Journal of Psychiatric Research 35 (2001) 307–312

[Link]/locate/jpsychires

Affective instability and impulsivity in borderline personality and

bipolar II disorders: similarities and differences
Chantal Henrya,b, Vivian Mitropoulouc, Antonia S. Newc, Harold W. Koenigsbergc,
Jeremy Silvermanc, Larry J. Sieverc,*
a
Service Universitaire de Psychiatrie, Centre Hospitalier Charles Perrens, 121 rue de la Béchade, 33076 Bordeaux, France
b
INSERM U-394, Neurobiologie Intégrative, rue Camille Saint Saëns, 33077 Bordeaux, France
c
The Bronx VA Medical Center, 116A, 130 West Kingsbridge Road, Bronx, NY 10468, USA

Received 22 February 2001; received in revised form 11 June 2001; accepted 17 August 2001

Abstract
Objectives: many studies have reported a high degree of comorbidity between mood disorders, among which are bipolar dis-
orders, and borderline personality disorder and some studies have suggested that these disorders are co-transmitted in families.
However, few studies have compared personality traits between these disorders to determine whether there is a dimensional overlap
between the two diagnoses. The aim of this study was to compare impulsivity, affective lability and intensity in patients with bor-
derline personality and bipolar II disorder and in subjects with neither of these diagnoses. Methods: patients with borderline per-
sonality but without bipolar disorder (n=29), patients with bipolar II disorder without borderline personality but with other
personality disorders (n=14), patients with both borderline personality and bipolar II disorder (n=12), and patients with neither
borderline personality nor bipolar disorder but other personality disorders (OPD; n=93) were assessed using the Affective Lability
Scale (ALS), the Affect Intensity Measure (AIM), the Buss–Durkee Hostility Inventory (BDHI) and the Barratt Impulsiveness
Scale (BIS-7B). Results: borderline personality patients had significantly higher ALS total scores (P< 0.05) and bipolar II patients
tended to have higher ALS scores than patients with OPD (P <0.06). On one of the ALS subscales, the borderline patients dis-
played significant higher affective lability between euthymia and anger (P<0.002), whereas patients with bipolar II disorder dis-
played affective lability between euthymia and depression (P< 0.04), or elation (P< 0.01) or between depression and elation
(P <0.01). A significant interaction between borderline personality and bipolar II disorder was observed for lability between anxiety
and depression (P< 0.01) with the ALS. High scores for impulsiveness (BISTOT, P< 0.001) and hostility (BDHI, P< 0.05) were
obtained for borderline personality patients only and no significant interactions between diagnoses were observed. Only borderline
personality patients tended to have higher affective intensity (AIM, P< 0.07). Conclusions: borderline personality disorder and
bipolar II disorder appear to involve affective lability, which may account for the efficacy of mood stabilizers treatments in both
disorders. However, our results suggest that borderline personality disorder cannot be viewed as an attenuated group of affective
disorders. # 2001 Elsevier Science Ltd. All rights reserved.
Keywords: Borderline personality disorder; Bipolar disorder; Affective instability; Affective Intensity; Aggression; Impulsiveness

1. Introduction et al., 1983; Frances et al., 1984; Perry, 1985; Zanarini et

al., 1998). Some family studies have also shown comor-
Many studies have reported a high frequency of bidity between affective disorders including bipolar dis-
comorbidity between borderline personality disorder orders and borderline personality (Akiskal, 1981;
and major mood disorders (bipolar disorder and major Loranger et al., 1982; Amadeo et al., 1992). Moreover,
depression), from 35 to 51.5% (McGlashan, 1983; Pope in a clinical case conference Bolton and Gunderson
(1996) have discussed how it is sometime difficult to
establish the differential diagnosis between bipolar dis-
* Corresponding author. Tel.: +1-718-584-9000 ext. 5225; fax: +1-
order and borderline personality.
718-364-3576. Some authors have suggested that borderline person-
E-mail address: [Link]@[Link] (L.J. Siever). ality disorder is on the spectrum of affective disorders
0022-3956/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0022-3956(01)00038-3
308 C. Henry et al. / Journal of Psychiatric Research 35 (2001) 307–312

