borderline personality disorder in patients with bipolar disorder and response to lamotrigine
TRANSCRIPT
www.elsevier.com/locate/jad
Journal of Affective Disorders 79 (2004) 297–303
Preliminary communication
Borderline personality disorder in patients with bipolar disorder and
response to lamotrigine
Gilbert A. Preston*, Barrie K. Marchant, Fredrick W. Reimherr,Robert E. Strong, Dawson W. Hedges
Department of Psychiatry, Mood Disorders Clinic, University of Utah School of Medicine, Salt Lake City, UT, USA
Received 14 January 2002; accepted 13 August 2002
Abstract
Background: Recent reports suggesting lamotrigine as an effective treatment in bipolar disorder, and perhaps borderline
personality disorder, a common comorbid personality disorder in bipolar patients, led us to retrospectively examine patients
from two bipolar studies to investigate this pattern of comorbidity, and to determine whether lamotrigine effected the
dimensions of borderline personality. Methods: Fifteen months following entry into either study, we retrospectively assessed
DSM-IV dimensions of borderline personality disorder pre- and post-treatment with lamotrigine in 35 bipolar patients.
Results: Forty percent met criteria for borderline personality disorder; this subgroup had a more frequent history of substance
abuse and childhood symptoms of attention deficit hyperactivity disorder (ADHD). Dimensions of borderline personality
improved significantly with treatment in both patient groups, and corresponded with response of bipolar symptoms. Six
(43%) comorbid bipolar patients endorsed three or fewer criteria of borderline personality during treatment with lamotrigine.
There was a trend for comorbid bipolar patients to require a second psychoactive medication in addition to lamotrigine
during extended treatment. Limitations: Criteria for borderline personality and improvement were assessed retrospectively in
an open manner. Conclusions: Dimensions of borderline personality disorder may respond to lamotrigine in comorbid bipolar
patients; controlled studies appear warranted. Bipolar studies should assess and specify the number of patients with
personality disorders in the trial.
D 2002 Elsevier B.V. All rights reserved.
Keywords: Lamotrigine; Bipolar; Borderline personality disorder
1. Introduction
The presence of borderline personality disorder in
bipolar disorder is a significant risk factor for in-
0165-0327/$ - see front matter D 2002 Elsevier B.V. All rights reserved.
doi:10.1016/S0165-0327(02)00358-0
* Corresponding author. Clinical Brain Disorders Branch,
National Institutes of Mental Health, 10 Center Drive, Room 4S235,
MSC1379, Bethesda, MD 20892-1379, USA. Tel.: +1-301-402-
2861; fax: +1-301-480-7795.
E-mail address: [email protected] (G.A. Preston).
creased morbidity and a more extensive use of mental
health resources (Bender et al., 2001). An incidence of
comorbid personality disorder as high as 45% and
65% has been reported in bipolar patients, with
borderline personality the most prevalent, followed
by histrionic personality disorder (Dunayevich et al.,
1996; O’Connell et al., 1991). Trials of various
psychotropic medication have been shown to be of
inconsistent benefit in borderline personality disorder,
including atypical antipsychotics (Schulz et al., 1999;
G.A. Preston et al. / Journal of Affective Disorders 79 (2004) 297–303298
Szigethy and Schulz, 1997), typical antipsychotics
(Soloff et al., 1993), anticonvulsants (Hollander et
al., 2001; Calabrese et al., 2002) and antidepressants
(Hirschfeld, 1997).
Lamotrigine, the best studied anticonvulsant, has
recently been shown to be effective in the treatment
of refractory bipolar disorder, bipolar depression
(Calabrese et al., 2002, 1999; Bowden et al., 1999),
and rapid cycling bipolar type II (Yatham et al.,
2002).
