Hypothesized Mechanisms of Change in Cognitive
Therapy for Borderline Personality Disorder

䊲

Amy Wenzel, Jason E. Chapman, Cory F. Newman,
Aaron T. Beck, and Gregory K. Brown

University of Pennsylvania

Preliminary evidence suggests that cognitive therapy (CT) is effective in
treating borderline personality disorder (BPD). According to cognitive theory,
BPD patients are characterized by dysfunctional beliefs that are relatively
enduring and inflexible and that lead to cognitive distortions such as
dichotomous thinking. When these beliefs are activated, they lead to
extreme emotional and behavioral reactions, which provide additional con-
firmation for the beliefs. It is hypothesized that a change in dysfunctional
beliefs is the primary mechanism of change associated with CT. How-
ever, additional mechanisms of change are likely also at work in CT, includ-
ing enhancement of skills, reduction in hopelessness, and improvement
in attitude toward treatment. Each of these mechanisms is discussed in
light of cognitive theory, data from an open clinical CT trial, relevant lit-
erature, and therapeutic interventions. Findings from the CT trial support
the role of cognitive change during therapy and its continuation after
termination. © 2006 Wiley Periodicals, Inc. J Clin Psychol 62: 503–
516, 2006.

Keywords: borderline personality disorder; cognitive therapy; cognitive
theory; dysfunctional beliefs

A salient feature of patients who have borderline personality disorder (BPD) is their
tendency to engage in dichotomous thinking, as they often evaluate themselves, other
people, and their environment in extreme terms and demonstrate little flexibility in assim-
ilating new information to modify their rigid beliefs (e.g., Arntz, 1994; Veen & Arntz,
2000). This pattern of distorted perception often results in angry outbursts, impulsive
behavior, and/or severe and sudden symptoms of anxiety and depression. One purpose of
cognitive therapy (CT) is to help patients develop tools to identify and evaluate such
cognitive distortions, given that a realistic appraisal of one’s circumstances will reduce

Correspondence concerning this article should be addressed to: Aaron T. Beck, M.D., Psychopathology Research
Unit, Department of Psychiatry, University of Pennsylvania, 3535 Market St., Room 2032, Philadelphia, PA
19103; e-mail: [e-mail].

JOURNAL OF CLINICAL PSYCHOLOGY, Vol. 62(4), 503–516 (2006)

© 2006 Wiley Periodicals, Inc.

Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jclp.20244

the severity of accompanying distress (e.g., Beck, 1995; Beck, Freeman, & Davis, 2004;
Beck, Rush, Shaw, & Emery, 1979). In addition, individuals who have BPD often have
dysfunctional thoughts about their own desirability and the trustworthiness of others, and
well-developed CT strategies have been designed to modify these beliefs. Although CT
comprises techniques with established utility in addressing cognitive distortions and dys-
functional beliefs, to date only one open clinical trial has examined the efficacy of stan-
dard CT in the treatment of BPD (Brown, Newman, Charlesworth, Crits-Christoph, &
Beck, 2004). Results from this trial indicated that patients reported significant decreases
in suicide ideation, hopelessness, depression, and borderline symptoms at the end of
treatment and suggest that CT is worthy of future systematic research in the treatment of
this disorder.

Although there is preliminary support for the effectiveness of CT for BPD, less is

known about the factors that are responsible for the changes that occur during the course
of treatment. The identification, measurement, and evaluation of the “active ingredients”
in CT for the treatment of BPD are important steps toward the continued refinement of
this intervention. Thus, the purpose of this article is to propose specific mechanisms of
change in CT for BPD. According to cognitive theory, the major hypothesized mecha-
nism of change involves a change in dysfunctional core beliefs that BPD patients hold
about themselves and the trustworthiness of others (Beck et al., 2004). However, we also
propose some additional mechanisms of change, including enhancement of skills, reduc-
tion in hopelessness, and improvement in attitude toward treatment. We discuss each of
these mechanisms in light of cognitive theory, preliminary empirical data, relevant liter-
ature, and therapeutic interventions associated with CT.

Cognitive Theory of Borderline Personality Disorder

At the heart of cognitive theory for BPD is the presence of dysfunctional beliefs that
influence patients’ perceptions about themselves, others, and their environment (Beck
et al., 2004; Layden, Newman, Freeman, & Morse, 1993). According to cognitive theory,
dysfunctional beliefs stem from negative learning experiences in childhood (Newman,
1998) that inhibit the development of flexible information processing (Arntz, 1994). In
adulthood, these dysfunctional beliefs are relatively enduring and inflexible, making it
difficult for patients to respond to complex and ambiguous demands. Moreover, these
beliefs are self-perpetuating because they influence the manner in which patients process
and interpret information in their environment, making it difficult for them to attend to
information that disconfirms their beliefs. Complicating this pattern of cognition further
is the observation that in many cases, two or more seemingly contradictory sets of beliefs
are activated in rapid succession. This cognitive profile, in turn, leads to extreme, dichot-
omous interpretations; heightened anxiety; depression, frustration, and shame; and urges
to engage in extreme behaviors to reduce the tension created by this antagonism (cf.
Layden et al., 1993).

