ORIGINAL ARTICLE

Cognitive and Behavioral Differentiation of Those With Borderline

Personality Disorder and Bipolar Disorder
Adam J. Bayes, MBBS (Hons), PhD*† and Gordon B. Parker, MD, DSc, PhD*†

addition, a large latent class analysis of BPD and BP symptoms in

Abstract: The current study sought to identify features offering differentiation 34,653 individuals from the National Epidemiologic Survey on Alcohol
of borderline personality disorder (BPD) from bipolar disorder (BP). Participants and Related Conditions generated a three-factor solution (i.e., BPD, de-
were clinically assessed and assigned diagnoses based on the Diagnostic and Sta- pression, and mania) with pairwise correlations between the factors in
tistical Manual of Mental Disorders criteria. A 113-item self-report questionnaire keeping with a model suggesting two separate syndromes (BPD and
was completed, comprising cognitive and behavioral constructs weighted to a BP) albeit allowing that they can be coterminous (de la Rosa et al., 2017).
borderline personality style. A total of n = 53 participants were assigned to Structured measures screening for BP disorder offer minimal ca-
BPD, n = 83 to BP, with comorbid participants excluded. Twenty items were pacity to differentiate those with BPD or a BP disorder. For example, in
highly endorsed (>95%) by the BPD group, with most of the features capturing those screening positive for BP disorder on the Mood Disorder Ques-
emotional dysregulation (ED) and identity disturbance; however, many items tionnaire (MDQ), the frequency of BPD was 80% of the rate of BP
were also highly endorsed by the participants with BP. Thirty-eight items offered (Zimmerman et al., 2018). Vöhringer et al. (2016), in another study
differentiation of BPD from BP, with identity disturbance overrepresented. The using the MDQ, found that BP was strongly predicted by the presence
study findings indicate that the transdiagnostic nature of ED (a feature of both of elevated mood, increased goal-directed activity, and episodicity (sen-
conditions) means it is less useful for diagnostic decisions, whereas identity dis- sitivity = 88.7%, specificity = 81.4%, area under receiver operator char-
turbance is both intrinsic to BPD and offers specificity in differentiation from BP. acteristic [auROC] = 0.91), although BPD was only predicted by female
Key Words: Borderline personality disorder, bipolar disorder, diagnosis, sex and with an auROC = 0.67. BPD criteria in the Diagnostic and Sta-
cognition, behavior tistical Manual of Mental Disorders, 5th Edition (DSM-5) (as for its
predecessor DSM-IV), require the individual to display “a pervasive pat-
(J Nerv Ment Dis 2019;207: 620–625)
tern of instability of interpersonal relationships, self-image, and affects,
and marked impulsivity, beginning in early adulthood and present in a
D ifferentiation of borderline personality disorder (BPD) and bipolar
disorders (BPs) is commonly judged as diagnostically difficult
(Bassett, 2012; Bayes et al., 2014). Individuals with each disorder
variety of contexts,” as indicated by nine descriptors, of which five
(or more) must be met to assign a diagnosis. BP is listed in DSM-5 as
one of the key differential diagnoses for BPD, and conversely, BPD is
frequently present with altered mood states, suicidality, and deliberate
listed as a differential diagnosis for both bipolar I and II disorders
self-harm (DSH), as well as displaying behaviors such as excessive
(BP I and BP II, respectively) because of the “substantial symptomatic
spending, gambling, sexual promiscuity, and drug and alcohol misuse
overlap….since mood lability and impulsivity are common to both con-
(Ghaemi et al., 2014). Features common to both disorders are considered
ditions” (p. 132) (American Psychiatric Association, 2013). However,
“transdiagnostic,” including emotional dysregulation (ED) and impulsivity,
other than defining criteria for each condition, there is no specificity
the presence of which may reduce diagnostic accuracy (Yen et al., 2015).
about which items might be more useful in delineating each diagnosis.
At cross-sectional assessment, periods of ED in patients with BPD
The current study aims to determine if cognitive and behavioral
can resemble hypomania (Kernberg and Yeomans, 2013), leading to a
constructs both defining a borderline personality style and differentiat-
false diagnosis of a BP disorder (Ruggero et al., 2010), whereas
ing it from BP can be identified. In two prior studies by our group, we
interepisode residual symptoms in those with a BP condition, such as
sought to differentiate the two conditions across multiple domains, in-
chronic dysphoria (Paykel et al., 2006), can also compromise diagnosis.
cluding a) family history, developmental antecedents, clinical symptoms,
The relationship between BPD and BP has been the subject of
illness correlates, and a 113-item self-report measure of cognitive and be-
much debate including whether they are independent or interdependent
havioral borderline constructs (Bayes et al., 2016a), and b) emotion reg-
conditions. A recent genome-wide association study (Witt et al., 2017)
ulation strategies, using the Difficulties in Emotional Regulation Scale
found genetic overlap of BPD with BP disorder (as well as with schizo-
and the Cognitive Emotion Regulation Questionnaire (Bayes et al.,
phrenia and major depressive disorder), suggesting a shared etiology.
2016b). The current study seeks to examine the full 113 set of items to
At the symptom level, some authors have argued that the fluctuating
determine a) those items most prevalent in the BPD group, b) those items
mood symptoms observed in BPD position it as an “ultrarapid cycling”
that most discriminate BPD from BP disorder, and c) whether the item set
variant of BP disorder and that it should be placed on the bipolar spec-
establishes substantive differentiation alone. In essence, are there sui
trum (Mackinnon and Pies, 2006). However, as reviewed by Paris and
generis cognitive and behavioral characteristics to those with a BPD?
Black (2015), there appear to be important differences in the time scale
of affective changes between the conditions and certain BPD features,
such as interpersonal difficulties and micropsychotic symptoms, that METHODS
are not clearly explainable as resulting from mood fluctuations. In
Participants
*School of Psychiatry, University of New South Wales; and †Black Dog Institute,
As the study design has been reported in detail (Bayes et al.,
Sydney, NSW, Australia. 2016a), we only summarize key aspects. We sought participants 18 years
Send reprint requests to Adam J. Bayes, MBBS (Hons), PhD, Black Dog Institute, and older who had received a prior diagnosis of BPD or a BP (I or II)
Hospital Road, Randwick, Sydney, NSW 2031, Australia. disorder, and recruited participants from several clinical services, in-
E‐mail: a.bayes@unsw.edu.au.
Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.
cluding a tertiary referral depression clinic, two public psychiatric hos-
ISSN: 0022-3018/19/20708–0620 pital inpatient services, three dialectical behavior therapy outpatient
DOI: 10.1097/NMD.0000000000001024 clinics, three private psychiatric clinics, and a public hospital outpatient

