BRIEF REPORTS
6. Spencer RL, Kim PJ, Kalman BA, Cole MA: Evidence for mineral-
ocorticoid receptor facilitation of glucocorticoid receptor-de-
pendent regulation of hypothalamic-pituitary-adrenal axis ac-
tivity. Endocrinology 1998; 27182726
7. Gesing A, Bilang-Bleuel A, Droste SK, Linthorst ACE, Holsboer F,
Reul JMHM: Psychological stress increases hippocampal miner-
alocorticoid receptor levels: involvement of corticotropin-re-
leasing hormone. J Neurosci 2001; 21:48224829
8. Young EA, Lopez JF, Murphy-Weinberg V, Watson SJ, Akil H: The
role of mineralocorticoid receptors in hypothalamic-pituitary-
Brief Report
Is Comorbidity of Posttraumatic Stress Disorder
and Borderline Personality Disorder
Related to Greater Pathology and Impairment?
Caron Zlotnick, Ph.D.
C. Laurel Franklin, Ph.D.
Mark Zimmerman, M.D.
Objective: The authors examined whether patients with co-
morbid borderline personality disorder and posttraumatic
stress disorder (PTSD) have a more severe clinical profile than
patients with either disorder without the other.
Method: Outpatients with borderline personality disorder
without PTSD (N=101), PTSD without borderline personality dis-
order (N=121), comorbid borderline personality disorder and
PTSD (N=48), and major depression without PTSD or borderline
P osttraumatic stress disorder (PTSD) and borderline per-
sonality disorder are DSM-IV diagnoses that are grounded
independently in well-articulated theoretical and em-
pirical frameworks. The more recent literature has sug-
gested a possible relationship between these disorders.
Comorbidity studies have shown a high frequency of
PTSD among patients with borderline personality disor-
der (56%) (1) and a high frequency of borderline personal-
ity disorder among patients with PTSD (68%) (2). Among
individuals in the community, it has been estimated that
one-third of those with borderline personality disorder
meet criteria for PTSD (3). In addition, it has been pro-
posed that both disorders share a common pathogenesis,
in that childhood sexual abuse has been found to have a
strong association with both borderline personality disor-
der and PTSD (4, 5).
1940
adrenal axis regulation in humans. J Clin Endocrinol Metab
1998; 83:33393345
9. Arvat E, Maccagno B, Giordano R, Pellegrino M, Broglio F, Gian-
otti L, Maccario M, Camanni F, Ghigo E: Mineralocorticoid re-
ceptor blockade by canrenoate increases both spontaneous
and stimulated adrenal function in humans. J Clin Endocrinol
Metab 2001; 86:31763181
10. Smythe JW, Murphy D, Timothy C, Costall B: Hippocampal min-
eralocorticoid, but not glucocorticoid, receptors modulate anx-
iety-like behavior in rats. Pharmacol Biochem Behav 1997; 56:
507513
personality disorder (N=469) were assessed with structured in-
terviews for psychiatric disorders and for degree of impairment.
Results: Outpatients with diagnoses of comorbid borderline
personality disorder and PTSD were not significantly different
from outpatients with borderline personality disorder without
PTSD, PTSD without borderline personality disorder, or major
depression without PTSD or borderline personality disorder in
severity of PTSD-related symptoms, borderline-related traits, or
impairment.
Conclusions: The additional diagnosis of PTSD or borderline
personality disorder does little to augment the pathology or
dysfunction of patients who have either disorder without the
other.
(Am J Psychiatry 2002; 159:19401943)
Trauma experts (e.g., Herman et al. [5]) have suggested
that borderline personality disorder may be a complex
variant of PTSD because many core borderline personality
disorder features, such as affective instability, anger, dis-
sociative symptoms, and impulsivity, are present in indi-
viduals with chronic PTSD. The concept that the effects of
trauma can lead to personality disturbances is reflected in
the provisional inclusion in the ICD-10 diagnostic system
of the diagnosis of enduring personality changes after cat-
astrophic stress.