The contents of beliefs associated with BPD cut across themes associated with a

number of Axis II pathologies. For example, Butler, Brown, Beck, and Grisham (2002)
identified 14 items from their Personality Beliefs Questionnaire (PBQ) that discrimi-
nated between patients who have BPD and patients with other personality disorders.
These items were from their Dependent, Paranoid, Avoidant, and Histrionic scales and
included beliefs such as dependency (e.g., “I am needy and weak”), helplessness (e.g., “I
am helpless when left on my own”), distrust (e.g., “ People will get at me if I don’t get
them first”), fears of abandonment (e.g., “If people get close to me, they will discover the
real me and reject me”), fears of losing control (e.g., “Unpleasant feelings will escalate

504

Journal of Clinical Psychology, April 2006

Journal of Clinical Psychology

DOI 10.1002/jclp

and get out of control”), and attention seeking behavior (e.g., “ People will pay attention
only if I act in extreme ways”). Arntz, Dietzel, and Dreessen (1999) identified a series of
20 assumptions specific to BPD and found that the degree to which BPD patients believed
these assumptions did not vary with the induction of positive and negative moods; that
finding suggests that faulty beliefs associated with this pathology are persistent and
engrained. Recently, Arntz, Dreessen, Schouten, and Weertman (2004) reported a revised
version of their inventory of BPD assumptions, which included themes of loneliness,
unlovability, rejection, abandonment of others, and viewing of oneself as bad and deserv-
ing of punishment.

Identification of dysfunctional beliefs is of paramount importance in CT because it

serves as a framework to understand and address seemingly contradictory emotional
reactions and behavioral responses. By accurately identifying their patients’ core beliefs,
cognitive therapists will understand the rules and assumptions that guide their patients’
experience, the automatic thoughts that are most likely to emerge in particular situations,
the intensity of negative affect associated with difficult situations, and subsequent behav-
ioral responses. Many BPD patients have a limited behavioral repertoire caused by inad-
equate learning experiences during childhood, and the activation of dysfunctional beliefs
further decreases the probability that they will choose the appropriate response in demand-
ing situations. For example, BPD patients frequently experience intense and unstable
interpersonal relationships and may exhibit behaviors that are violent, dependent, demand-
ing, self-harming, or suicidal in nature. Thus, their skills deficits contribute to failed
relationships and difficulty in reaching goals, which further intensify their negative affect.

In addition, BPD patients often engage in harmful and impulsive behaviors as a way

to release tension, to make a dramatic statement to others, and/or to be distracted from
emotional pain. These are some of the compensatory strategies that provide the BPD
patients with temporary “relief” but that actually serve to make their situation worse. The
result is a vicious circle of distorted cognition, negative affect, strong impulses, and
maladaptive behavioral responses that confirm or perpetuate the original distorted beliefs.
Further, the strains that BPD patients’ behaviors place on their relationships also cycle
back and “confirm” their unlovability, incompetence, and abandonment beliefs. This cycle
creates a sense of hopelessness in these patients, as they believe that they are powerless
to make positive changes in their life.

Thus, there is a high likelihood that BPD patients will view themselves as not being

able to have normal relationships and that they will actually suffer damage in their alli-
ances with important others, including therapists. Arntz (1994) observed that patients
who have BPD are often ambivalent toward psychotherapy; although they desire help and
acceptance from the therapist, they also fear rejection. Because of their ambivalence,
these patients are at risk for noncompliance and early termination from treatment. After a
series of unsuccessful courses of psychotherapy, not surprisingly patients who have BPD
develop a negative attitude toward treatment, which decreases the probability that a strong
therapeutic relationship will develop, that they will invest in treatment, and ultimately,
that treatment will be successful.

Hypothesized Mechanisms of Change

Belief Change

It is hypothesized that the principal mechanism of change in CT for BPD is the modifi-
cation of dysfunctional beliefs. In the early stages of therapy, modification of beliefs
begins to occur as a result of restructuring the automatic thoughts that emerge in particular

Change in BPD

505

Journal of Clinical Psychology

DOI 10.1002/jclp

situations. Beliefs are absolute representations of the individual, the world, and others but
are often difficult for patients to articulate, particularly in the early stages of CT. How-
ever, as patients become socialized to the cognitive model and observe themes associated
with their automatic thoughts, they gain insight into the beliefs that drive these thoughts
and other cognitive dysfunctions. Automatic thoughts and beliefs are treated as hypoth-
eses to be tested.

Some empirical support exists for the notion that change in dysfunctional cognition

is an important mechanism of change in CT for depression. For example, Jarrett and
Nelson (1987) examined specific components of CT in order to identify those associated
with patient change. The components were self-monitoring (i.e., self-guided monitoring
of thoughts and assumptions as they relate to mood), logical analysis (i.e., the skill of
systematically evaluating and revising inaccurate thoughts by evaluating evidence and
identifying thinking errors), and hypothesis testing (i.e., the skill of developing “experi-
ments” to test the accuracy of thoughts). A sample of depressed participants received
each of the three components in a predetermined order, and participants’ thoughts were
measured by using the Automatic Thoughts Questionnaire (Hollon & Kendall, 1980).
Results indicated that exposure to the logical analysis and hypothesis testing components
of CT were associated with reductions in both the frequency and the degree of belief in
depression-related dysfunctional thoughts. Additionally, DeRubeis and associates (1990)
found some support for the role of patients’ attitudes, attributional style, and hopeless-
ness as mediators of change in CT of depression. Moreover, Oei and Sullivan (1999)
examined the efficacy of group cognitive-behavioral therapy for depression and reported
that recovered patients endorsed fewer negative automatic thoughts and dysfunctional
attitudes than nonrecovered patients. Although preliminary, these findings provide some
evidence for cognitive change as a mechanism of change in CT of depression, though the
extent to which these findings hold in the treatment of BPD remains unclear.