620 www.jonmd.com The Journal of Nervous and Mental Disease • Volume 207, Number 8, August 2019

Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.

The Journal of Nervous and Mental Disease • Volume 207, Number 8, August 2019 Differentiation of BPD and BP

psychotherapy clinic. In addition, an invitation was placed on our institute's

Volunteer Research Register and website, and advertisements were TABLE 1. Sociodemographic Features
placed in newspapers. Site ethical approval was ratified by the Univer-
BPD BP
sity of New South Wales Human Research Ethics Committee, and writ-
(n = 53) (n = 83) Test p
ten consent was obtained from all participants. Exclusion criteria
included English language limitations, psychotic or significant organic (BP I = 4;
features, and current substance dependence. BP II = 79)
Age (SD), yrs 32.9 (11.3) 35.4 (13.4) F = 1.9 0.17
Diagnostic Assessment Sex, %
Study diagnoses respected DSM-IV criteria for both condi- Female 84.9 53.7 χ2 = 14.0 <0.001
tions (although clinical diagnoses were also generated). In practice, Male 15.1 46.3
a DSM-IV BP I or BP II disorder diagnosis was determined by adminis- Marital status, %
tering the Mini International Neuropsychiatric Index to all participants. A Divorced 7.7 6.1 χ2 = 5.53 0.14
DSM-IV diagnosis of BPD was generated after administering the Diag- Separated 1.9 3.7
nostic Interview for Personality Disorders IV–Borderline Personality
Married/de facto 23.1 41.5
Disorder section to all interviewees, and those who scored two on five
or more of the nine criteria received a DSM-IV BPD diagnosis. Never married 67.3 48.8
Employment status, %
Instruments Unemployed 17.3 7.4 χ2 = 16.64 0.02
Participants completed a booklet assessing sociodemographic Retired 1.9 2.5
characteristics, current and past mood history, family history, and treat- Disability/sickness 25 9.9
ment data. The booklet also included a 113-item self-report measure of benefits
personality items developed by Ruiz, Fletcher, and Parker (unpub- Pensioner 1.9 0
lished) to assess cognitive and behavioral constructs weighted to a bor- Student 23.1 21
derline personality style, as well as items addressing mood, self-harm Home duties 3.8 1.2
and suicidality, interpersonal relationships, and childhood experiences. Part-time 11.5 24.7
The items were derived by a detailed review of BPD literature and clin- Full-time 15.4 33.3
ical experience and sought to capture a broad range of cognitive and be- Highest level education, %
havioral features. Participants were instructed to answer the personality Some secondary school 15.1 14.6 χ2 = 8.15 0.15
questions based on their general functioning over time (not just during Completed secondary 28.3 29.3
mood swing periods). Questions were either coded on a 5-point Likert school
scale (1 = “not at all characteristic” to 5 = “very characteristic”) or di- Diploma 24.5 13.4
chotomously (“yes” versus “no”).
Bachelor degree 17 20.7
Graduate diploma 9.4 3.7
Data Analysis
Postgraduate degree 5.7 18.3
Likert scale scores were dichotomized, with those nominating a
1 or 2 score assigned as answering “no,” and those scoring 3, 4, or 5
assigned as answering “yes.” Prevalence estimates were calculated for
each of the 113 items as reported by the participants with BPD and
those with BP. Seeking to optimize the most discriminating items, we endorsed by 100% of the subjects with BP, thus not offering any differ-
refined the set by comparing the mean scores between groups using entiation between the conditions. Of the 20 items, 11 captured ED con-
the continuity corrected chi-square test and setting p < 0.01 as the sta- structs (e.g., “If someone is critical of something I do, I feel
tistically significant level. All tests were performed using SPSS, version devastated”) or worry (e.g., “I worry over minor things”); 7 related to
22 (SPSS Inc, Chicago). self-identity (e.g., “I often feel that I have no idea of who I am or any
clear identity”); and the 2 remaining items included one related to rela-
tionships (“I feel a need to make people feel happy”) and one suggestive
RESULTS “bipolar” item (“I have periods where I'm extremely productive at work
A total of 226 participants met provisional criteria for having and other times when I'm not”; endorsed by 98.1% of those with BPD
completed the diagnostic interview and self-report measures. Those and 95.1% of those with BP). Thus, these analyses identified highly
who failed to meet diagnostic criteria for either condition—as well as prevalent constructs acknowledged by those with a BPD, but estab-
those who met criteria for both conditions—were excluded. Table 1 de- lished that they also had high prevalence rates in those with a BP con-
tails basic sociodemographic characteristics of study participants. A to- dition, a finding that goes some way to explain why diagnostic
tal of 53 participants received a DSM diagnosis of BPD and 83 a BP differentiation of the two conditions can be difficult.
diagnosis (4 with BP I and 79 with BP II). Females were overrepre- Of the 113 items, 38 showed significant differences in preva-
sented in assignment to BPD versus BP disorder (χ2 = 12.63; lence rates between the BPD and BP groups (p < 0.01), as detailed in
p < 0.001), and there were a greater proportion of participants unem- Table 3. Participants with BPD returned higher scores than the subjects
ployed or receiving sickness benefits in the BPD compared with the with BP on all but two items. The most differentiating item—“I tend to
BP group. idealize others (i.e., put them on a pedestal) but then often seek to hurt
The 20 most highly endorsed items by those with BPD (preva- them back if I judge them as hurtful to me”—weighted a characteristic
lence rates >95%) are listed in Table 2 (with the corresponding preva- generally judged as integral to those with a BPD but rarely reported by
lence rates returned by those with a BP disorder also listed). Three of patients with BP. The next two most differentiating items related to self-
the items were endorsed by 100% of the subjects with BPD (“I am harm—“There have been times where I have harmed myself ” and
tough on myself,” “I have always been self-critical,” and “I tend to take “When I feel irritable or angry, I sometimes hurt myself to relieve
things too personally”). However, the former two items were also stress.” The fourth most differentiating item pertained to stress-related

© 2019 Wolters Kluwer Health, Inc. All rights reserved. www.jonmd.com 621

Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.