(Akiskal et al., 1981; McGlashan, 1983). This has led to We hypothesized that (1) both borderline personality
the following hypotheses: (1) borderline personality is a disorder and bipolar II patients would display higher
variant of affective disorder, (2) borderline personality levels of affective lability and intensity, as measured by
predisposes the patient to depression, or (3) the two the Affective Lability Scale (ALS) and the Affect Inten-
disorders have etiologic features in common (Gunder- sity Measure (AIM), than patients without these dis-
son and Phillips, 1991; Koenigsberg et al., 1999). Biolo- orders; (2) borderline personality patients, but not
gical and treatment studies have shown that borderline bipolar II patients would display a higher level of
personality and affective disorders display similarities impulsiveness compared to patients without these dis-
and differences in their underlying biology and treat- orders; (3) and bipolar II patients would display affec-
ment response (Koenigsberg et al., 1999). tive lability even between major affective episodes.
However, most studies assessing the relationship
between borderline personality and bipolar disorder
raise the question of comorbidity or describe the impact 2. Methods
of comorbid borderline personality on the features,
progression or treatment response of bipolar disorder. 2.1. Subjects
Bipolar patients with a comorbid borderline personality
have a less favorable outcome, with more suicide We studied 148 patients with personality disorders
attempts, earlier onset and poorer response to treatment from the Mood and Personality Disorder Program at
than those without borderline personality (Kutcher et the Mount Sinai Medical Center and the Bronx VA
al., 1990; Sato et al., 1999). Medical Center. All subjects were healthy outpatients at
To our knowledge, this is the first study to assess the time of the study, as evaluated by laboratory tests,
dimensional traits in these disorders, identified from a physical and neurological examination, and had not
personality disorder cohort, in an adequately large taken any medication including psychotropic medica-
sample, to determine phenomenologically the simila- tion, for at least 14 days.
rities and differences between these disorders. This Patients with bipolar I disorder, schizophrenia, and
dimensional approach may make it possible to elucidate other Axis I psychotic disorders were excluded from this
the possible links between these two disorders and to study sample; only bipolar II patients were included.
identify the biological features that they share. This The study was described in detail to the subjects and
approach is based on the model that there are psycho- informed written consent obtained.
biological dimensions such as affective instability, that Diagnoses were determined by one or more trained
predispose individuals and may cut across Axis I/II interviewers, using the Structured Interview for the
boundaries (Siever and Davis, 1991). Diagnosis of Personality Disorders for DSM-III-R
Affective instability and impulsive behavior appear to (SIPD-R, Stangl et al., 1989; kappa 0.81 for borderline
be appropriate traits for comparison in the two dis- personality disorder) and the Schedule for Affective
orders. Indeed, affective instability is explicitly included Disorders and Schizophrenia (SADS; Spitzer and End-
as a criterion for borderline personality, whereas icott, 1975; Kappa for BP II=0.72).
‘‘autonomic lability’’ including a high level of inter-
personal sensitivity and a low level of emotional stabi- 2.2. Dimensional assessments
lity has been suggested to be a valid antecedent of
unipolar depression and bipolar disorders (Clayton et Each subject completed the Affective Lability Scale
al., 1994; Lauer et al., 1997). In addition, borderline and (ALS, Harvey et al., 1989). This is a 54-item self-report
bipolar patients (more specifically bipolar II disorder) scale that measures changeable affects. ALS items assess
frequently display suicidal behavior. Moreover, poor subjects perception of their tendency to change between
impulse control may be a strong risk factor for recur- what they consider to be their normal (euthymic) mood
rent suicidal behavior, suggesting that impulsivity is a and affects of anger (ANG), depression (DEP), elation
trait that should be investigated in both disorders. (ELA), and anxiety (ANX), and their tendency to
Finally, it has been suggested that both affective oscillate between states of depression and elation (BIP),
instability and impulsive personality disorder traits and between states of anxiety and depression
rather than the personality disorder per se have an (ANXDEP). Each item is rated on a four-point scale
heritable component (Silverman et al., 1991; Torgersen, (scored 0–3 inclusive) from ‘‘very undescriptive’’ to
2000). Thus, affective instability and impulsiveness ‘‘very descriptive’’ of themselves. ALS total is the mean
appear to be appropriate dimensions for investigation in of the six subscales for individual affect shifts.
these disorders. We compared bipolar II disorder Subjects also completed the Affect Intensity
patients (rather than patients with bipolar I disorder) to Measure (AIM, Larsen and Diener, 1985; Larsen et
borderline personality patients, as these two disorders al., 1986) a self-report scale consisting in a 40-items.
share similar clinical presentations (Benazzi, 2000). Affect intensity refers to individual differences in the
 C. Henry et al. / Journal of Psychiatric Research 35 (2001) 307–312 309