In this preliminary study we retrospectively exam-
ined the incidence of borderline personality dimen-
sions in a cohort of patients who participated in two
multi-center studies of lamotrigine in the prevention
of relapse of bipolar mania or depression sponsored
by Glaxo–Wellcome to determine whether non-
comorbid bipolar and comorbid bipolar patients
would respond to treatment in the same manner. We
classified four dimensions of borderline personality
disorder—impulsivity, suicidality/self harm, affective
instability, and inability to control anger—as ‘affec-
tive dimensions’ in view of their clinical and opera-
tional similarities to DSM-IV symptom criteria of
depression and/or mania. The five remaining dimen-
sions we classified as ‘nonaffective dimensions’ of
borderline personality disorder. We were also inter-
ested in learning if specific affective or non-affective
dimensions of the personality disorder would respond
to lamotrigine, and if so, whether response was a
function of comorbidity.
Data from the multi-center study have been
reported (Bowden et al., 2001); lamotrigine showed
positive effects in preventing relapse of mania and
depression in bipolar I patients.
2. Methods
In the two original studies (GW605, GW606),
large multi-center clinical trials designed to evaluate
lamotrigine in the prevention of relapse in bipolar
disorder, subjects were required to meet DSM-IV
criteria for bipolar affective disorder, type I, and
either current major depression, or mania or mixed
state. Exclusion criteria included: rapid cycling,
recent history of alcohol or drug abuse, personality
disorder severe enough to interfere with the study
protocol, or significantly abnormal results on safety
evaluations (routine laboratory testing, thyroid pro-
file, ECG, urine toxicology screen and serum preg-
nancy tests). Information was collected regarding
the history of affective disorder and past psychiatric
symptoms including drug or alcohol abuse. The
Wender Utah Rating Scale (WURS) (Ward et al.,
1993) was administered to assess history of atten-
tion deficit hyperactivity disorder symptoms as a
child because it has been suggested that childhood
attention deficit hyperactivity disorder (ADHD)
symptoms may be a putative marker for early onset
bipolarity (Sachs et al., 2000). After a detailed
explanation of risks and procedures, patients signed
written informed consent for the original study. On
entry into the original studies, patients were initially
treated with lamotrigine in an open manner, thus
allowing all patients to experience initial exposure
to lamotrigine, and were continued on previously
prescribed medication. If they stabilized, previous
prescribed medications were withdrawn, and they
were observed on lamotrigine monotherapy. If they
remained stable, they were allowed to enter the
double-blind phase of the study. During this phase,
they were assigned in a double-blind manner to
lamotrigine, lithium, or placebo. Following comple-
tion of each study, patients were offered extended
treatment at no charge to insure that they were
stable prior to arranging long-term community care.
During the period of continued care, adjunctive
medication was employed when needed to maintain
mood stabilization.
All patients we could locate from the original
studies who were treated at the Mood Disorders
Clinic, Department of Psychiatry, and University of
Utah Health Sciences Center were re-interviewed.
Data reported here reflect only the Utah cohort. The
Institutional Review Board at the University of Utah
approved the study. Patients were re-interviewed at
the time of follow-up by one of the two authors not
involved in clinical care of the patients (GAP, BKM).
The Structured Clinical Interview for DSM-IV sched-
ule I (SCID-I) was used to reconfirm the diagnosis of
bipolar disorder, type I (American Psychiatric Asso-
ciation, 1994). The Structured Clinical Interview for
DSM-IV schedule II (SCID-II) was used for the
diagnosis of comorbid borderline personality disorder
(American Psychiatric Association, 1994). This inter-
view was conducted an average 15.3 months after
G.A. Preston et al. / Journal of Affective Disorders 79 (2004) 297–303 299
patients entered the original multi-center study.
Patients were not specifically evaluated for additional
Axis II disorders.
Subjects were diagnosed with comorbid borderline
personality disorder based on the self-reported histor-
ical presence of DSM-IV criteria for borderline per-
sonality disorder during euthymic periods. SCID-II
interview results were reviewed with the patient’s
treating psychiatrist. If the treating psychiatrist dis-
agreed with the SCID-II diagnosis, the patient was re-
interviewed by a second investigator who was told
there was a disagreement, but not the specific nature
of the disagreement. Following this second assess-
ment, the interviewers and treating psychiatrist
reached a diagnostic consensus in all cases. Medica-
tion status, dosage, and response was determined from
clinical records and patient interview and confirmed
by the treating psychiatrist.