Bayes and Parker The Journal of Nervous and Mental Disease • Volume 207, Number 8, August 2019

TABLE 2. Cognitive and Behavioral Items Endorsed By More Than 95% of Those With BPD

Prevalence
Item BPD, % BP, %
“I am tough on myself ” 100 100
“I've always been very self-critical” 100 100
“I tend to take things too personally” 100 90.1
“If someone is critical of something I do, I feel devastated” 98.1 90.2
“If someone upsets me, I am not able to easily put it out of my mind” 98.1 98.8
“I have periods where I'm extremely productive at work and other times when I'm not” 98.1 95.1
“I worry over minor things” 98.1 86.6
“My value as a person depends enormously on what others think of me” 98.1 71.6
“I tend to worry more than most people, particularly when under stress” 98.1 87.7
“I stew over things” 98.1 96.3
“I do not know who I am really in terms of my identity” 98.1 68.3
“I worry about being criticized for things I have said or done” 98.1 87.7
“I feel a need to make people feel happy” 98.1 96.3
“I can never be really sure if someone approves of me” 98.1 84.1
“I often feel that I have no idea or who I am or any clear identity” 96.2 73.2
“I always notice if someone doesn't respond to me” 96.2 96.3
“Other people tend to have a major impact on my mood” 96.2 90.2
“I worry about the effect I have on other people” 96.2 93.9
“I fear that my feelings will overwhelm people” 96.2 85.2
“If other people knew what I am really like, they would think less of me” 96.2 77.8

paranoid or interpersonal sensitivity ideation, “During times of stress, I weighted toward a BPD style, some items may have applicability to
often feel that others are deliberately mean to me,” followed by, “I tend other personality disorders—a limitation of the categorical system of
to have suicidal thoughts during and after a break-up or being rejected classifying personality disorders, such that meeting criteria for one per-
by someone.” Of the remaining significantly discriminating items, sonality disorder frequently means that patients meet criteria for other
those offering the least level of discrimination included “I can never personality disorders. Likewise, certain items may also be applicable
be really sure if someone approves of me” and “I tend to take things to other psychiatric diagnoses, for example, a number of BPD con-
too personally.” Of the 38 discriminating items, 10 related to self- structs have features in common with depression (e.g., suicidality,
identity, 7 to idealization/devaluation or unstable relationships, 6 each DSH) and anxiety (e.g., abandonment fears). A key study strength is
to inappropriate anger and abandonment fears, 4 to DSH and the preponderance of participants with BP II in the sample, which gen-
suicidality, 2 to ED, and 1 to emptiness. Of the 38 items, 6 were also en- erated a more diagnostically challenging and relevant bipolar group for
dorsed by more than 95% of participants with BPD, 5 of which pertain identifying items distinguishing the two conditions. Future studies
to identity disturbance. could use more than one rater to quantify the extent of interrater vari-
We then undertook analyses examining the degree to which both ability on the structured measure, as well as validate the refined item
the 10 and 20 most highly differentiating individual items distinguished set in a new patient sample.
the BPD from the BP participant groups as a set. Receiver operator Two of the three items affirmed by all subjects with BPD
characteristic analyses were undertaken and the Youden's index was pertained to self-criticism, in keeping with a core feature of BPD being
calculated (sensitivity + specificity – 1) for the 10-item set, which es- a “noxious sense of self ” (Meares et al., 2011) and with the third item
tablished a cutoff score of 5 or more indicating a probable BPD diagno- relating to ED. Overall, most of the items endorsed totally or near totally
sis and with this score having a sensitivity coefficient of 0.88 and a by participants with BPD were weighted to identity disturbance and to
specificity coefficient of 0.81, whereas the cutoff score for the 20- ED, with the latter considered a transdiagnostic feature also occurring in
item set was 10 or more and which generated sensitivity coefficients BP disorder (Yen et al., 2015). The other core BPD DSM domains—
of 0.94 and 0.73, respectively. including abandonment fears, self-harm, anger, stress-related paranoid
ideation, impulsivity, and relationship difficulties—were not so highly
endorsed. Intriguingly, one of the highly prevalent items in the BPD
DISCUSSION group was a “bipolar” item (“I have periods where I'm extremely pro-
Limitations of the current study include participants being vol- ductive at work and other times when I'm not”), which was endorsed
unteers and thus likely weighted toward having less severe disorders. more highly (though not statistically significant) by the BP group—
Participants may have biased their symptom reporting in line with their with such a high BPD endorsement possibly relating to the item tapping
prior diagnoses. Potential clinician rater biases include diagnostic ex- productivity changes secondary to ED rather than bipolar-related mood
pectation related to the differing recruitment portals and potential for episodes. Of note, in the highly endorsed items by the participants with
measurement error as there was only a single rater for the majority of BPD, there was also a high level of endorsement of many of those items
structured DSM diagnoses. Females were overrepresented in the BPD by those with a BP, indicating again that many of these personality and
group, a potential confounding factor if any sex-specific personality cognitive items are not specific to BPD. One explanation may be that,
features are overrepresented in women. Despite our measure being although the current study did exclude those with comorbid BPD and