intensity of response to a given level of emotion-pro- bipolar II disorder patients who did not meet the cri-
voking stimulation. teria for borderline personality disorder, 12 patients
Aggressiveness and impulsiveness were assessed using who met the criteria for both borderline personality
the Buss–Durkee Hostility Inventory (BDHI, Buss and disorder and bipolar II disorder and 93 patients who did
Durkee, 1957) and the Barratt Impulsiveness Scale not meet the criteria for either (but met criteria for other
(BIS-7B, Barratt et al., 1965), both of which are also personality disorders; see Table 1 for description of the
self-report scales. The BDHI provides a dimensional study population).
measure of lifetime history of aggression, and the sub- Age and sex distribution differed between groups, so
scale for irritability and assaultiveness includes compo- we adjusted the analysis for age and sex [borderline
nents related to anger, hostility, physical and verbal personality patients were younger than bipolar II sub-
aggression. It is widely used in the study of impulsive jects, and a higher proportion were female F(1,3)=2.7,
aggression and high scores for the scale as a whole and d.f.=148, P=0.1, Table 1.
for the composite subscale, irritability and assaultive- A two-way analysis of variance (with presence or
ness, have been shown to be associated with a blunted absence of borderline personality disorder and bipolar
response to the serotonin releasing agent fenfluramine II disorder as factors), adjusted for sex and age, for total
(Coccaro et al., 1989). The BIS characterizes four ALS score demonstrated a significant effect of border-
aspects of impulsiveness: speed of cognitive response, line personality disorder diagnosis [F(1,5)=4.3,
lack of impulse control, extroversion, and risk taking. d.f.=148, P < 0.04]; patients with borderline personality
had significantly higher ALS scores than patients with-
out this disorder. Patients meeting the criteria for bipo-
3. Statistical analysis lar II disorder tended to have higher ALS scores
compared to patients who did not meet criteria for
Statistical analysis was performed with SPSS version bipolar II disorder [F(1,5)=3.6, d.f.=148, P < 0.06].
9.0 computer software. As no other large studies have There was a tendency towards interaction between bor-
assessed borderline personality and bipolar II disorders derline personality and bipolar II disorder diagnosis,
together, we conducted a two-way analysis of covar- with patients meeting both diagnoses tending to have
iance with borderline personality disorder and bipolar the highest scores [F(1,5)=3.0, d.f.=148, P < 0.09]. This
disorder diagnoses (presence or absence) as factors, to model was significant after adjustement for current
investigate the contribution of each diagnosis separately depressive episode [F(1,4)=0.39, P < 0.05].
and to test for possible interaction effects between these As borderline personality diagnosis had a significant
diagnoses, after adjusting for age and sex. effect on total ALS score, we investigated whether there
was a difference in ALS subscore between the two
diagnoses. We investigated the subscale scores for
4. Results depression (DEP), anger (ANG), anxiety (ANX) and
elation (ELA), and for shifts between elation and
The sample consisted of 29 borderline patients who depression (BIP) and between anxiety and depression
did not meet the criteria for bipolar II disorder, 14 (ANXDEP).
Affective lability scores profiles differed between the
Table 1
two diagnostic groups (Table 2). Patients meeting the
Characteristics of the four groups of patients criteria for bipolar II disorder had significantly higher
scores for DEP, ELA and BIP [F(1,5)=4.3, d.f. 148,
Mean age Sex
P=0.04 for DEP; F(1,5)=6.8, d.f. 148, P=0.01 for
(S.D.) (% male)
ELA, F(1,5)=6.8, d.f. 148, P=0.01 for BIP]. In con-
Borderline no Bipolar II n=29 32.3 ( 7.47) P=0.027 31 trast, patients meeting the criteria for borderline per-
Bipolar II no Borderline n=14 42.6 ( 11) P=0.021 64.3 sonality disorder had significantly higher scores for
Borderline and Bipolar II n=12 39.2 ( 10.1) 58.3
OPD no Bipolar II n=93 39 ( 10.61) 61.3
ANG [F(1,5)=9.7, d.f. 148, P=0.002) and a tended to
have higher scores for ANX (F(1,5)=2.8, d.f. 148,