A competing view is that the overlap between borderline
personality disorder and PTSD is often the result of a pre-
morbid borderline personality disorder, which increases
vulnerability to the deleterious effects of trauma exposure
(6). This view postulates that a distinction between the
symptoms of PTSD and the traits of borderline personality
disorder can be drawn on the basis of long-standing devel-
Am J Psychiatry 159:11, November 2002

opmental processes. Thus, the literature suggests that
PTSD and borderline personality disorder share common
features but that there remains some controversy regard-
ing the origins of these features. Additionally, there is lim-
ited knowledge regarding the impact of comorbid PTSD or
borderline personality disorder on the clinical presenta-
tion of either disorder, despite the presumed link between
borderline personality disorder and PTSD and the phe-
nomenological interface between these disorders.
To our knowledge, only one study to date (7) has exam-
ined whether patients with a dual diagnosis of PTSD and
borderline personality disorder represent a more impaired
group than those with the single diagnosis of PTSD. In this
study of female patients with PTSD and a history of child-
hood sexual abuse, no differences were found between
women with and without borderline personality disorder
in severity or frequency of PTSD symptoms, although the
women with both disorders scored higher on measures of
borderline traits (7). Limitations of this study include a
small sample size and the lack of a comparison group of
individuals with borderline personality disorder without
PTSD.
To address some gaps in the literature, we present here a
report from the Rhode Island Methods to Improve Diag-
nostic Assessment and Services project that examined
whether the presence of a comorbid disorder (PTSD or
borderline personality disorder) is associated with greater
pathology and dysfunction than PTSD without borderline
personality disorder or borderline personality disorder
without PTSD. More specifically, we compared four groups
of treatment-seeking outpatients (those with comorbid
PTSD and borderline personality disorder, PTSD without
borderline personality disorder, borderline personality
disorder without PTSD, and major depression without
PTSD or borderline personality disorder) along dimen-
sions that are considered markers of borderline personal-
ity disorder and/or PTSD, as well as several other indices
of impairment.
Method
Semistructured diagnostic interviews were administered to
1,500 outpatients seeking treatment at a general hospital private
practice. Exclusion criteria for the study were age less than 18
years, a history of a developmental disability, or difficulty with the
English language.
Trained diagnostic raters administered the Structured Clinical
Interview for DSM-IV (SCID) (8) and the Structured Interview for
DSM-IV Personality (9). All raters achieved adequate interrater re-
liability before independently administering the interviews (10).
The interrater kappa coefficient for 47 SCID-based PTSD diag-
noses was 0.91.
Disorders were considered current if patients met DSM-IV di-
agnostic criteria for the disorder at the evaluation. The hospitals
institutional review committee approved the research protocol,
and all patients provided written informed consent.
Ninety participants were not given the Structured Interview for
DSM-IV Personality and were excluded from the analyses. From
the remaining 1,410 participants, 739 met criteria for one of the
Am J Psychiatry 159:11, November 2002
BRIEF REPORTS
four targeted diagnostic groups and were included in the analy-
ses: comorbid current PTSD and borderline personality disorder
(N=48), borderline personality disorder without current PTSD
(N=101), current PTSD without borderline personality disorder
(N=121), and current major depression without current PTSD or
borderline personality disorder (N=469).
Besides comparing the four groups on severity or frequency of
PTSD symptoms and borderline personality disorder traits, we
also examined whether a specific cluster of borderline personality
disorder traits (affective instability, anger, dissociative symptoms,
and impulsivity) identified by experts as characteristic of both
disorders (5, 6) was more likely to be associated with borderline
personality disorder than PTSD.
The four diagnostic groups were compared along several mea-
sures of impairment. Patients were interviewed about the num-
ber of times they had been hospitalized for psychiatric reasons
and history of suicide attempts. Items from the Schedule for Af-
fective Disorders and Schizophrenia (11) were used to assess so-
cial and work impairment, including number of days out of work
due to emotional or psychiatric problems and social functioning
in the last 5 years.