622 www.jonmd.com © 2019 Wolters Kluwer Health, Inc. All rights reserved.

Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.

The Journal of Nervous and Mental Disease • Volume 207, Number 8, August 2019 Differentiation of BPD and BP

TABLE 3. Statistically Significant* Items Discriminating BPD From BP

Prevalence, %
Item BPD BP
“I tend to idealize others (i.e., put them on a pedestal) but then often seek to hurt them back if I judge them as hurtful to me” 75.5 32.9
“There have been times where I have harmed myself ” 84.9 45.1
“When I feel irritable or angry, I sometimes hurt myself to relieve stress” 75 38.3
“During times of stress, I often feel that others are deliberately mean to me” 82.7 47.6
“I tend to have suicidal thoughts during and after a break-up or being rejected by someone” 86.8 53.8
“If someone I care about is late to meet me, I immediately start worrying” 92.3 59.8
“I often feel hatred towards someone I care about and need” 76.9 46.9
“I do not know who I am really in terms of my identity” 98.1 68.3
“My relationships tend to be short-lived” 76.5 47.6
“I believe I have more difficulties with relationships than the average person my age” 92.2 63.4
“During times of stress, I often try to hurt or criticize others before they can hurt me” 60.4 31.7
“My value as a person depends enormously on what others think of me” 98.1 71.6
“I have intentionally harmed myself, for instance by cutting myself or taking too many pills” 82.4 56.1
“I often feel like the victim in many situations” 88.5 63.4
“I often feel that I have no idea of who I actually am” 98.1 74.1
“I feel the need to always be in a relationship” 73.1 49.4
“I can only believe that something I have done is good when someone tells me” 94 70.7
“I avoid saying what I think for fear of being rejected” 94.2 70.7
“I often feel that I have no idea or who I am or any clear identity” 96.2 73.2
“I've felt empty inside for as long as I can remember” 86.5 64.6
“When I get angry, I throw, break or punch things” 59.6 37.8
“I often feel very dependent on others but then need to get a lot of distance from them” 86.3 65.4
“If others knew the real me, they would not like me” 88.7 68.3
“I am very afraid of being alone” 83 63.4
“In situations when I've been extremely angry at someone, I will be quick to retaliate physically (i.e., punch, throw things, yell)” 61.5 42
“I am usually very reserved at parties and other social situations” 84.9 65.9
“If other people knew what I am really like, they would think less of me” 96.2 77.8
“I tend to get angry and lose my cool when stressed” 88.7 70.4
“I tend to get angry over things that most people wouldn't consider a big deal” 88.7 70.7
“I think a lot about being deserted by loved ones” 81.1 63.4
“I often profit at the expense of others, without feeling bothered by the pain I may cause them” 32.1 15.1
“I've taken revenge out on someone by damaging their property” 36.5 19.5
“I don't like people to really know me” 84.9 68.3
“I worry about how strong my important relationships really are” 92.5 76.8
“I can never be really sure if someone approves of me” 98.1 84.1
“I tend to take things too personally” 98.1 84.1
“I have distinct periods of time where my mood can crash following a high” 77.4 97.6
“I'm fairly happy with who I am most of the time” 62.3 86.4
*p < 0.01.