Table 2
Mean (SEM) for the subscales of ALS in the four groups patients

DEP ELAT ANXDEP ANX ANG BIP TOT

Borderline No Bipolar II (n=29) 1.6 (0.08) 1.3 (0.08) 2.0 (0.1) 1.6 (0.1) 1.6 (0.1) 1.3 (0.09) 1.6 (0.08)
Bipolar II no BPD (n=14) 1.7 (0.1) 1.4 (0.1) 1.9 (0.1) 1.4 (0.1) 1.2 (0.1) 1.3 (0.1) 1.5 (0.1)
BPD & BP II (n=12) 1.6 (0.1) 1.5 (0.1) 1.7 (0.1) 1.5 (0.1) 1.5 (0.1) 1.5 (0.1) 1.6 (0.1)
OPD, No BPII (n=93) 1.3 (0.05) 1.0 (0.05) 1.4 (0.7) 1.0 (0.06) 0.9 (0.07) 1.3 (0.1) 1.6 (0.1)
310 C. Henry et al. / Journal of Psychiatric Research 35 (2001) 307–312

Table 3
Total scores for the various scales used in the four groups

ALSTOT Mean (S.E.M.) AIM Mean (S.E.M.) BDHI Mean (S.E.M.) BISTOT Mean (S.E.M.)

Borderline no Bipolar II n=29 1.6 (0.08) P=0.039 3.6 (0.09) P=0.07 41.6 (2.1) P=0.018 45.9 (3.5) P=0.000
Bipolar II no Borderline n=14 1.5 (0.1) P=0.058 3.4 (0.1) 35.1 (3.6) 30.8 (4.5)
Borderline and Bipolar II n=12 1.6 (0.1) P=0.08 3.4 (0.1) 43.6 (3.9) 55.6 (6.2)
OPD no Bipolar II n=93 1.1 (0.05) 3.4 (0.05) 34.8 (1.2) 32.7 (1.5)

