Physical
J.
Stuart
and Sexual Abuse Histories With Borderline Personality Goldman,
Eugene J. D’Angelo, Ph.D., David and Enrico Mezzacappa, M.D.
M.D.,
Objective: The purpose ofthe study was to determine whether abuse is more common in children with borderline personality evaluated in the same outpatient abuse in 44 children diagnosed children.
Results:
The
borderline with
J
(Am
B
personality
combined groups.
personality
the
psychiatric clinic. Method: with borderline personality
borderline
lence ofphysical and significantly between
Among Disorder
disorder
physical/sexual Conclusions:
disorder
supports
group
the hypothesis
that
Ray DeMaso,
a history disorder
ofphysical than in other
The authors disorder and had
rates abuse
a history
M.D.,
or sexual children
contrasted rates in I 00 comparison
a significantly
abuse. Sexual abuse The finding of greater
Children
greater
preva-
alone did not in the group
of trauma
of
differ with
is associated
disorder. Psychiatry
1992;
149:1723-1726)
_____
Psychodynamic, family, and biologic etiologies have been hypothesized (12-15). Recently, reports of a 60%-80% prevalence of childhood abuse, particularly sexual abuse, in adults with borderline personality disorder have emerged (16-18). It has been noted that the prevalence of sexual abuse differentiates patients with borderline personality disorder from comparison groups in both inpatient and outpatient psychiatric set-
ported a higher than average prevalence of physical abuse among their sample of 24 children with bonderline personality disorder who were evaluated in an inpatient setting. Grapentine et al. (unpublished 1990 paper) noted a greater than average prevalence of sexual abuse among hospitalized adolescent girls diagnosed with borderline personality disorder. Most recently, Famulano et al. (20) have described trauma and PTSD in a sample of 19 children with borderline personality disorder. The present study sought to determine whether a histony of physical or sexual abuse is more common in children with borderline personality disorder than in other children evaluated in the same outpatient psychiatnic clinic. This investigation reports on a much larger and younger group of children with borderline person-
tings
ality
orderline sive
personality
attention
tune
(1-8).
line
personality
disorder
in the
Through
adult has
consensus
(17,
18).
regarding
been
On
the
the
basis
received
exten-
clinical
litera-
child
use of DSM-III-R disorder
and reliably diagnosed syndrome long-term emotional and behavioral While diagnostic criteria have is little
has
and
criteria, shown
border-
to
be
(9,
10) with consequences been improved, cause
of these
of
the
a valid
serious, ( 1 1). there disorder.
findings,
Herman
and van den Kolk (1 9) have hypothesized that psychological trauma, specifically abuse, plays an important role in the genesis of borderline personality disorder. They speculated that borderline personality disorder has symptomatic parallels with posttnaumatic stress disorder (PTSD). In contrast, the reports of childhood abuse in the histories of children with borderline personality disorder have been limited in nature and have tended to be confined
to
case
reports
(6,
7).
Bemporad
et
al.
(6)
re-
Received Nov. 6, 1 99 1 ; revision received March 1 1 , I 992; accepted April 10, 1992. From Children’s Hospital and Harvard Medical School, Boston. Address reprint requests to Dr. Goldman, Department of Psychiatry, Children’s Hospital, 300 Longwood Ave., Boston, MA 02115. Copyright © 1992 American Psychiatric Association.
Am
J
Psychiatry
149:12,
December
1992
disorder
than
has
previously
appeared
in the
lit-
enature.
METhOD The study was conducted in the outpatient psychiatry clinic of a tertiary care pediatric hospital. Patients were referred from a wide range of sources within and outside the hospital. The child and family member or canegiver each underwent a 1-hour diagnostic interview. Semistructuned interviews were conducted by a staff psychiatrist, psychologist, social worker, or child psychiatry resident. The interview used a standard format to generate information in the following areas: current behavior and symptoms of the child; medical, educational, developmental, and family histories; and mental status examination.
I 723
BORDERLINE
TABLE Adult
PERSONALITY
1. DSM-III-R
Criteria for Borderline
Perso nality
Disorder Adapted to Account for Developme ntal Differences
Criterion
1. A pattern
Adapted of unstable
personal
and
relationships
alternation idealization 2.
DISORDER
intense
characterized
between extremes and devaluation
Impulsiveness potentially
in at least self-damaging,
sex, substance
inter-
two
A pattern
by of over-
areas that are e.g., spending,
use, shoplifting,
reckless
driving, binge eating (do not include suicidal or self-mutilating behavior covered in item 5) 3. Affective instability: marked shifts from baseline mood to depression, irritability, or anxiety, usually lasting a few hours
and only
rarely
more
than
a few days
4.
Inappropriate, intense anger or lack of control of anger, e.g., frequent displays of temper, constant anger, recurrent physical fights S. Recurrent suicidal threats, gestures, or behavior or self-mutilating behavior
6.