BP, it is possible that a proportion of the BP sample exhibited borderline BPD versus a BP diagnosis related to core features of DSH and
traits. Alternately, if personality traits are viewed dimensionally, then those suicidality. Overall, the majority of the most discriminating BPD items
with a BP disorder would be expected to lie somewhere on a borderline related to deficits in self-identity. Those with BPD generally experi-
continuum, and this may partially explain why there is frequent difficulty ence a disruption to their sense of self with associated “painful inco-
in distinguishing BPD from BP disorders. Furthermore, the current domi- herence,” that is, emotional pain as a result of a fragmented sense of
nant syndromal approach to psychiatric diagnosis, and the lack of “gold self (Meares et al., 2011). Those with BP, in contrast, tend to have a
standard” diagnoses based on underlying etiopathophysiology, means that more coherent sense of self, for example, those with BP II tend to expe-
symptom overlap is not uncommon, yet the underlying pathogenic mech- rience self-deficits only when depressed—and a grandiose self when
anisms giving rise to symptoms may eventually be found to differ. hypomanic—with stability of self-identity generally when euthymic
Turning to the most differentiating items, we established that (Renaud et al., 2012).
they did not correspond with the most prevalent BPD items, for exam- Highly discriminating items were also overrepresented by
ple, one of the items endorsed by the entire borderline group—“I tend to themes of idealization/devaluation and unstable relationships, suggest-
take things too personally”—was in fact one of the least discriminating ing that evaluating the capacity to have meaningful relationships may
items. Instead, three of the top five overrepresented items favoring a assist diagnosis. Those with BPD demonstrate immaturity in their views

© 2019 Wolters Kluwer Health, Inc. All rights reserved. www.jonmd.com 623

Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.

Bayes and Parker The Journal of Nervous and Mental Disease • Volume 207, Number 8, August 2019