P=0.09]. A significant interaction between the two However, this dimensional approach to borderline
diagnoses was seen only for the ANXDEP subscale personality and bipolar II disorders also shows differ-
[F(1,5)=6.5, d.f. 148, P=0.01]. ences in emotional pattern between the two disorders.
A similar analysis with bipolar II disorder and bor- Emotional lability corresponds to short-term mood
derline personality disorder diagnoses as factors, and fluctuations with the same affective qualities, differing
total scores for AIM, BDHI and BISTOT as the only by degree, such as major mood episodes with both
dependent variables, demonstrated significant or trend polarities, one tending towards elation and an the other
effects for borderline personality but not for bipolar II towards depression. All items in the ALS specify that
disorder [F(1,5)=3.2, d.f. 148, P=0.07 for AIM; the affective shifts occurred very rapidly. Thus, this is
F(1,5)=5.9, d.f. 146, P=0.018 for BDHI; F(1,5)=22.7, not an evaluation of hypomanic and depressive epi-
d.f. 147, P=0.000 for BISTOT]. For all these scales, sodes, but of the pre-existing or residual instability
patients meeting the criteria for borderline personality between episodes. Akiskal (1994) suggested that bipolar
disorder had higher scores than those who did not. II disorder often involves dysthymic, irritable and
Bipolar II disorder diagnosis had no effect on scores for cyclothymic temperaments. However, this study shows
the BDHI and BIS or for the irritability and aggression that affective lability occurs even between major mood
subscale of the BDHI (Table 3). episodes in bipolar patients; hence, it could be con-
sidered a temperamental trait. Thus, the emotional
lability present in both borderline personality and
5. Discussion bipolar disorders may be a more specific trait, for which
there is some evidence of neurobiological dysregulation
In this study, we have found affective lability to be a (for review see Gurvits et al., 2000) where the choliner-
trait shared by both borderline personality and bipolar gic system may play a role (Steinnberg et al., 1997).
II patients. However, the two diagnoses presented dif- Borderline personality disorder patients differed from
ferent patterns of affective lability. Shifts to euthymia bipolar patients and subjects with other personality dis-
from anger and anxiety were associated with borderline orders in being more impulsive and aggressive. These
personality disorder in the absence of bipolar II dis- traits, impulsiveness and aggressiveness, seem to be very
order diagnosis whereas shifts from euthymia to characteristic of borderline personality disorder, as sug-
depression and elation, and shifts to elation from gested by previous studies (Virkkunen, 1976; Pattison
depression were characteristic of bipolar II patients. and Kahan, 1983; New et al., 1999). This is not surpris-
High scores on scales of impulsiveness and aggressive- ing because self-mutilation, increased suicidal gestures
ness were obtained only for borderline personality sub- (seen as acts of aggression against oneself) and impul-
jects. The difference between the groups was less siveness are among the diagnostic criteria for borderline
marked for the affective intensity. personality disorder. Further evidence of the sig-
Thus, affective lability may be a trait common to both nificance of impulsive aggressive symptoms in patients
borderline personality and bipolar II disorders, with personality disorders is provided by the frequency
accounting for the overlap in efficacy of mood stabi- of borderline personality in the criminal population
lizers, and the high frequency of comorbidity for these (Virkkunen et al., 1996; Dutton et al., 1996).
two diagnoses (46% for this sample). Mood stabilizers These data suggest that the neurobiological basis of
such as carbamazepine, lithium, and valproate have borderline personality disorder and bipolar disorder
been reported to be effective in borderline personality may include elements common to both disorders,
patients, but they principally reduce impulsiveness or resulting in a common core of affective lability. How-
improve overall function, rather than reduce depression ever, this affective lability is associated with greater
(Cowdry and Gardner, 1988; Stein et al., 1995). Thus, impulsivity in borderline personality disorder patients,
mood stabilizers may be effective treatment for border- leading to potentially self-damaging impulsiveness,
line personality disorder because they help to control inappropriate or uncontrolled anger, recurrent suicide
affective instability. threats or gestures. These data support the hypothesis
 C. Henry et al. / Journal of Psychiatric Research 35 (2001) 307–312 311