In order to measure the response to medication, we
calculated scores for each borderline personality dis-
order dimension before and after treatment. We
assigned a score of one if a symptom was endorsed
at the threshold level, as defined by DSM-IV SCID-II.
A symptom that failed to reach the defined threshold
was scored zero. Response to treatment was scored
zero if a symptom was in complete remission, 0.5 if it
responded partially, and 1 if it failed to respond. The
cohort burden was the sum of the scores of nine DSM-
IV dimensions for borderline personality disorder
prior to treatment. The reduction in cohort burden
was the sum of response scores for nine dimensions
while on treatment subtracted from the initial cohort
burden. To decide whether a patient no longer met
criteria for the diagnosis of borderline personality
disorder, we modified the suggestion of Pinto and
Akiskal (1998) and required that the patient endorse
three or fewer DSM-IV borderline personality dimen-
sions following treatment.
In the original multi-center relapse prevention
studies, response was defined as having Clinical
Global Severity scores of 3 (mildly ill) or better for
4 weeks in a row. We classified as ‘bipolar respond-
ers’ patients whose mania or depression improved
during either of the two original studies or during
the period of extended treatment. Bipolar responders
who stabilized on lamotrigine alone were further
classified as ‘monotherapy responders’. Bipolar res-
ponders who stabilized on lamotrigine with adjunctive
lithium and/or antidepressant medication were classi-
fied as ‘combined therapy responders’. Patients who
failed to stabilize with lamotrigine monotherapy or
combined therapy were classified as ‘bipolar non-
responders’. Bipolar patients comorbid for borderline
personality disorder whose borderline dimensional
scores improved with pharmacotherapy were classi-
fied as ‘borderline responders’.
The Wilcoxon matched-pairs signed-ranks test
with Spearman correlation coefficient was used to
compare the cohort burden of the affective and non-
affective symptoms (prior to treatment) as well as
changes in mean borderline dimensional scores pre-
and post-treatment. Fisher’s exact test was used to
compare categorical variables. Continuous variables
were compared using paired sample or independent
sample t-tests, as appropriate. All statistics were
calculated with the SPSS version 6.1.4 statistical
package.
3. Results
We were able to locate 37 of 56 patients (66%). At
the follow-up interview, it was reconfirmed that all of
the patients fulfilled DSM-IV criteria for bipolar
disorder, type I. Two patients were unable to complete
the retrospective interview; both were excluded from
analysis of the outcome data leaving 35 subjects. We
compared reassessed patients with those lost to fol-
low-up on demographic, historic, and severity meas-
ures. Twelve of the patients (63%) excluded from the
analysis had a history of more than two serious
suicide attempts versus 11 of patients (32%) whom
we were able to locate. The difference was statistically
significant (Fisher exact test P < 0.05). This important
psychiatric symptom was not related to improvement
of the affective disorder in the original studies or
improvement of borderline dimensions in this study.
There were no other statistically significant demo-
graphic, historical or clinical differences between
patients in the study and those excluded from the
analysis.