of others and experience ongoing interpersonal conflicts (Kernberg and Bayes AJ, McClure G, Fletcher K, Román Ruiz del Moral YE, Hadzi-Pavlovic D,
Yeomans, 2013). By contrast, those with BP (when euthymic) are less Stevenson JL, Manicavasagar VL, Parker GB (2016a) Differentiating the bipolar dis-
likely to show pathological relationships and tend to maintain stable rela- orders from borderline personality disorder. Acta Psychiatr Scand. 133:187–195.
tionships (Kernberg and Yeomans, 2013). The overrepresentation of Bayes A, Parker G, Fletcher K (2014) Clinical differentiation of bipolar II disorder
abandonment fears as differentiating BPD from BP disorder is also in from borderline personality disorder. Curr Opin Psychiatry. 27:14–20.
keeping with those with BPD being sensitive to rejection (Gunderson, Bayes A, Parker G, McClure G (2016b) Emotional dysregulation in those with bipolar
2011) and which may precipitate anger—with a number of items disorder, borderline personality disorder and their comorbid expression. J Affect
pertaining to the latter found to offer discrimination from BP disorder. Disord. 204:103–111.
Previous studies have demonstrated affect shifts from euthymia to anger
de la Rosa I, Oquendo MA, García G, Stanley B, González-Pinto A, Liu SM, Blanco C
in BPD, in contrast to those individuals with BP disorder, who display (2017) Determining if borderline personality disorder and bipolar disorder are alter-
more affective lability from euthymia to depression or to elation, and native expressions of the same disorder: Results from the national epidemiologic
from elation to depression (Henry et al., 2001). survey on alcohol and related conditions. J Clin Psychiatry. 78:e994–e999.
ED was quantified as an overrepresented domain in the most
Ghaemi SN, Dalley S, Catania C, Barroilhet S (2014) Bipolar or borderline: A clinical
highly prevalent items in subjects with BPD and, in contrast to identity
overview. Acta Psychiatr Scand. 130:99–108.
disturbance, was not found to be as discriminating a feature. When we
specifically examined emotion regulation strategies in our previous re- Gunderson JG (2011) Borderline personality disorder. N Engl J Med. 364:2037–2042.
lated study (Bayes et al., 2016b), we did find those with BPD (com- Henry C, Mitropoulou V, New AS, Koenigsberg HW, Silverman J, Siever LJ (2001)
pared with BP) returned higher total scores on measures of ED; yet, Affective instability and impulsivity in borderline personality and bipolar II disor-
on a number of subscales, ED was equally severe in both groups. In ad- ders: Similarities and differences. J Psychiatr Res. 35:307–312.
dition, it has been found that compared with controls, those with BP Kernberg OF, Yeomans FE (2013) Borderline personality disorder, bipolar disorder,
display emotion regulation deficits even during periods of euthymia depression, attention deficit/hyperactivity disorder, and narcissistic personality
(van Rheenen et al., 2015). Thus, ED is likely not pathognomonic of disorder: Practical differential diagnosis. Bull Menninger Clin. 77:1–22.
BPD, as it can occur in BP disorder (Henry et al., 2001), with neurobi- Koenigsberg HW (2010) Affective instability: Toward an integration of neuroscience
ological studies suggesting that ED in those with a BP may be internally and psychological perspectives. J Pers Disord. 24:60–82.
driven, in contrast to the reactivity to social cues observed in those with
Mackinnon DF, Pies R (2006) Affective instability as rapid cycling: Theoretical
BPD (Koenigsberg, 2010). Overall, items pertaining to deficits in self-
and clinical implications for borderline personality and bipolar spectrum dis-
identity occurred with both a high and differential prevalence in BPD orders. Bipolar Disord. 8:1–14.
and offered superior discrimination.
Meares R, Gerull F, Stevenson J, Korner A (2011) Is self disturbance the core of bor-
derline personality disorder? An outcome study of borderline personality factors.
CONCLUSIONS Aust N Z J Psychiatry. 45:214–222.