Results
Most participants were women (N=497, 67.3%), white
(N=638, 86.3%), unmarried (N=448, 60.6%), and had com-
pleted high school (N=594, 80.4%). The patients mean age
was 37.1 (SD=12.1); mean age significantly differed among
the groups (F=20.18, df=3, 738, p<0.001) (Table 1). There
were no other significant demographic differences.
We analyzed our data as a two-by-two factorial design,
with presence versus absence of borderline personality
disorder and presence versus absence of PTSD as the
grouping factors. We performed two multivariate analyses
of covariance (MANCOVAs), covarying for age in both. In
one MANCOVA, the dependent variables were symptoms/
criteria of PTSD and borderline personality disorder traits,
and in the other the dependent measures were indices of
impairment.
The combined dependent variables (symptoms/criteria
of PTSD and borderline personality disorder traits) were
significantly affected by PTSD status (Wilkss lambda=
0.58, F=105.57, df=5, 730, p<0.001), borderline personality
disorder status (Wilkss lambda=0.37, F=245.96, df=5, 730,
p<0.001), and their interaction (Wilkss lambda=0.98, F=
2.35, df=5, 730, p=0.04, controlling for age).
With a Bonferroni correction of 0.008, between-subjects
analyses of variance (ANOVAs) were conducted. The
ANOVAs showed that no variables were significant for the
interaction. Total number of PTSD symptoms and symp-
toms in each criterion of PTSD were significant for the
PTSD without borderline personality disorder group, and
total number of PTSD symptoms, criterion C symptoms,
borderline personality disorder traits, and borderline per-
sonality disorder cluster traits were significant for the bor-
derline personality disorder without PTSD group (Table 1).
When the impairment indices were used as dependent
variables, controlling for age, the interaction was not sig-
nificant (Wilkss lambda=1.00, F=0.57, df=5, 728, p=0.73).
Significant main effects were found for the PTSD without
1941
BRIEF REPORTS

TABLE 1. Clinical Characteristics of 739 Patients With Diagnoses of Borderline Personality Disorder But Not Posttraumatic
Stress Disorder (PTSD), PTSD But Not Borderline Personality Disorder, PTSD and Borderline Personality Disorder, or Major
Depression But Not PTSD or Borderline Personality Disorder
Comorbid PTSD
Borderline and Borderline
Personality Disorder Personality Disorder Major Depression
Characteristic (N=101)a PTSD (N=121)b (N=48)c (N=469)d
Mean SD Mean SD Mean SD Mean SD
Age (years) 31.7 9.2 34.9 11.0 30.8 8.3 39.5 12.5
Number of DSM-IV symptoms
PTSD
Total 4.64e 4.74 11.74e 2.67 12.73 2.35 3.01 3.91
Criterion B 1.62 1.73 3.56e 1.20 3.67 1.15 1.00 1.43
Criterion C 1.64g 1.89 4.45e 1.22 5.06 1.21 1.05 1.55
Criterion D 1.38 1.70 3.73e 1.06 4.00 0.97 0.96 1.42
Borderline personality disorder
Total 6.13e 1.06 1.79 1.36 6.06 1.04 1.15 1.30
Specific clusterf 2.97e 0.81 1.01 0.98 3.02 0.70 0.62 0.80
Number of previous psychiatric hospitalizations 0.91g 1.41 0.66h 1.19 1.23 1.74 0.34 0.84
Number of suicide attempts 1.33g 2.62 0.82g 1.49 2.02 3.37 0.22 0.57
Impairment in last 5 yearsh
Social functioning 3.40g 1.36 3.24 1.25 3.79 1.27 3.15 1.25
Days out of work 3.01g 2.33 2.93g 2.31 3.94 2.93 2.09 1.70
Comorbid diagnosesi 4.28g 3.26 4.25g 1.89 5.17 1.65 2.79 2.55
a Patients with borderline personality disorder but not PTSD made up 13.7% of the study group.
b Patients with PTSD but not borderline personality disorder made up 16.4% of the study group.
c Patients with comorbid PTSD and borderline personality disorder made up 6.5% of the study group.