Fourteen of 35 bipolar patients (40%) endorsed
symptoms in five or more dimensions of borderline
personality disorder and were retrospectively diag-
nosed as having both bipolar disorder and comorbid
borderline personality disorder. Table 1 compares
Table 1
Historical and severity measures of 21 bipolar and 14 comorbid bipolar patients expressed as either mean (S.D.) or as a percent of patients
exhibiting the attribute
Variable Bipolar Comorbid bipolar ta df P
Age at entry into study 34.9(9.7) 40.9(11.8) 1.57 33 0.11
At onset of affective symptoms 15.3(5.7) 20.5(9.4) 2.05 33 0.05
Number of mood episodes 27.0(15.8) 27.6(18.1) 0.11 32 0.91
Entry mania rating (n= 10) 25.8(4.8) 26.5(4.9) 0.20 8 0.85
Entry HAMD-31 (n= 24) 36.9(5.9) 39.1(3.3) 1.11 23 0.28
Entry CGI-S 4.4(0.5) 4.6(0.5) 1.10 33 0.28
Entry GAF 48.1(10.7) 49.1(3.3) 0.36 33 0.73
WURS (child) 37.4(19.4) 54.4(25.2) 2.07 28 0.05
WURS (adult) 46.9(13.7) 53.9(15.9) 1.27 27 0.21
Percent of patients exhibiting attribute Fisher’s exact test
Entered depressed 62% 86% 0.25
Male 52% 14% 0.03
First degree relative with mood disorder 95% 79% 0.28
History of suicide/self harm 43% 50% 0.74
History of psychosis 52% 29% 0.30
Bipolar responder 62% 71% 0.47
Monotherapy responders 77% 40% 0.09
History of alcohol abuse 33% 36% 0.99
History of substance abuse 33% 77% 0.02
Abbreviations: HAMD-31, 31-item Hamilton Psychiatric Rating Scale for Depression; CGI-S, Clinical Global Severity Scale; GAF, Global
Assessment of Functioning; WURS, Wender Utah Rating Scale.a Independent samples t-test.
G.A. Preston et al. / Journal of Affective Disorders 79 (2004) 297–303300
demographic, historical and severity measures for
bipolar and comorbid bipolar patients. The two groups
were similar on measures of symptom severity but
they differed in other ways. The bipolar patients were
evenly split between males and females, but the
comorbid bipolar patients were more frequently fe-
male (Fisher’s exact test, P= 0.05). Compared to the
bipolar patients, comorbid bipolar patients endorsed
significantly more childhood ADHD symptoms as
measured by WURS scores (t = 2.07 df 28, P= 0.05)
even though the comorbid bipolar patients were
predominantly female. They were also more likely
to have a history of drug abuse (Fisher’s exact test,
P= 0.02). There was no difference in the frequency of
alcohol abuse.
Table 2 displays the percentage of bipolar and
comorbid bipolar patients that endorsed each of the
nine dimensions for borderline personality disorder.
As expected, the comorbid group endorsed affective
and non-affective borderline dimensions at a higher
rate than did the bipolar group. Bipolar patients
endorsed affective borderline dimensions at a higher
rate than they did non-affective dimensions (Z = 2.04,
P < 0.05). Five bipolar patients endorsed four bor-
derline personality criteria, and six endorsed three
criteria.
Table 2 displays the change in borderline dimen-
sional scores in response to treatment. Treatment
resulted in a reduction of symptom burden in all
dimensions of borderline personality disorder at
levels that were both clinically and statistically
significant. While it appeared that both groups
showed more improvement in the affective dimen-
sions than the non-affective ones, this difference was
not statistically significant. With treatment, the av-
erage decrease in borderline personality disorder
symptom burden in comorbid bipolar patients was
45% (pre-treatment mean = 7.0F 1.2; post-treatment
mean = 3.9F 2.4); the average decrease in borderline
personality symptom burden for bipolar patients was
38% (pre-treatment mean = 2.5F 1.7; post-treatment
mean = 1.5F 1.3). Bipolar symptoms in both groups
of patients improved with treatment at the same rate:
62% of the bipolar patients versus 71% for the
comorbid bipolar patients. In the bipolar group, ten
of 21 patients (48%) were monotherapy responders.
Table 2
Endorsement and response of the dimensions of borderline personality disorder to lamotrigine by 21 bipolar and 14 comorbid patients
Endorsement by Dx Fisher’s Response by dimension Response by diagnosis
Bipolar Comorbidexact test
Pre Post % Changea Z scoreb P Bipolar Comorbid
Non-affective dimensions
I. Abandonment 19% 57% 0.05 12 8 33% 2.27 0.05 13%(4) 44%(8)
II. Unstable relationships 19% 79% 0.01 15 9.5 37% 2.43 0.05 50%(4) 32%(11)
III. Unstable self image 19% 71% 0.005 14 10.5 25% 2.65 0.01 13%(4) 30%(10)
VII. Chronic Emptiness 19% 93% 0.001 17 9.5 44% 3.03 0.005 50%(4) 42%(13)
IX. Dissociation/Paranoia 29% 79% 0.0059 17 11 35% 2.58 0.01 33%(6) 36%(11)
Affective dimensions
IV. Impulsiveness 48% 86% 0.07 22 11.5 48% 3.39 0.001 30%(10) 63%(12)
V. Self harm/suicidality 19% 57% 0.05 12 5 58% 2.65 0.01 50%(4) 63%(8)
VI. Affective instability 48% 100% 0.001 24 14 42% 3.40 0.001 40%(10) 43%(14)
VIII. Anger 29% 79% 0.0059 17 7.5 56% 3.27 0.005 58%(6) 55%(11)
a Percent improvement in cohort burden (number of patients who initially experienced the symptom).b Wilcoxon matched-pairs signed-rank test.