Our findings of a high prevalence of many “borderline” person- Paris J, Black DW (2015) Borderline personality disorder and bipolar disorder: What is
ality items in the BP group offer an explanation as to why differentiation the difference and why does it matter? J Nerv Ment Dis. 203:3–7.
of the two conditions is often challenging. This central finding has two Paykel ES, Abbott R, Morriss R, Hayhurst H, Scott J (2006) Sub-syndromal and
nuances. First, in deriving diagnostic criteria for any diagnosis, if high syndromal symptoms in the longitudinal course of bipolar disorder. Br J Psychia-
prevalence alone of a criterion is adopted and there is failure to recog- try. 189:118–123.
nize that the feature may have high prevalence in other conditions Renaud S, Corbalan F, Beaulieu S (2012) Differential diagnosis of bipolar affective
(i.e., is transdiagnostic), then it has distinct limitations as a criterion. disorder type II and borderline personality disorder: Analysis of the affective di-
Second, if such transdiagnostic criteria were weighted by clinicians, mension. Compr Psychiatry. 53:952–961.
then it becomes clear as to why many then find differentiation of the
Ruggero CJ, Zimmerman M, Chelminski I, Young D (2010) Borderline personal-
two conditions difficult in a clinical setting. Our study highlights spe- ity disorder and the misdiagnosis of bipolar disorder. J Psychiatr Res. 44:
cific cognitive and behavioral features of BPD, namely, identity distur- 405–408.
bance and ED, as more prevalent than many of the other DSM domains.
ED, although found to offer some differentiating capacity, was less use- Van Rheenen TE, Murray G, Rossell SL (2015) Emotion regulation in bipolar dis-
order: Profile and utility in predicting trait mania and depression propensity.
ful as a distinguishing feature, possibly owing to its transdiagnostic sta-
Psychiatry Res. 225:425–432.
tus. Seemingly, identity disturbance is both an intrinsic feature of BPD
and offers specificity in diagnostic differentiation from BP. The study Vöhringer PA, Barroilhet SA, Alvear K, Medina S, Espinosa C, Alexandrovich K,
details BPD clinical features that offer minimal specificity and those Riumallo P, Leiva F, Hurtado ME, Cabrera J, Sullivan M, Holtzman N, Ghaemi
that offer superior differentiation from BP, and which may be used to SN (2016) The International Mood Network (IMN) Nosology Project: Differenti-
advance diagnostic precision of the two disorders. ating borderline personality from bipolar illness. Acta Psychiatr Scand. 134:
504–510.
Witt SH, Streit F, Jungkunz M, Frank J, Awasthi S, Reinbold CS, Treutlein J,
ACKNOWLEDGMENTS Degenhardt F, Forstner AJ, Heilmann-Heimbach S, Dietl L, Schwarze CE,
The authors thank Yolande Roman Ruiz del Moral and Kathryn Schendel D, Strohmaier J, Abdellaoui A, Adolfsson R, Air TM, Akil H, Alda
Fletcher for the use of the self-report questionnaire. M, Alliey-Rodriguez N, Andreassen OA, Babadjanova G, Bass NJ, Bauer M,
Baune BT, Bellivier F, Bergen S, Bethell A, Biernacka JM, Blackwood DHR,
Boks MP, Boomsma DI, Børglum AD, Borrmann-Hassenbach M, Brennan P,
DISCLOSURE Budde M, Buttenschøn HN, Byrne EM, Cervantes P, Clarke TK, Craddock N,
The authors declare no conflict of interest. Cruceanu C, Curtis D, Czerski PM, Dannlowski U, Davis T, de Geus EJC,
Di Florio A, Djurovic S, Domenici E, Edenberg HJ, Etain B, Fischer SB, Forty
REFERENCES L, Fraser C, Frye MA, Fullerton JM, Gade K, Gershon ES, Giegling I, Gordon
SD, Gordon-Smith K, Grabe HJ, Green EK, Greenwood TA, Grigoroiu-
American Psychiatric Association (Ed) (2013) Diagnostic and statistical manual of Serbanescu M, Guzman-Parra J, Hall LS, Hamshere M, Hauser J, Hautzinger
mental disorders (5th ed). Arlington, VA: American Psychiatric Publishing. M, Heilbronner U, Herms S, Hitturlingappa S, Hoffmann P, Holmans P, Hottenga
Bassett D (2012) Borderline personality disorder and bipolar affective disorder. Spec- JJ, Jamain S, Jones I, Jones LA, Juréus A, Kahn RS, Kammerer-Ciernioch J, Kirov
tra or spectre? A review. Aust N Z J Psychiatry. 46:327–339. G, Kittel-Schneider S, Kloiber S, Knott SV, Kogevinas M, Landén M, Leber M,