d Patients with major depression but not PTSD or borderline personality disorder made up 63.5% of the study group.
e Significant main effect in the multivariate analysis of covariance (MANCOVA), covarying for age (p<0.008).
f Total of four traits commonly associated with borderline personality disorder and PTSD: 1) inappropriate, intense anger or difficulty control-
ling anger, 2) impulsivity in at least two areas that are potentially self-damaging, 3) affective instability/marked reactivity of mood, and 4)
transient, stress-related paranoid ideation or severe dissociative symptoms.
g Significant main effect in the MANCOVA, covarying for age (p<0.01).
h Based on items from the Schedule for Affective Disorders and Schizophrenia.
i Total current axis I comorbid diagnoses (index diagnosis included).

borderline personality disorder group (Wilkss lambda=
0.93, F=10.50, df=5, 728, p<0.001) and the borderline per-
sonality disorder without PTSD group (Wilkss lambda=
0.88, F=19.64, df=5, 728, p<0.001). Between-subjects anal-
yses of covariance revealed that all variables except social
functioning were significant for both groups, with a Bon-
ferroni correction of 0.01 (Table 1).
A power analysis using an alpha level of 0.05 showed
that, for a study group of 192, the power of 1.0, as defined
by Cohen (12), was acceptable for a large effect to be
detected.
Discussion
The present report found that outpatients with comor-
bid PTSD and borderline personality disorder did not have
a more severe clinical profile than outpatients with PTSD
without borderline personality disorder or those with bor-
derline personality disorder without PTSD. More specifi-
cally, comorbidity of PTSD and borderline personality dis-
order was not significantly associated with greater or less
severity of PTSD symptoms, borderline personality disor-
der traits, symptoms in each of the PTSD clusters, border-
line personality disorder traits associated with PTSD and
borderline personality disorder, or overall impairment.
Our findings confirm and expand on previous research
showing that patients with comorbid PTSD and borderline
personality disorder have a similar pattern of PTSD symp-
toms as those with PTSD and no borderline personality
disorder (7). These results suggest that the additional diag-
nosis of PTSD or borderline personality disorder does little
to exacerbate existing pathology or dysfunction. Perhaps
there is little interaction between the clinical difficulties
associated with each disorder; treatment that targets one
disorder may not affect the other. Obviously, longitudinal,
prospective research will elucidate whether these two dis-
orders covary with one another.
The level of impairment found among the group of pa-
tients with comorbid PTSD and borderline personality
disorder may have been unaffected by either diagnosis in
this study because both disorders are chronic conditions
that have severe effects on impairment (13, 14), and,
therefore, the presence of a comorbid disorder did little to
augment the level of existing dysfunction.
Another finding of this report was that patients with bor-
derline personality disorder and PTSD reported a greater
number of PTSD symptoms and criterion C symptoms
(avoidant cluster) than PTSD patients without borderline
personality disorder. A possible explanation for these find-
ings is that PTSD has been conceptualized as a disorder of
affect dysregulation (15), a prominent area of dysfunction
among patients with borderline personality disorder (16).
The avoidant cluster of PTSD incorporates dimensions
that tap aspects of affect dysregulation and dissociation,

1942 Am J Psychiatry 159:11, November 2002


such as restrictions in affect and in relationships, which
play a principal role in borderline personality disorder.
In contrast, we found that PTSD patients without bor-
derline personality disorder did not have higher levels of
borderline personality disorder traits or traits related to
both PTSD and borderline personality disorder than pa-
tients without PTSD, which suggests that associated fea-
tures of PTSD may not include borderline traits, as previ-
ously postulated (5). Future research needs to examine the
etiological role of childhood sexual abuse in PTSD and
borderline personality disorder, since the negative se-
quelae of childhood sexual abuse are strongly associated
with borderline personality disorder and PTSD (4, 5) and
may account for the presence of PTSD symptoms among
patients with borderline personality disorder.