In order to examine the degree to which borderline beliefs change with CT, we

administered the BPD scale of the Personality Beliefs Questionnaire (PBQ; Butler et al.,
2002) to patients enrolled in our open clinical trial (Brown et al., 2004). The mean PBQ-
BPD scores at baseline, treatment termination, and 18-month follow-up are presented in
Table 1. Paired samples t tests were conducted to determine whether PBQ-BPD scores
changed during these intervals. Because the small sample size is not optimal for uncov-
ering statistical significance, we also calculated effect sizes for correlated samples to
establish the magnitude of these differences (Dunlap, Cortina, Vaslow, & Burke, 1996).
Results of these analyses revealed a statistically significant reduction in the PBQ-BPD
score between baseline and treatment termination, t(22)

⫽ 4.96, p ⬍ .001, d ⫽ .88, and

between baseline and follow-up, t(16)

⫽ 5.14, p ⬍ .001, d ⫽ 1.19, but not between

treatment termination and follow-up, t(17)

⫽ 1.36, p ⫽ .19, d ⫽ .24. According to the

Table 1
Mean Personality Beliefs Questionnaire—Borderline Personality Disorder Scores

Total Sample

Responders

Nonresponders

Baseline

31.15 (10.60; n

⫽ 27)

31.65 (11.46; n

⫽ 22)

28.92 (5.72; n

⫽ 5)

Termination

22.17 (12.11; n

⫽ 25)

21.32 (13.00; n

⫽ 20)

25.60 (7.64; n

⫽ 5)

18-Month follow-up

17.37 (11.26; n

⫽ 19)

16.27 (11.38; n

⫽ 15)

21.50 (11.27; n

⫽ 4)

Note. The first values in parentheses are standard deviations. Sample sizes vary because there were not complete data sets on
some patients.

506

Journal of Clinical Psychology, April 2006

Journal of Clinical Psychology

DOI 10.1002/jclp

guidelines described by Cohen (1988), the effect sizes characterizing changes between
baseline and termination and baseline and follow-up were large, and the effect size char-
acterizing changes between termination and follow-up was small.

In addition, we evaluated PBQ-BPD change with respect to patient treatment re-

sponse status. Treatment responders were defined as patients who had fewer than five
Diagnostic and Statistical Manual of Mental Disorders, fourth edition (DSM-IV), BPD
criteria at 18-month follow-up (cf. Brown et al., 2004). Of the 19 data sets available from
the 18-month follow-up period, 15 (79%) were considered treatment responders. Results
of an independent samples t test revealed a nonsignificant difference between the groups,
t(15)

⫽ 1.84, p ⫽ .09, but a large effect size, d ⫽ ⫺.99.

Item-level PBQ-BPD analyses also were performed to evaluate change that occurred

between baseline and 18-month follow-up. Results of paired samples t tests and effect
size estimates for each of the items are presented in Table 2. These exploratory findings
demonstrate that statistically significant reductions occurred in patients’ degree of belief
in 11 of the 14 (79%) PBQ-BPD items between baseline and the 18-month follow-up.
More importantly, the effect size estimates ranged from .39 to 1.35 (M

⫽ .76, SD ⫽ .33).

In summary, BPD patients who participated in the open clinical CT trial reported

significant reductions in borderline beliefs between baseline and termination. These gains
persisted 18 months after treatment. The greatest belief changes were associated with
items assessing distorted self-perceptions, such as “If people get close to me, they will
discover the ‘real’ me and reject me.” Belief changes were of a more moderate level for
items assessing mistrust, such as “I have to be on guard at all times.” Although these data
suggest that BPD patients substantially altered dysfunctional beliefs during the course of
the treatment trial, the lack of control group precludes the ability to draw conclusions
about the degree to which these changes are above and beyond those that would be
expected by the passage of time, and the lack of comparison to another form of psycho-
therapy precludes the ability to draw conclusions about the degree to which these changes
are specific to CT. Additionally, little is known about the manner in which these belief
changes occur, such as whether new beliefs replace previously held beliefs, new beliefs
compete with existing ones, or existing beliefs are modified (Robins & Hayes, 1993).
Nevertheless, the impressive effect sizes suggest that change in beliefs is a viable mech-
anism of change in CT, in accordance of predictions made by cognitive theory.

Other Hypothesized Mechanisms of Change

Enhancement of Behavioral Skills. The central proposed mechanism by which treat-

ment change is achieved in CT is a change in dysfunctional beliefs. However, the sys-
tematic evaluation of cognitive distortions reduces negative affect, thereby allowing patients
to select appropriate behavioral responses in a more thoughtful way and to practice the
skills learned in therapy. Thus, CT allows patients who have to develop and implement
new behaviors to negotiate interpersonal conflict and manage personal distress. In the fol-
lowing section we briefly address additional skill-based mechanisms of change on the basis
of existing empirical findings, cognitive theory, and our targeted intervention for BPD.

Behavioral experiments involve the formulation of specific, testable hypotheses about

the patient’s behavior with the aim of gathering evidence for the evaluation and refine-
ment of underlying beliefs (Brown & Newman, 1999). In CT for BPD, these experiments
are used as methods to test currently held beliefs and predictions and translate newly
developed beliefs into behavioral change. Research examining the role of behavioral
experiments in changes associated with CT is limited. Bennett-Levy (2003) found that
among members who participated in a CT training engaging in behavioral experiments

Change in BPD

507

Journal of Clinical Psychology

DOI 10.1002/jclp

T

able

2

Personality

Beliefs

Questionnair

e—Bor

derline

Personality

Disor

der

Individual

Item

Change

Fr

om

Baseline

to

18-Month

Follow-Up

PBQ-BPD

Items

MS

D

Lower

95%

CI

Upper

95%

CI

td

f

Sig.

d

If

people

get

close

to

me,

they

will

discover

the

“real”

me

and

reject

me.

1.47

1.28

.812

2.129

4.74

16

⬍

.001

1.21

Unpleasant

feelings

will

escalate

and

get

out

of

control.

1.06

1.35

.367

1.750

3.25

16

.005

0.88

Any

signs

of

tension

in

a

relationship

indicate

the

relationship

has

gone

bad;

therefore,

I

should

cut

it

of

f.

0.82

1.33

.138

1.509

2.55

16

.022

0.64

I

am

needy

and

weak.