624 www.jonmd.com © 2019 Wolters Kluwer Health, Inc. All rights reserved.

Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.

The Journal of Nervous and Mental Disease • Volume 207, Number 8, August 2019 Differentiation of BPD and BP

Leboyer M, Li QS, Lissowska J, Lucae S, Martin NG, Mayoral-Cleries F, Ripke S, Mobascher A, Rujescu D, Lieb K, Roepke S, Schmahl C, Bohus M,
McElroy SL, McIntosh AM, McKay JD, McQuillin A, Medland SE, Middeldorp Rietschel M (2017) Genome-wide association study of borderline personality dis-
CM, Milaneschi Y, Mitchell PB, Montgomery GW, Morken G, Mors O, order reveals genetic overlap with the bipolar disorder, major depression and
Mühleisen TW, Müller-Myhsok B, Myers RM, Nievergelt CM, Nurnberger JI, schizophrenia. Transl Psychiatry. 7:e1155.
O'Donovan MC, Loohuis LMO, Ophoff R, Oruc L, Owen MJ, Paciga SA,
Yen S, Frazier E, Hower H, Weinstock LM, Topor DR, Hunt J, Goldstein TR, Gold-
Penninx BWJH, Perry A, Pfennig A, Potash JB, Preisig M, Reif A, Rivas F,
stein BI, Gill MK, Ryan ND, Strober M, Birmaher B, Keller MB (2015) Border-
Rouleau GA, Schofield PR, Schulze TG, Schwarz M, Scott L, Sinnamon GCB, Stahl
line personality disorder in transition age youth with bipolar disorder. Acta
EA, Strauss J, Turecki G, Van der Auwera S, Vedder H, Vincent JB, Willemsen G,
Psychiatr Scand. 132:270–280.
Witt CC, Wray NR, Xi HS; Bipolar Disorders Working Group of the Psychiatric
Genomics Consortium; Major Depressive Disorder Working Group of the Psychi- Zimmerman M, Chelminski I, Dalrymple K, Martin J (2018) Screening for bipolar dis-
atric Genomics Consortium; Schizophrenia Working Group of the Psychiatric Ge- order and finding borderline personality disorder: A replication and extension.
nomics Consortium, Tadic A, Dahmen N, Schott BH, Cichon S, Nöthen MM, J Pers Disord. 23:1–11.

© 2019 Wolters Kluwer Health, Inc. All rights reserved. www.jonmd.com 625

Copyright © 2019 Wolters Kluwer Health, Inc. All rights reserved.