that there are biological predispositions (such as affec- Akiskal HS. Subaffective disorders: dysthymic, cyclothymic and bipo-
tive instability in this instance) that cut across diag- lar II disorders in the borderline realm. Psychiatric Clinics of North
America 1981;4(1):25–46.
nostic boundaries (Siever and Davis, 1991) and that
Akiskal HS. The temperamental borders of affective disorders. Acta
affective instability and impulsiveness are key dimen- Psychiatrica Scandinavica 1994;379:32–7.
sions of borderline personality disorder. In bipolar Atre-Vaidya N, Hussain S. Borderline personality disorder and bipo-
patients, this affective instability, probably associated lar mood disorders: Two distinct disorders or a continuum? Journal
with different neurobiological disturbances, leads to of Nervous and Mental Disorders 1999;57(5):313–5.
more enduring dysphoric mood tending towards either Barratt ES. Factor analysis of some psychometric measures of impul-
siveness and anxiety. Psychiatric Reports 1965;16:547–54.
elation or depression. These results suggest that bor- Benazzi F. Borderline personality disorder and bipolar II disorder in
derline personality disorder is not simply an attenuated private practice depressed outpatients. Comprehensive Psychiatry
subgroup of affective disorders. Our results are con- 2000;41(2):106–10.
sistent with those of a preliminary study (Atre-Vaidya Bolton S, Gunderson JG. Distinguishing borderline personality dis-
and Hussain, 1999) reporting that borderline person- order from bipolar disorder: differential diagnosis and implications.
Am J Psychiatry 1996;153(9):1202–7.
ality disorder could be distinguished from bipolar dis- Buss AH, Durkee A. An inventory for assessing different kinds of
order on the basis of temperament and character, hostility. Journal of Conscience Psychology 1957;21:343–9.
assessed by the model of personality proposed by Clo- Clayton PJ, Ernst C, Angst J. Premorbid personality traits of men
ninger (TPQ, Cloninger et al., 1994). It was found that who develop unipolar or bipolar disorders. European Archives of
borderline personality disorder patients scored higher Psychiatry and Clinical Neurosciences 1994;243:340–6.
Cloninger CR, Przybeck TR, Svrakic DM, Wetzel R. TCI-the tem-
for harm avoidance and lower for self-directedness and perament and character inventory: a guide to its development and
cooperativeness than bipolar patients. However, Vaidya use. St. Louis, Montana: Center for Psychobiology of Personality,
and Hussain only evaluated 10 borderline personality Washington University, 1994.
patients and 13 bipolar patients; it would be interesting Coccaro EF, Siever LJ, Klar HM, Maurer G, Cochrane K, Cooper TB,
to compare borderline personality, bipolar and control Mohs RC, Davis KL. Serotonergic studies in patients with affective
and personality disorders. Correlates with suicidal and impulsive
subjects in the same study to determine whether bor- aggressive behaviour. Arch Gen Psychiatry 1989;46(7):587–99.
derline personality and bipolar patients can be dis- Cowdry RW, Gardner DL. Pharmacotherapy of borderline person-
tinguished from controls in term of harm avoidance. ality. Archives of General Psychiatry 1988;45:111–9.
One limitation of this study is its reliance on self- Dutton DG, Starzomski A, Ryan L. Antecedents of abusive person-
ality and abusive behavior in wife assaulters. Journal of Family
reports to assess the various dimensions. However, these
Violence 1996;11(2):113–32.
ALS measures are significantly associated with the Frances A, Clarkin JF, Gilmore M, Hurt SW, Brown R. Reliability of
diagnostic criterion of affective instability rated by a criteria for borderline personality disorder: a comparaison of DSM-
diagnostic interviewer (Koenigsberg, personal commu- III and the diagnostic for borderline patients. American Journal of
nication), demonstrating the validity of ALS. Psychiatry 1984;141:1080–4.
These data also suggest other possible ways to explore Gunderson JG, Phillips KA. A current view of the interface between
borderline personality disorder and depression. Am J Psychiatry
the boundaries of affective disorders. Indeed, affective 1991;148:967–75.
instability may be a trait in bipolar disorders and a Gurvits IG, Koenigsberg HW, Siever LJ. Neurotransmitter dysfunc-
marker of susceptibility to bipolar disorder in relatives. tion in patients with borderline personality disorder. The Psychiatric
Further investigations are now required to explore Clinics of North America 2000;23(1):27–40.
bipolar I patients and the relatives of bipolar patients Harvey PD, Greenberg BR, Serper MR. The affective lability scales:
development, reliability and validity. Journal of Clinical Psychology
determine whether this is indeed the case. 1989;45:786–93.
Koenigsberg HW, Anwunah I, New AS, Mitropoulou V, Schopick F,
Siever LJ. Relationship between depression and borderline person-
Acknowledgements ality disorder. Depression and Anxiety 1999;10:158–67.
Kutcher SP, Marton P, Korenblum M. Adolescent bipolar illness and
This work was supported by the Programme Hospi-
personality disorder. Journal of American Academic Child and
talier de Recherche Clinique (PHRC 2000) and by Adolescent Psychiatry 1990;29(3):355–8.
grants from the National Institutes of Health; National Larsen RJ, Diener E. A multitrait-multimethod examination of affect
Center for Research Resources (5 M01 RR00071), for structure: hedonic level and emotional intensity. Personality Indivi-
the Mt. Sinai General Clinical Research Center; dual Differences 1985;6:631–6.
National Institute of Mental Health grant R01-MH- Larsen RJ, Diener E, Emmons RA. Affect intensity and reactions to daily
life events. Journal of Personality and Social Psychology 1986;51:803–14.
41131; and Department of Veterans Affairs Merit Lauer CJ, Bronisch T, Kainz M, Schreiber W, Holsboer F, Krieg JC.
Award 7609004. Pre-morbid psychometric profile of subjects at high familial risk for
affective disorder. Psychological Medicine 1997;27:355–62.
Lorenger AW, Oldham JM, Tulis EH. Familial transmission of DSM-
References III borderline personality disorder. Archives of General Psychiatry
1982;39( 7):795–9.
Amadeo S, Abbar M, Fourcade ML, Scharbach H, Selin D, Bretome McGlashan TH. The borderline syndrome: II: is it a variant of schi-
A, Madec A, Castelnau D, Besancon G. Clinical study of 5 families zophrenia or affective disorder? Archives of General Psychiatry
with bipolar disorder. Encephale 1992;18(1):78–82. 1983;40:1319–29.
312 C. Henry et al. / Journal of Psychiatric Research 35 (2001) 307–312