Marked and persistent identity disturbance manifested by uncertainty about at least two of the following: self-image, sexual orientation, long-term goals or career choice, type of friends desired, preferred values 7. Chronic feelings of emptiness or boredom
of unstable
personal
abuse,
Affective from bility,
hours
and
relationships
alternation idealization distortion Impulsiveness potentially
taking,
Childhood
intense
Examples
inter-
characterized
by
between extremes of overand devaluation and/or marked of the nature of the relationship in at least two areas that are self-damaging, e.g., reckless risk
running sex,
away,
binge
stealing,
Describing
teacher
chronic
inability
ships
Walking sniffing
despite
as a “girlfriend,”
to maintain
wish
across glue
friend-
to do so
railroad
bridge
railing,
substance
eating
instability: marked, rapid baseline mood to depression, or anxiety lasting less than
and only rarely
days. Episodes distortions of Inappropriate, trol of anger, per, constant
in Childhood
Criterion
more
than
shifts irnitaa few
Early-afternoon persecutory successful
game
a few
may include transient reality intense anger or lack of cone.g., frequent displays of temanger, recurrent physical fights
Recurrent suicidal threats, gestures, or behavior or self-mutilating or selfendangering acts Marked and persistent disturbance in selfperception and -presentation characterized by confusion regarding two of the following: gender identity or roles, friendships, socially appropriate behaviors, school or career plans, self-image Chronic feelings of emptiness or boredom
anxiety delusions, participation
attack with followed by in soccer
in late afternoon
Easily provoked, frequent threatens and attempts therapist out window
fights, to throw
Carves boyfriend’s name into arm, multiple episodes of being struck by car Chronic cross-dressing, running for class president despite having no friends
Chronic
complaints
of boredom, unin appropriate activities Continual concern that therapist will not be there at next appointment,
able to invest
8. Frantic efforts abandonment self-mutilating
to avoid real or imagined (do not include suicidal or behavior covered in item 5)
Frantic pation
efforts with
to avoid or major preoccureal or imagined abandonment
The interview data were reviewed at the time of the diagnostic session by an interdisciplinary team headed by the first author. A consensus diagnosis was made for all subjects according to DSM-III on DSM-III-R cnitenia. In cases in which patients had been diagnosed before the DSM-III revision, all relevant records were reviewed to determine whether patients met the DSMIII-R criteria. All subjects in this study maintained their initial diagnosis after this careful record review. Subject
Selection
while
parent
is
na makes the diagnosis of borderline personality disorden highly likely. The convergent validity of the two groups was also demonstrated by using the Diagnostic Interview for Borderline Patients-Record Review ( DIB-R) (22). Each record was reviewed in a blind manner and cumulative DIB-R scores were derived (23). The analysis indicated that the entire group with a definite diagnosis of borderline personality disorder and 17 of the 22 subjects with a probable diagnosis were at or above the DIB-R cutoff score of 7. Of the five remaining children in
Table 1 presents DSM-III-R criteria for borderline personality disorder that were adapted to account for developmental differences across childhood. From just under 2,000 consecutive subjects evaluated since 1982, all subjects who received a definite or probable diagnosis of borderline personality disorder were selected for the study. A diagnosis of definite borderline personality disorder was made when a child met at least five of the eight DSM-III-R criteria. A diagnosis of probable borderline personality disorder was made when a child met at least four of the criteria. The group with a probable diagnosis of borderline personality disorder was included in this study on the basis of Clankin et al.’s finding (21 ) that the presence of four of the symptom cnite-
1724
refusal to leave house at work
(do not include suicidal or self-mutilating behavior included in item 5)
the
group
with
a probable
diagnosis,
four
were
point and the fifth subject was 2 points below the cutoff score. A comparison group of 100 subjects was randomly selected from patients who received diagnostic evaluations in the outpatient clinic concurrently with the study group. They all underwent the same diagnostic process and clinical review as described for the study group and had a wide range of diagnoses other than borderline personality disorder. Subjects Of the 44 children in the group with sonality disorder, 32 were boys (mean
Am
J
Psychiatry
borderline age=10.8
1 49:1 2, December
penyears,
1992
I
GOLDMAN,
SD=3.6) and 12 were girls (mean age=12.4, SD=4.S). The families’ mean Hollingshead and Redlich occupation score (24) was 4.6 (SD=1.S). The comparison group consisted of 62 boys (mean age=99 years, SD= 4.3) and 38 girls (mean age=l0.0 years, SD=3.8). Their families’
mean
Hollingshead
score was 5.1 (SD=1 ferences in age, sex, the two groups. Assessment
and
.8). There on families’
of Childhood
Redlich
occupation
were no significant occupation scones
diffor
Abuse
The semistructuned interview data were reviewed for each subject to identify the presence or absence of childhood physical or sexual abuse histories obtained during the diagnostic assessment. It was noted that the subjects had
been
abused
only
when
clear
data,
such
as physical
evidence on direct statements by the child, parent, or protective agency, were obtained. Cases in which abuse was possible or even probable but in which there was no substantiating data were not considered cases of abuse for this study. None of the subjects in cases of abuse was embedded in the complex context of custody, visitation, or divorce disputes, a context that might diminish the validity of the complaint.
RESULTS Seventeen of the 44 children with borderline personality disorder (38.6%) and nine ofthe 100 clinical cornpanison children (9%) had an abuse history. The cornplex chi-squane analysis revealed that the children with borderline personality disorder had a significantly greater frequency of abuse than did the comparison group (x2=2S-S df=3, p