G.A. Preston et al. / Journal of Affective Disorders 79 (2004) 297–303 301
In the comorbid bipolar group only four of 14
patients (29%) were monotherapy responders. This
difference is not significant (Fisher Exact test
P= 0.31), but suggests that bipolar patients may be
more responsive to lamotrigine as a monotherapy.
Improvement in borderline dimensions was directly
related to improvement in bipolar illness and was
not a function of comorbidity. Bipolar responders
and bipolar nonresponders were essentially identical
prior to treatment with regard to the number of
borderline dimensions endorsed (mean 4.2 and 4.1,
respectively). With treatment, the two groups dif-
fered, endorsing an average of 1.7 and 3.5 dimen-
sions, respectively. This difference was significant
(Wilcoxon signed-ranks test at Z = 2.23, P < 0.05).
Three patients had no borderline symptoms initially
and couldn’t be categorized. Six comorbid patients
(42%) endorsed three or fewer borderline personality
disorder criteria following treatment, and no longer
met DSM-IV criteria for the personality disorder. All
six were also bipolar responders. Two (33.3%) were
monotherapy responders, and four (66.6%) were
combined therapy responders. This difference
approached significance (Fisher exact test P= 0.054).
4. Discussion
Forty-one percent of the patients in this study were
diagnosed with comorbid borderline personality dis-
order. This result corresponds with that of O’Connell
et al. (1991); studies with less structured interviews
(Charney et al., 1981; Gavaria et al., 1982) report
lower incidences of comorbidity.
The demographic, historic and severity measures
of our cohort are comparable to bipolar I samples
collected by Schulz et al. (1999) and McElroy et al.
(2001). Our comorbid bipolar patients’ average en-
dorsement of more than seven dimensions of border-
line personality disorder is comparable to that of Pinto
and Akiskal (1998), who also reported lamotrigine
effective in the treatment of borderline personality
disorder, suggesting that our sample is similar to those
studies. However, we cannot rule out the possibility
that our exclusion criteria (rapid cycling, recent alco-
hol abuse or dependency, and personality disorder
severe enough to interfere with the study protocol)
may have introduced a selection bias. Our failure to
locate more than one-third of patients for this end of
study follow-up should also be taken into consider-
ation in light of their significant differences in suici-
dality. However, we found no evidence to indicate that
past suicide attempts were related to response of either
bipolar symptoms or borderline dimensions. More-
over, our clinical measures (CGI-S, HAMD-31, Ma-
nia Rating, and GAF) indicated equivalent impairment
in our patient population and those we could not
locate.
The dimensions of borderline personality disorder
were reduced by about 40% during lamotrigine
G.A. Preston et al. / Journal of Affective Disorders 79 (2004) 297–303302
treatment, without significant differences in response
between affective and non-affective dimensions.
Others have reported similar findings. Frankenburg
and Zanerini (2002) studied divalproex sodium treat-
ment of 30 women with comorbid borderline per-
sonality disorder and bipolar II disorder (the more
commonly reported setting for bipolar and borderline
personality disorder comorbidity) divalproex was
superior to placebo in diminishing irritability, anger,
the tempestuousness of their relationships, and im-
pulsive aggressiveness.