1.35

1.66

.501

2.204

3.37

16

.004

1.03

I

need

somebody

around

available

at

all

times

to

help

me

to

carry

out

what

I

need

to

do

or

in

case

something

bad

happens.

0.63

1.41

⫺

.125

1.375

1.78

15

.096

0.62

I

am

helpless

when

left

on

my

own.

0.59

1.18

⫺

.016

1.193

2.06

16

.056

0.71

I

can’

t

cope

as

other

people

can.

1.41

1.28

.755

2.069

4.56

16

.000

1.29

People

will

get

at

me

if

I

don’

t

get

them

first.

0.59

0.94

.105

1.071

2.58

16

.020

0.39

People

will

pay

attention

only

if

I

act

in

extreme

ways.

1.35

1.17

.752

1.954

4.77

16

.000

1.35

I

cannot

trust

other

people.

0.88

1.09

.295

1.455

3.22

15

.006

0.62

I

have

to

be

on

guard

at

all

times.

0.65

1.22

.019

1.275

2.18

16

.044

0.46

People

will

take

advantage

of

me

if

I

give

them

the

chance.

0.88

1.80

⫺

.042

1.807

2.02

16

.060

0.62

People

often

say

one

thing

and

mean

something

else.

0.53

1.00

.011

1.047

2.17

16

.046

0.40

A

person

whom

I

am

close

to

could

be

disloyal

or

unfaithful.

0.71

0.99

.199

1.212

2.95

16

.009

0.48

508

Journal of Clinical Psychology, April 2006

Journal of Clinical Psychology

DOI 10.1002/jclp

was associated with changes in both behavior and beliefs. Furthermore, DeRubeis and
Feeley (1990) found that the CT strategy of testing the accuracy of beliefs was signifi-
cantly associated with changes in mood.

In addition, patients who have BPD also benefit from CT by developing problem

solving skills. The process of problem and goal definition, generation of alternative
responses, evaluation of the alternatives, and implementation and evaluation of the selected
strategy is an important component of the intervention that serves to equip BPD patients
with an adaptive behavioral repertoire for coping with stress. Similarly, we propose that
relationship enhancement skills, such as anger management, effective communication,
and assertiveness, are associated with improvements in interpersonal relationships and
decreases in mood fluctuation related to interpersonal conflict. Though research evaluat-
ing these domains as possible mechanisms of change is limited, our CT protocol and
clinical experience with BPD patients suggest that the development of these skills is
crucial to treatment success and should be studied systematically as a mechanism of
change.

Reduction in Hopelessness. CT aims to increase patients’ hope by systematically

targeting negative views of the future (Beck et al., 1979). With respect to the treatment of
BPD, this occurs as the therapist, while validating patients’ emotions, models hopeful-
ness in the ability to improve their current situation through the use of effective problem
solving skills and the evaluation of biased cognitions. Previous empirical research sup-
plies evidence that supports the hypothesis that reduction of hopelessness is a mechanism
of change associated with CT. For example, Kuyken (2004) found that depressed patients
who had greater reductions in hopelessness early in CT experienced improved outcomes
as compared to those who had smaller reductions in hopelessness, even when baseline
levels of depression and hopelessness were controlled. Similarly, Rush, Beck, Kovacs,
Weissenburger, and Hollon (1982) reported that patients who received an average of 11
weeks of CT experienced greater reductions in hopelessness than those who received a
pharmacological intervention (imipramine). Oei and Sullivan (1999) indicated that recov-
ered depressed patients reported lower levels of hopelessness after group cognitive-
behavioral therapy than nonrecovered depressed patients. This evidence suggests that
reduction of patient hopelessness is an important factor associated with successful CT for
depression.

We examined reduction in hopelessness in our open clinical trial of CT for BPD.

Patients completed the Beck Hopelessness Scale (BHS; Beck & Steer, 1989) at intake,
treatment termination, and 18-month follow-up. The mean BHS scores at each of these
intervals are presented in Table 3. Results of a paired samples t test between each of the
three time points revealed that BHS scores decreased significantly between baseline and

Table 3
Mean Beck Hopelessness Scale Scores

Total Sample

Responders

Nonresponders

Baseline

13.95 (5.79; n

⫽ 29)

13.94 (5.92; n

⫽ 24)

14.00 (5.79; n

⫽ 5)

Termination

9.59 (5.49; n

⫽ 29)

9.58 (5.79; n

⫽ 24)

9.60 (4.34; n

⫽ 5)

18-Month follow-up

7.04 (6.07; n

⫽ 27)

6.59 (6.12; n

⫽ 22)

9.00 (6.04; n

⫽ 5)

Note: The first values in parentheses are standard deviations. Sample sizes vary because there were not complete data sets on
some patients.

Change in BPD

509

Journal of Clinical Psychology

DOI 10.1002/jclp

termination, baseline and follow-up, and termination and follow-up: t(28)

⫽ 4.04, p ⬍

.001, d

⫽ .77; t(26) ⫽ 5.70, p ⬍ .001, d ⫽ 1.24; and t(26) ⫽ 3.49, p ⫽ .002, d ⫽ .43,

respectively. The effect size characterizing changes was medium between baseline and
termination, was large between baseline and follow-up, and was small between termina-
tion and follow-up. In addition, we evaluated BHS change with respect to patient treat-
ment response status, as operationally defined previously. There were 22 (81%) treatment
responders who had available BHS data. Results of an independent samples t test revealed
a nonsignificant difference between the groups at 18-month follow-up, t(25)

⫽ 0.80, p ⫽

0.43, but a small to medium effect size, d

⫽ .40. Thus, these data support the speculation

that reduction in hopelessness is a mechanism of change associated with CT for BPD,
although the effect size characterizing the difference between responders and nonrespond-
ers was smaller than the similar effect size for change in borderline beliefs.