Received July 12, 2001; revisions received Nov. 7, 2001, and May
10, 2002; accepted June 3, 2002. From the Department of Psychiatry
and Human Behavior, Brown University and Butler Hospital; and the
Department of Psychiatry and Human Behavior, Rhode Island Hospi-
tal, Providence. Address reprint requests to Dr. Zlotnick, Butler Hos-
pital, 345 Blackstone Blvd., Providence, RI 02906; caron_zlotnick@
brown.edu (e-mail).
Supported in part by NIMH grants MH-56404 and MH-48732 (Dr.
Zimmerman).
References
1. Zanarini MC, Frankenburg FR, Dubo ED, Sickel AE, Trikha A,
Levin A, Reynolds V: Axis I comorbidity of borderline personal-
ity disorder. Am J Psychiatry 1998; 155:17331739
2. Shea MT, Zlotnick C, Weisberg RB: Commonality and specificity
of personality disorder profiles in subjects with trauma histo-
ries. J Personal Disord 1999; 13:199210
3. Swartz M, Blazer D, George L, Winfield I: Estimating the preva-
lence of borderline personality disorder in the community. J
Personal Disord 1990; 4:257272
4. Rowan AB, Foy DW, Rodriguez N, Ryan S: Posttraumatic stress
disorder in a clinical sample of adults sexually abused as chil-
dren. Child Abuse Negl 1994; 182:145150
Am J Psychiatry 159:11, November 2002
BRIEF REPORTS
5. Herman JL, van der Kolk BA: Traumatic antecedents of BPD, in
Psychological Trauma. Edited by van der Kolk BA. Washington,
DC, American Psychiatric Press, 1987, pp 111126
6. Gunderson JG, Sabo AN: The phenomenological and concep-
tual interface between borderline personality disorder and
PTSD. Am J Psychiatry 1993; 150:1927
7. Heffernan K, Cloitre M: A comparison of posttraumatic stress
disorder with and without borderline personality disorder
among women with a history of childhood sexual abuse: etio-
logical and clinical characteristics. J Nerv Ment Dis 2000; 188:
589595
8. First MB, Spitzer RL, Gibbon M, Williams JBW: Structured Clini-
cal Interview for DSM-IV Axis I Disorders (SCID). Washington,
DC, American Psychiatric Press, 1997
9. Pfohl B, Blum N, Zimmerman M: Structured Interview for DSM-
IV Personality: SIDP-IV. Washington, DC, American Psychiatric
Press, 1997
10. Zimmerman M, Mattia JI: Is posttraumatic stress disorder un-
derdiagnosed in routine clinical settings? J Nerv Ment Dis 1999;
187:420428
11. Endicott J, Spitzer RL: A diagnostic interview: the Schedule for
Affective Disorders and Schizophrenia. Arch Gen Psychiatry
1978; 35:837844
12. Cohen J: Statistical Power Analyses for the Behavioral Sciences.
Hillsdale, NJ, Lawrence Erlbaum Associates, 1988
13. Ballenger JC, Davidson JRT, Lecrubier Y, Nutt DJ, Foa E, Kessler
RC, McFarlane AC, Shalev AY: Consensus statement on post-
traumatic stress disorder from the International Consensus
Group on Depression and Anxiety. J Clin Psychiatry 2000; 61:
6065
14. Skodol AE, Gunderson JG, McGlashan TH, Dyck IR, Stout RL,
Bender DS, Grilo CM, Shea MT, Zanarini MC, Morey LC, Sanislow
CA, Oldham JM: Functional impairment in patients with schizo-
typal, borderline, avoidant, or obsessive-compulsive personal-
ity disorder. Am J Psychiatry 2002; 159:276283
15. Taylor GJ, Bagby RM, Parker JDA: Disorders of Affect Regula-
tion: Alexithymia in Medical and Psychiatric Illness. Cambridge,
UK, Cambridge University Press, 1997
16. Linehan MM, Kehrer CA: Borderline personality disorder, in
Clinical Handbook of Psychological Disorders, 2nd ed. Edited
by Barlow DH. New York, Guilford, 1993, pp 396441
1943