The ADHD symptoms reported by our comorbid
patients may be markers of early onset and severity in
bipolar disorder (Sachs et al., 2000), but might also
have contributed to severe perturbations of psychoso-
cial and interpersonal development that have been
shown to be markers of vulnerability to, and subse-
quent severity of, affective illness (Post et al., 2001).
Finally, the high incidence of affective disorder (79–
95%) in first degree relatives of our patients may also
play an etiopathogenic role in the severity, age of
onset, and borderline features we report (Post et al.,
2001). We cannot exclude the possibility that some of
our bipolar patients were retrospectively endorsing
state-based symptoms that cleared with treatment
despite their assertion that these criteria were present
when they were in a euthymic state. Alternatively, it
may be the case that our patients have only one
disease, bipolar I disorder, which, as described by
Judd et al. (2002) can be characterized as a severe,
chronic disorder of idiosyncratically cycling affective
states of fluctuating severity and polarity; however,
our retrospective review of these patients based on an
extensive collection of historical data and information
from extended contact with them over a period of 12–
24 months support the diagnostic conclusions drawn
from our research interviews. In addition, the obser-
vation that there were a number of characteristics
separating the comorbid bipolar patients suggests we
correctly identified a distinct group of bipolar patients.
Improvement in affective and nonaffective symp-
toms with lamotrigine with or without adjunctive
lithium and/or antidepressant in this subset of comor-
bid bipolar I patients is significant in light of the
widely acknowledged resistance to treatment of bor-
derline personality disorder, and its negative effect on
the treatment response of comorbid bipolar disorder.
Based on this study and those of others, comorbid
personality disorder is a consistent and significant
phenomenon in bipolar disorder. A larger, prospec-
tive, long-term, controlled trial of lamotrigine for the
treatment of comorbid borderline personality and
bipolar disorder is warranted.
Acknowledgements
Support for this study was provided in part by
Glaxo–Wellcome. The authors would like to thank
Lorna S. Benjamin, Ph.D., Professor, Department of
Psychology, University of Utah, for reviewing and
critiquing the manuscript.
References
American Psychiatric Association, 1994. Diagnostic and Statistical
Manual of Mental Disorders, 4th Edition. American Psychiatric
Association, Washington, DC.
Bender, D.S., Dolan, R.T., Skodol, A.E., Sanislow, C.A., Dyck,
I.R., McGlashan, T.H., Shea, M.T., Zanarini, M.C., Oldham,
J.M., Gunderson, J.G., 2001. Treatment utilization by patients
with personality disorders. Am. J. Psychiatry 158, 295–302.
Bowden, C.L., Calabrese, J.R., McElroy, S.L., Rhodes, L.J., Keck
Jr., P.E., Cookson, J., Anderson, J., Bolden-Watson, C., Ascher,
J., Monaghan, E., Zhou, J., 1999. The efficacy of lamotrigine in
rapid cycling and non-rapid cycling patients with bipolar disor-
der. Biol. Psychiatry 45, 953–958.
Bowden, C.L., Calabrese, J.R., Bari, M., Feiger, A., Khan, A.,
Rapaport, M., Ascher, J.A., Spaulding, T., 2001. Meta-analysis
of two large placebo-controlled 18-month trials of lamotrigine
and lithium maintenance treatment in bipolar I disorder. Poster
presentation. In: 42nd Annual NCDEU 2001.
Calabrese, J.R., Bowden, C.L., Sachs, G.S., Ascher, J.A., Mona-
ghan, E., Rudd, G.D., 1999. A double blind placebo controlled
study of lamotrigine monotherapy in outpatients with bipolar 1
depression. J. Clin. Psychiatry 60, 79–88.
Calabrese, J.R., Shelton, M.D., Rapport, D.J., Kimmel, S.E., 2002.
Bipolar disorders and the effectiveness of novel anticonvulsants.
J. Clin. Psychiatry. 63 (Suppl. 3), 5–9.
Charney, D.S., Nelson, J.C., Quinlan, D.M., 1981. Personality
traits and disorder in depression. Am. J. Psychiatry 138,
1601–1604.