Attitude Toward Treatment. Previous studies have found that patients who have pos-

itive expectancies for treatment endorse greater reductions in symptoms than patients
who have negative expectancies for treatment (e.g., Stewart et al., 1993; Westra, Dozois,
& Boardman, 2002). To examine this issue in our open clinical trial (Jeglic, Levy-Mack,
Wenzel, Beck, & Brown, 2004), we asked 29 patients to complete the Attitudes and
Expectations questionnaire adapted from a form used by Elkin and colleagues (1989) in
the NIMH Treatment of Depression Collaborative Research Program. Patients were iden-
tified as having either a positive or a negative attitude toward treatment by their response
to the question “What is your attitude toward talking with a therapist/counselor as treat-
ment for your problem?” It was scored on a scale from 1 (very positive) to 7 (very
negative). Patients who rated 1–3 on this question were regarded as having a positive
attitude, and patients who rated 4–7 on this question (where 4

⫽ neutral) were regarded as

having a negative attitude.

Results indicated that 66.7% of the patients who had positive attitudes toward treat-

ment no longer met criteria for BPD after 12 months of treatment, as compared to only
14.3% of the patients who had a negative attitude toward treatment. Moreover, as com-
pared to patients who had a negative attitude toward treatment, patients who had a pos-
itive attitude had higher Global Assessment of Functioning (GAF) scores, lower levels of
self-reported suicidal ideation, lower levels of self-reported depression, and lower levels
of interviewer-rated depression after 12 months of treatment (Jeglic et al., 2004). In all,
these findings raise the possibility that a positive attitude toward treatment creates a
context in which substantial symptom reduction can occur. Interestingly, change in bor-
derline beliefs as measured by the PBQ-BPD scale did not vary as a function of attitude
toward treatment; that finding raises the possibility that these potential mechanisms of
change work independently. However, it is important to acknowledge that attitude toward
treatment was not measured throughout the course of treatment; therefore, it is not pos-
sible to evaluate whether this variable interacts with therapist skill or the quality of the
therapeutic relationship over time to affect outcome.

Clinical Applications

The cognitive theory of BPD, including the hypothesized mechanisms of change, lends
itself to several clinical applications in the treatment of this disorder. These applications
include the development of a case conceptualization and the implementation of specific
strategies for identifying and changing dysfunctional core beliefs, improving behavioral
skills, decreasing hopelessness, and addressing patient attitude toward treatment. An over-
view of these techniques is provided in the following section.

510

Journal of Clinical Psychology, April 2006

Journal of Clinical Psychology

DOI 10.1002/jclp

Case Conceptualization

The cognitive case conceptualization of BPD is distinguished from alternate therapeutic
models by the explicit focus on the patient’s beliefs, assumptions, rules, and compensa-
tory strategies. Cognitive therapists recognize that many factors contribute to the etiology
of BPD, including genetic loading for poor impulse control, deficient emotion regulation,
skills deficits, and early abuse histories (Layden, 1998; Newman, 1998). However, accord-
ing to the cognitive model, it is the expression of these serious problems in the form of
dysfunctional ways of viewing themselves and others that serves to perpetuate patients’
problems, inhibit corrective learning, feed into interpersonal conflict, and reinforce a
sense of helplessness and hopelessness. The goals of the cognitive conceptualization are
(1) to develop a comprehensive and specific model of the patient’s background, beliefs,
assumptions, and compensatory strategies that accounts for his or her emotional and
behavioral problems and (2) to identify specific therapeutic interventions to target the
beliefs and compensatory strategies directly.

The conceptualization process is ongoing and incorporates frequent revisions as new

clinical information is acquired (Fossel & Wright, 1999). Formulating a workable, evolv-
ing case conceptualization is important in the treatment of any patient, but it is particu-
larly vital in making sense of the apparently chaotic and “illogical” life of the BPD
patient. As these patients often have a bewildering array of problems, many of which
seem deceptively unpredictable and unrelated to each other, it is necessary to develop a
conceptual road map that provides a sense of context, comprehensiveness, and direction
for the development of a treatment plan and treatment goals.

Belief Identification

The therapist’s task is to use guided discovery techniques to identify or hypothesize the
BPD patient’s beliefs (Brown & Newman, 1999). As previously described, the most
common beliefs observed for these patients involve those associated with dependency,
helplessness, distrust, fears of abandonment, and fears of losing control. Therapists may
utilize several sources of clinical data in order to begin the process of identifying patients’
relevant beliefs, including the presenting problem, crises that emerge during treatment,
and critical negative experiences in the patient’s history. During crises or when discus-
sion of critical negative experiences, the therapist may develop hypotheses about the core
beliefs by thinking about which beliefs might lead a patient to react in a specific way.
Once the therapist begins to formulate hypotheses about the patient’ beliefs, he or she
may also look for opportunities to discuss these thoughts with the patient in order to
obtain feedback. The best opportunities to solicit feedback occur either when an emo-
tional topic is being discussed or when an emotional reaction occurs during the session.
At this time, the patient may actually articulate a belief as an automatic thought.

There are several other strategies for identifying beliefs: (1) looking for the expres-

sion of a belief as an automatic thought, (2) recognizing a common theme across auto-
matic thoughts or situations, (3) directly asking the patient what he or she thinks the
belief is, (4) reviewing a completed PBQ, (5) using the downward arrow technique (Burns,
1980), and (6) completing a cognitive conceptualization diagram form (Layden et al.,
1993; Beck, 1995). In using the cognitive conceptualization diagram form, the therapist
and patient identify critical childhood incidents, core beliefs that are associated with
these incidents, conditional assumptions or intermediate beliefs, and compensatory strat-
egies as they affect several notable situations that epitomize the patient’s current diffi-
culties. An alternative strategy for identifying beliefs is the downward arrow technique

Change in BPD

511

Journal of Clinical Psychology

DOI 10.1002/jclp

(Burns, 1980), in which the therapist may ask the patient a series of questions such as
“What does this mean to you?” or “What does this mean about other people?” Regardless
of the specific technique chosen for identifying the core beliefs, therapists need to keep in
mind that identification of the core belief(s) is crucial for guiding the treatment.