New AS, Sevin EM, Mitropoulou V, Reynolds D, Novotny SL, Call- Spitzer RL, Endicott J. The schedule of affective disorders and schi-
aban A, Trestman RL, Siever LJ. Serum cholesterol and impulsivity zophrenia (SADS). New York Biometrics Research, NY: State
in personality disorders. Psychiatry Research 1999;85(2):145–50. Psychiatric Institute, 1975.
Pattison E, Kahan J. The deliberate self-harm syndrom. American Stangl D, Pfohl B, Zimmerman M, Bowers W, Corenthal C. Struc-
Journal of Psychiatry 1983;140(7):867–72. tured interview for the DSM-III personality disorders. Archives of
Perry JC. Depression in borderline personality disorder: lifetime pre- General Psychiatry 1985;42:591–6.
valence at interview and longitudinal course of symptoms. Am J Stein DJ, Simeon D, Frenkel M, Islam MN, Hollannder E. An open
Psychiatry 1985;142(1):15–21. trial of valproate in borderline personality disorder. Journal of
Pope HG, Jonas JM, Hudson JI, Cohen BM, Gunderson JG. The Clinical Psychiatry 1995;56:506–10.
validity of DSM-III borderline personality disorder: a phenomen- Steinberg BJ, Trestman R, Mitropoulou V, Serby M, Silverman J,
ologic, family history, treatment response, and long-term follow-up Coccaro E, Weston S, de Vegvar M, Siever LJ. Depressive response
study. Archives of General Psychiatry 1983;40:23–30. to physostigmine challenge in borderline personality disorder
Sato T, Sakado K, Uehara T, Narita T, Hirano S. Personality patients. Neuropsychopharmacology 1997;17(4):264–73.
disorder comorbidity in early-onset versus late-onset major depression Torgersen S. Genetics of patients with borderline personality disorder.
in Japan. Journal of Nervous and Mental Disease 1999;187(4):237– Psychiatry Clinics of North America 2000;23(1):1–9.
42. Virkkunen M. Reactive hypoglycemic tendency among habitually
Siever LJ, Davis KL. A psychobiological perspective on the person- violent offenders. Nutritional Review 1976;44(Suppl.):94–103.
ality disorders. Am J Psychiatry 1991;148:1647–58. Virkkunen M, Eggert M, Rawlings R, Linnoila M. A prospective fol-
Silverman JM, Pinkham L, Horvath TB, Coccaro EF, Howard K, low-up study of alcoholic violent offenders and fire setters. Archives
Schear S, Apter S, Davidson M, Mohs R, Siever LJ. Affective and of General Psychiatry 1996;53:523–9.
impulse personality disorder traits in the relatives of patients with Zanarini MC, Frankenburg FR, Dubo ED, Sickel AE, Trikha A,
borderline personality disorder. American Journal of Psychiatry Levin A, Reynolds V. Axis I comorbidity of borderline personality
1991;148(10):1378–85. disorder. Am J Psychiatry 1998;155(12):1733–9.