Dunayevich, E., Strakowski, S.M., Sax, K.W., Sorter, M.T., Keck
Jr., P.E., McElroy, S.L., McConville, B.J., 1996. Personality
disorders in first- and multiple-episode mania. Psychiatry Res.
64, 69–75.
Frankenburg, F.R., Zanerini, M.C., 2002. Divalproex sodium treat-
ment of women with borderline personality disorder and bipolar
II disorder: a double-blind placebo-controlled pilot study.
J. Clin. Psychiatry 63, 442–446.
Gavaria, M., Flaherty, J.A., Val, E., 1982. A comparison of bipolar
G.A. Preston et al. / Journal of Affective Disorders 79 (2004) 297–303 303
patients with and without a borderline personality disorder. Psy-
chiatr. J. Univ. Ottawa 7, 190–195.
Hirschfeld, R.M., 1997. Pharmacotherapy of borderline personality
disorder. J. Clin. Psychiatry 58, 48–52.
Hollander, E., Allen, A., Lopez, R.P., Bienstock, C.A., Grossman,
R., Siever, L.J., Merkatz, L., Stein, D.J., 2001. A preliminary
double-blind, placebo-controlled trial of divalproex sodium in
borderline personality disorder. J. Clin. Psychiatry 62, 199–203.
Judd, L.L., Akiskal, H.S., Schettler, P.J., Endicott, J., Maser, J.,
Solomon, D.A., Leon, A.C., Rice, J.A., Keller, M.B., 2002.
The long-term natural history of the weekly symptomatic
status of bipolar I disorder. Arch. Gen. Psychiatry 59 (6),
530–537.
McElroy, S.L., Altshuler, L.L., Suppes, T., Keck, P.E., Frye, M.A.,
Denicoff, K.D., Nolen, W.A., Kupka, R.W., Leverich, G.S.,
Rochussen, J.R., Rush, J.A., Post, R.M., 2001. Axis I psychi-
atric comorbidity and its relationship to historical illness varia-
bles in 288 patients with bipolar disorder. Am. J. Psychiatry
158, 420–426.
O’Connell, R.A., Mayo, J.A., Sciutto, M.S., 1991. PDQ-R person-
ality disorders in bipolar patients. J. Affect. Disord. 23, 217–221.
Pinto, O.C., Akiskal, H.S., 1998. Lamotrigine as a promising ap-
proach to borderline personality: an open case series without
concurrent DSM-IV major mood disorder. J. Affect. Disord.
51, 333–343.
Post, R.M., Leverich, G.S., Xing, G., Weiss, R.B., 2001. Develop-
mental vulnerabilities to the onset and course of bipolar disor-
der. Dev. Psychopathol. 13, 581–598.
Sachs, G.S., Baldassano, C.F., Truman, C.J., Guille, C., 2000.
Comorbidity of attention deficit hyperactivity disorder with
early- and late-onset bipolar disorder. Am. J. Psychiatry 157,
466–468.
Schulz, C.S., Camlin, K.L., Berry, S.A., Jesberger, J.A., 1999.
Olanzapine safety and efficacy in patients with borderline per-
sonality disorder and co-morbid dysthymia. Biol. Psychiatry 46,
1429–1435.
Soloff, P.H., Perel, J.M., Ulrich, R.F., 1993. Continuation pharma-
cotherapy of borderline personality disorder with haloperidol
and phenelzine. Am. J. Psychiatry 150, 1843–1848.
Szigethy, E.M., Schulz, S.C., 1997. Risperidone in comorbid bor-
derline personality disorder and dysthymia. J. Clin. Psychophar-
macol. 17, 326–327.
Ward, M.F., Wender, P.H., Reimherr, F.W., 1993. The Wender Utah
Rating Scale: an aid in the retrospective diagnosis of childhood
attention deficit hyperactivity disorder. Am. J. Psychiatry 150,
885–890.
Yatham, L.N., Kusumakar, V., Calabrese, J.R., Rao, R., Scarrow,
G., Kroeker, G., 2002. Third generation anticonvulsants in bi-
polar disorder: a review of efficacy and summary of clinical
recommendations. J. Clin. Psychiatry 63, 275–283.