T.R., Trost, W.T., Salsman, N., Linehan, M.M., 2007. Dialectical behavior therapy for borderline personality disorder. Ann. Rev. Clin. Psychol. 3, 181–205. Malhi, G.S., Tanious, M., Fritz, K., Coulston, C.M., Bargh, D.M., Phan, K.L., Calhoun, V., Das, P., 2013. Differential engagement of the fronto-limbic network during emotion processing distinguishes bipolar and borderline personality disorder. Mol. Psychiatry 18, 1247–1248. Morriss, R., 2002. Clinical importance of inter-episode symptoms in patients with bipolar affective disorder. J. Affect. Disord. 72 (Suppl 1), S3–S13. Muehlenkamp, J.J., Ertelt, T.W., Miller, A.L., Claes, L., 2011. Borderline personality symptoms differentiate non-suicidal and suicidal self-injury in ethnically diverse adolescent outpatients. J. Child Psychol. Psychiatry Allied Discip. 52, 148–155. Nehls, N., 1998. Borderline personality disorder: gender stereotypes, stigma, and limited system of care. Issues Mental Health Nurs. 19, 97–112. Nilsson, A.K., Jorgensen, C.R., Straarup, K.N., Licht, R.W., 2010. Severity of affective temperament and maladaptive self-schemas differentiate borderline patients, bipolar patients, and controls. Compr. Psychiatry 51, 486–491. O’Connell, R.A., Mayo, J.A., Sciutto, M.S., 1991. PDQ-R personality disorders in bipolar patients. J. Affect. Disord. 23, 217–221. Paris, J., 2002. Chronic suicidality among patients with borderline personality disorder. Psychiatr. Serv. 53, 738–742. Perugi, G., Akiskal, H.S., 2002. The soft bipolar spectrum redefined: focus on the cyclothymic, anxious-sensitive, impulse-dyscontrol, and binge-eating connection in bipolar II and related conditions. Psychiatr. Clin. N. Am. 25, 713–737. Perugi, G., Toni, C., Travierso, M.C., Akiskal, H.S., 2003. The role of cyclothymia in atypical depression: toward a data-based reconceptualization of the borderline-bipolar II connection. J. Affect. Disord. 73, 87–98. Peselow, E.D., Sanfilipo, M.P., Fieve, R.R., 1995. Relationship between hypomania and personality disorders before and after successful treatment. Am. J. Psychiatry 152, 232–238. Preston, G.A., Marchant, B.K., Reimherr, F.W., Strong, R.E., Hedges, D.W., 2004. Borderline personality disorder in patients with bipolar disorder and response to lamotrigine. J. Affect. Disord. 79, 297–303. Rathus, J.H., Miller, A.L., 1995. Life Problems Inventory. Montefiore Medical Center/ Einstein College of Medicine, Bronx, New York. Rathus, J.H., Miller, A.L., 2002. Dialectical behavior therapy adapted for suicidal adolescents. Suicide Life-Threatening Behav. 32, 146–157. Ripoll, L.H., 2013. Psychopharmacologic treatment of borderline personality disorder. Dialogues Clin. Neurosci. 15, 213–224. Rossi, A., Marinangeli, M.G., Butti, G., Scinto, A., Di Cicco, L., Kalyvoka, A., Petruzzi, C., 2001. Personality disorders in bipolar and depressive disorders. J Affect Disord 65, 3–8. Ruocco, A.C., Amirthavasagam, S., Choi-Kain, L.W., McMain, S.F., 2013. Neural correlates of negative emotionality in borderline personality disorder: an activation-likelihood-estimation meta-analysis. Biol. Psychiatry 73, 153–160. Ruocco, A.C., Medaglia, J.D., Ayaz, H., Chute, D.L., 2010. Abnormal prefrontal cortical response during affective processing in borderline personality disorder. Psychiatry Res. 182, 117–122. Sansone, R.A., Sansone, L.A., 2009. Borderline personality and criminality. Psychiatry (Edgmont) 6, 16–20. Sansone, R.A., Sansone, L.A., 2013. Responses of mental health clinicians to patients with borderline personality disorder. Innov. Clin. Neurosci. 10, 39–43. Schmahl, C., Bremner, J.D., 2006. Neuroimaging in borderline personality disorder. J. Psychiatr. Res. 40, 419–427. Shaffer, D., Gould, M.S., Brasic, J., Ambrosini, P., Fisher, P., Bird, H., Aluwahlia, S., 1983. A children's global assessment scale (CGAS). Arch. Gen. Psychiatry 40, 1228–1231.

T.M. Fonseka et al. / Journal of Affective Disorders 170 (2015) 39–45

Skodol, A.E., Stout, R.L., McGlashan, T.H., Grilo, C.M., Gunderson, J.G., Shea, M.T., Morey, L.C., Zanarini, M.C., Dyck, I.R., Oldham, J.M., 1999. Co-occurrence of mood and personality disorders: a report from the collaborative longitudinal personality disorders study (CLPS). Depress. Anxiety 10, 175–182. Smith, D.J., Muir, W.J., Blackwood, D.H., 2004. Is borderline personality disorder part of the bipolar spectrum? Harv. Rev. Psychiatry 12, 133–139. Sullivan, A.E., Judd, C.M., Axelson, D.A., Miklowitz, D.J., 2012. Family functioning and the course of adolescent bipolar disorder. Behav. Ther. 43, 837–847. Vieta, E., Colom, F., Martinez-Aran, A., Benabarre, A., Gasto, C., 1999. Personality disorders in bipolar II patients. J Nerv Ment Dis 187, 245–248. Weissman, M.M., Wickramaratne, P., Adams, P., Wolk, S., Verdeli, H., Olfson, M., 2000. Brief screening for family psychiatric history: the family history screen. Arch. Gen. Psychiatry 57, 675–682. Winograd, G., Cohen, P., Chen, H., 2008. Adolescent borderline symptoms in the community: prognosis for functioning over 20 years. J. Child Psychol. Psychiatry Allied Discip. 49, 933–941. Zanarini, M.C., Frankenburg, F.R., Dubo, E.D., Sickel, A.E., Trikha, A., Levin, A., Reynolds, V., 1998. Axis I comorbidity of borderline personality disorder. Am. J. Psychiatry 155, 1733–1739.

45

Zanarini, M.C., Frankenburg, F.R., Hennen, J., Reich, D.B., Silk, K.R., 2004. Axis I comorbidity in patients with borderline personality disorder: 6-year follow-up and prediction of time to remission. Am. J. Psychiatry 161, 2108–2114. Zanarini, M.C., Frankenburg, F.R., Hennen, J., Reich, D.B., Silk, K.R., 2005. The McLean study of adult development (MSAD): overview and implications of the first six years of prospective follow-up. J. Personality Disord. 19, 505–523. Zimmerman, M., Martinez, J., Young, D., Chelminski, I., Morgan, T.A., Dalrymple, K., 2014. Comorbid bipolar disorder and borderline personality disorder and history of suicide attempts. J. Personal. Disord. 28, 358–364. Zimmerman, M., Mattia, J.I., 1999. Axis I diagnostic comorbidity and borderline personality disorder. Compr. Psychiatry 40, 245–252. Zimmerman, M., Morgan, T.A., 2013. Problematic boundaries in the diagnosis of bipolar disorder: the interface with borderline personality disorder. Curr. Psychiatry Rep. 15, 422.