Belief Modification

During treatment, one or more dysfunctional beliefs may be identified. If more than one
belief is identified, then the therapist and patient should decide which beliefs should be
the focus of treatment (Brown & Newman, 1999). In choosing which belief to address,
the therapist should identify the core belief as the belief that is most strongly endorsed
by the patient, is relatively enduring, and is clearly associated with the patient’s present-
ing problems. Patient-therapist collaboration in identifying and modifying specific beliefs
is of critical importance, and there are several strategies to facilitate this collaboration.
For example, patients may rate the extent to which they agree with the core belief(s) on
a scale of 0 to 100 (completely agree) so that the most strongly endorsed belief is addressed.
Once it has been determined that the patient is willing to work on the belief, there are a
variety of verbal, behavioral, and imagery strategies that can be utilized.

Verbal strategies for modifying core beliefs include (1) writing down the beliefs and

rating fluctuations in the degree of certainty of these beliefs over the course of a session,
day, or week (Beck, 1995); (2) completing a core belief worksheet in which evidence that
contradicts the core belief and supports a more adaptive one is noted (Beck, 1995); (3)
simply writing down the modified core belief; and (4) completing a dysfunctional thought
record (DTR). The core belief worksheet helps patients to articulate their beliefs and to
rate fluctuations in their degree of certainty about these beliefs over the course of a week.
Patients may then generate new, therapeutically desirable beliefs and, likewise, rate their
degree of certainty regarding this new way of thinking. In order to facilitate belief change,
patients are encouraged to document evidence that contradicts the dysfunctional core
belief and supports a more adaptive one. This approach of strengthening the new belief
and weakening the old belief produces a stronger therapeutic effect than either method
used alone (Layden et al., 1993). Although the DTR may also be used to modify core
beliefs (Layden et al., 1993), many BPD patients have difficulty completing it in the early
stages of treatment when cognitive processing skills are less developed. If patients do not
respond well to a standard cognitive therapy technique, it is advisable either to discon-
tinue or to modify it so that patients are more likely to be successful.

Patients may also be encouraged to monitor fluctuations in the degree of endorse-

ment of a particular belief. If changes in the degree of belief are observed only in given
situations, then the therapist and patient may agree to modify the belief so that it more
accurately reflects reality. Sometimes when patients write down their beliefs, they notice
that their written belief does not accurately capture the latent belief. In this case, patients
are encouraged to modify the written belief so that it is more accurate. This technique
often leads to shaping highly dysfunctional beliefs into more adaptive ones because they
are often revised to be less extreme or less pervasive.

In order to facilitate the patients’ translating their more adaptive beliefs into appro-

priate behavioral changes, there are several behavioral strategies that may be employed,
as described by Layden and her colleagues (1993): (1) Instruct patients to self-monitor
any adaptive behaviors that occur in response to adopting more adaptive beliefs and to
keep a journal of these behaviors for future reference. (2) Role-play various scenarios
with patients in session to improve interpersonal skills and to develop more adaptive

512

Journal of Clinical Psychology, April 2006

Journal of Clinical Psychology

DOI 10.1002/jclp

ways of viewing themselves and others. (3) Conduct behavioral experiments in which
patients formulate hypotheses about their belief and then, between sessions, act in such a
way as to test these hypotheses. Making small but steady behavioral changes advances a
patient’s adaptive behavioral repertoire so that dysfunctional beliefs may gradually erode
(Layden et al., 1993).

Imagery techniques are used to address beliefs that have a significant pictorial com-

ponent, such as vivid visual memories of early life events that were particularly upsetting
for the patient (Edwards, 1990). Imagery exercises may be used for three main purposes:
(1) to assess the content of a patient’s belief; (2) to reframe the meaning of a memory, so
that the belief is not maintained or “proved” without scrutiny; and (3) to reenact a painful
memory to develop a more positive self-affirming outcome. By exploring memories and
images of critical incidents in the past, therapists and their BPD patients can gain a better
understanding of the types of beliefs that may have formed. For example, as the patient
closes his eyes (perhaps after a brief relaxation induction) in order to visualize a humil-
iating moment in his childhood, the therapist asks the “child” in the image what he is
thinking and feeling. The therapist and patient then examine what the “child” is conclud-
ing about himself, other people, and his future as a function of the visualized event.
Further, the “child” is asked to offer his idiosyncratic explanation of why the remem-
bered event is happening and what lessons in life are learned from the negative experience.

A second function of the imagery exercise is to modify the “child’s” self-defeating

interpretations about the recalled event (Layden et al., 1993). For example, the adult
patient may be able to state that she does not blame herself for the sexual abuse that she
suffered as a child but may carry a great deal of residual self-hatred and guilt that are
associated with early maladaptive beliefs (e.g., “I’m bad, unlovable, worthless”). There-
fore, the intervention must attend not only to the adult patient in the “here and now,” but
also to the beliefs held by the “child” many years ago. Imagery exercises may be used to
address these beliefs held as a child. In addition, the therapist can use the image to help
the patient reenact a traumatic event in a healthier way. The therapist does so by asking
the patient to take control over the image, to manipulate the events in a way that was not
possible when the event really happened, and to produce a more favorable outcome. This
technique is not intended to deny the past but, rather, to help the patient gain a sense of
empowerment.

In summary, CT comprises a number of strategies to modify dysfunctional beliefs,

including cognitive restructuring, behavioral experiments, and the use of imagery. The
goal of each of these strategies is to improve patients’ ability to understand the manner in
which these beliefs influence their interpretations and emotional reactions and to evaluate
the accuracy of these beliefs. Regardless of the specific intervention strategies that are
applied, the most important task for the therapist is the development of an adequate case
conceptualization that guides treatment in the context of a collaborative and supportive
therapeutic relationship.

Enhancing Behavioral Skills

Although the techniques to modify cognitive distortions and dysfunctional beliefs are
central to CT for BPD, cognitive therapists incorporate behavioral strategies to manage
suicidal ideation, emotional dysregulation, and relationship distress for their own merit
as well. For example, cognitive therapists encourage suicidal BPD patients to develop
skills to “procrastinate” a suicide attempt so that they can think through their decision
more carefully when they are not in crisis. Moreover, they assist suicidal borderline

Change in BPD

513

Journal of Clinical Psychology

DOI 10.1002/jclp

patients in developing safety plans, including identification of warning signs of suicidal
behavior, short-term coping strategies, and emergency contacts. Cognitive therapists also
enlist a number of long-term behavioral strategies to manage mood and suicidal ideation,
including skills for increasing social support, enhancing basic self-care, developing self-
comforting and relaxation skills, improving problem solving strategies, and practicing
assertiveness and relationship enhancement skills. As mentioned previously, the cogni-
tive therapist makes use of behavioral experiments to evaluate patients’ distorted percep-
tions and provide an opportunity for them to practice skills.

Reducing Hopelessness

Hopelessness is a high-priority target for intervention in CT for BPD, as it is a risk factor
for eventual suicide (Beck, Steer, Kovacs, & Garrison, 1989). A patient’s report of increas-
ing hopelessness makes it the primary focus of the session. A number of strategies are
employed to reduce the patient’s hopelessness and facilitate adaptive coping. One of the
first steps toward targeting hopelessness entails the patient’s acknowledging that there is
a problem and, in collaboration with the therapist, adopting a problem solving stance.
The application of problem solving skills serves to instill hope by demonstrating the
availability of alternative and adaptive methods of responding to problems. In addition,
patients are taught to identify and evaluate cognitive distortions related to hopelessness,
to provide them with a less biased, extreme view of their future and highlight their ability
to solve problems effectively. Other therapeutic strategies for addressing hopelessness
include developing reasons for living, constructing a “Hope Kit” (memory aids that serve
as reminders for reasons to live in times of distress), and increasing social support.

Improving Attitude Toward Treatment

The general structure of CT includes several provisions for assessing and enhancing
patients’ attitude toward treatment (Beck, 1995). For example, cognitive therapists spend
a great deal of time educating patients about the cognitive model and clarifying their
expectations for the activities that will occur in and out of session. In addition, cognitive
therapists take a collaborative stance by working together with the patient to set agendas
for sessions, prioritize issues to be addressed in sessions, and developing homework
assignments to be completed outside sessions. Moreover, cognitive therapists solicit feed-
back from patients regarding whether they thought the topics addressed in session were
useful and whether there were any topics that should have been covered but were not.
These tactics communicate to patients that they are working collaboratively with their
therapists and that their ideas and insights are valued. If a negative attitude toward treat-
ment is detected, cognitive therapists apply cognitive restructuring techniques, in which
they help patients to identify negative automatic thoughts about treatment, evaluate the
evidence that supports and refutes them, and construct more realistic appraisals.

Conclusion

Although empirical research has confirmed that CT is efficacious for a number of Axis I
pathologies, only one open clinical trial on CT for BPD has been published to date. We
are currently conducting a randomized controlled trial to compare the efficacy of 6 months
of CT with treatment as usual for this population, and a similar trial is nearing completion
in the Netherlands. The identification and modification of core beliefs held by BPD

514

Journal of Clinical Psychology, April 2006

Journal of Clinical Psychology

DOI 10.1002/jclp

patients are hypothesized to be a mechanism of change associated with CT because the
individual cognitive case conceptualization serves as a framework for understanding their
emotional and behavioral reactions. Furthermore, this case conceptualization serves as a
guide for identifying specific therapeutic interventions that directly target dysfunctional
beliefs, skill deficits, hopelessness, and a negative attitude toward treatment. Our prelim-
inary data indicate that CT for BPD is effective in changing borderline beliefs and reduc-
ing hopelessness, and they suggest that symptom reduction varies with attitude toward
treatment. However, there is a paucity of empirical investigation of the manner in which
these potential mechanisms of change work together in the successful cognitive treatment
of BPD. Although we have proposed four potential mechanisms of change, we cannot
rule out the possibility that they are epiphenomena or consequences of other mechanisms
of change. Thus, we encourage future researchers to conduct research designed with the
primary purpose of identifying the specific and nonspecific ingredients that account for
the success of this intervention.

References

Arntz, A. (1994). Treatment of borderline personality disorders: A challenge for cognitive-

behavioural therapy. Behaviour Research and Therapy, 32, 419– 430.

Arntz, A., Dietzel, R., & Dreessen, L. (1999). Assumptions in borderline personality disorder:

Specific stability and relationship with etiological factors. Behaviour Research and Therapy,
37, 545–557.

Arntz, A., Dreessen, L., Schouten, E., & Weetman, A. (2004). Beliefs in personality disorders: A

test with the Personality Disorder Belief Questionnaire. Behaviour Research and Therapy, 42,
1215–1225.

Beck, A.T., Freeman, A., & Davis, D.D. (2004). Cognitive therapy of personality disorders (2nd

ed.). New York: Guilford Press.

Beck, A.T., Rush, A.J., Shaw, B.F., & Emery, G. (1979). Cognitive therapy of depression. New

York: Guilford Press.

Beck, A.T., & Steer, R.A. (1989). Manual for the Beck Hopelessness Scale. San Antonio, TX: The

Psychological Corporation.

Beck, A.T., Steer, R.A., Kovacs, M., & Garrison, B. (1989). Hopelessness and eventual suicide: A

10 year prospective study of patients hospitalized with suicidal ideation. American Journal of
Psychiatry, 42, 559–563.

Beck, J.S. (1995). Cognitive therapy: Basics and beyond. New York: Guilford Press.

Bennett-Levy, J. (2003). Mechanisms of change in cognitive therapy: The case of automatic thought

records and behavioural experiments. Behavioural and Cognitive Psychotherapy, 31, 261–277.

Brown, G.K., & Newman, C.F. (1999). Cognitive therapy treatment manual for borderline person-

ality disorder. Unpublished manuscript, University of Pennsylvania, Philadelphia, PA.

Brown, G.K., Newman, C.F., Charlesworth, S.E., Crits-Christoph, P., & Beck, A.T. (2004). An

open clinical trial of cognitive therapy for borderline personality disorder. Journal of Person-
ality Disorders, 18, 257–271.

Burns, D.D. (1980). Feeling good: The new mood therapy. New York: New American Library.

Butler, A.C., Brown, G.K., Beck, A.T., & Grisham, J.R. (2002). Assessment of dysfunctional beliefs

in borderline personality disorder. Behaviour Research and Therapy, 40, 1231–1240.

Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ:

Lawrence Erlbaum Associates, Inc.

DeRubeis, R.J., Evans, M.D., Hollon, S.D., Garvey, M.J., Grove, W.M., & Tuason, V.B. (1990).

How does cognitive therapy work? Cognitive change and symptom change in cognitive therapy

Change in BPD

515

Journal of Clinical Psychology

DOI 10.1002/jclp

and pharmacotherapy for depression. Journal of Consulting and Clinical Psychology, 58,
862–869.

DeRubeis, R.J., & Feeley, M. (1990). Determinants of change in cognitive behavioral therapy for

depression. Cognitive Therapy and Research, 14, 469– 482.

Dunlap, W.P., Cortina, J.M., Vaslow, J.B., & Burke, M.J. (1996). Meta-analysis of experiments with

matched groups or repeated measures designs. Psychological Methods, 1, 170–177.

Edwards, D.J. (1990). Cognitive therapy and the restructuring of early memories through guided

imagery. Journal of Cognitive Psychotherapy: An International Quarterly, 4, 33–50.

Elkin, I., Shea, M.T., Watkins, J.T., Imber, S.D., Sotsky, S.M., Collins, J.F., et al. (1989). National

Institute of Mental Health Treatment of Depression Collaborative Research Program: General
effectiveness of treatments. Archives of General Psychiatry, 46, 971–982.

Fossel, R.V., & Wright, J.H. (1999). Targeting core beliefs in treating borderline personality dis-

order: The case of Anna. Cognitive and Behavioral Practice, 6, 54– 60.

Hollon, S.D., & Kendall, P.C. (1980). Cognitive self-statements in depression: Development of the

Automatic Thoughts Questionnaire. Cognitive Therapy and Research, 4, 384–395.

Jarrett, R.B., & Nelson, R.O. (1987). Mechanisms of change in cognitive therapy of depression.

Behavior Therapy, 18, 227–241.

Jeglic, E.L., Levy-Mack, H.J., Wenzel, A., Beck, A.T., & Brown, G.K. (2004). Attitude toward

therapy and treatment outcome in cognitive therapy for borderline personality disorder. Manu-
script under review.

Kuyken, W. (2004). Cognitive therapy outcome: The effects of hopelessness in a naturalistic out-

come study. Behaviour Research and Therapy, 42, 631– 646.

Layden, M.A. (1998). Layden’s response to the case of Joe: Cognitive therapy for a case of bor-

derline lite. Cognitive and Behavioral Practice, 5, 309–315.

Layden, M.A., Newman, C.F., Freeman, A., & Morse, S.B. (1993). Cognitive therapy of borderline

personality disorder. Boston: Allyn & Bacon.

Newman, C.F. (1998). Cognitive therapy for borderline personality disorder. In L. VandeCreek &

S. Knapp (Eds.), Innovations in clinical practice: A sourcebook (Vol. 116). Sarasota, FL:
Professional Resource Press/Professional Resource Exchange.

Oei, T.P.S., & Sullivan, L.M. (1999). Cognitive changes following recovery from depression in a

group cognitive-behavior therapy program. Australian and New Zealand Journal of Psychia-
try, 33, 407– 415.

Robins, C.J., & Hayes, A.M. (1993). An appraisal of cognitive therapy. Journal of Consulting and

Clinical Psychology, 61, 205–214.

Rush, A.J., Beck, A.T., Kovacs, M., Weissenburger, J., & Hollon, S.D. (1982). Comparison of the

effects of cognitive therapy and pharmacotherapy on hopelessness and self-concept. American
Journal of Psychiatry, 139, 862–866.

Stewart, J.W., Mercier, M.A., Quitkin, F.M., McGrath, P.J., Numes, E., Young, J., et al. (1993).

Demoralization predicts nonresponse to cognitive therapy in depressed outpatients. Journal of
Cognitive Psychotherapy, 7, 105–116.

Veen, G., & Arntz, A. (2000). Multidimensional dichotomous thinking characterizes borderline

personality disorder. Cognitive Therapy and Research, 24, 23– 45.

Westra, H.A., Dozois, D.J.A., & Boardman, C. (2002). Predictors of treatment change and engage-

ment of cognitive-behavioral group therapy for depression. Journal of Cognitive Psychother-
apy: An International Quarterly, 16, 227–241.

516

Journal of Clinical Psychology, April 2006

Journal of Clinical Psychology

DOI 10.1002/jclp