Changes in personality in pre- and post-dialectical behaviour therapy BPD groups - A question of self-control, ♥ ...

[ Pobierz całość w formacie PDF ]
Australian Psychologist
, March 2010; 45(1): 59–66
Changes in personality in pre- and post-dialectical behaviour therapy
borderline personality disorder groups: A question of self-control
JANE DAVENPORT
1
, MILES BORE
1
, & JUDY CAMPBELL
2
1
School of Psychology, University of Newcastle, Newcastle and
2
Wesley Mission Private Hospital, Ashﬁeld,
New South Wales, Australia
Abstract
Dialectical behaviour therapy (DBT) is an evidence-based therapy for people with borderline personality disorder (BPD).
Past research has identiﬁed behavioural changes indicating improved functioning for people who undergo DBT. To date,
however, there has been little research investigating the underlying mechanism of change. The present study utilised a
between-subjects design and self-report questionnaires of Self-Control and the ﬁve factor model of personality and drew
participants from a metropolitan DBT program. We found that pre-treatment participants were signiﬁcantly lower on Self-
Control, Agreeableness and Conscientiousness when compared to both the post-treatment assessment and the norms for
each questionnaire. Neuroticism was signiﬁcantly higher both before and after treatment when compared to the norms.
These ﬁndings suggest that Self-Control may play a role in both the presentation of this disorder and the effect of DBT. High
levels of Neuroticism lend weight to the Linehan biosocial model of BPD development.
Key words:
Borderline personality disorder, dialectical behaviour therapy, personality, personality assessment,
psychological disorders, self-control.
The aim of this study was to investigate the impact of
dialectical behaviour therapy (DBT) on individuals
with borderline personality disorder (BPD) in terms
of changes in self-regulation and personality. There
is now a growing body of research investigating the
effectiveness of DBT (Elwood, Comtois, Holdcraft,
& Simpson, 2002; Linehan, Armstrong, Suarez,
Allmon, & Heard, 1991; van den Bosch, Koeter,
Stijnen, Verheul, & van den Brink, 2005). This
research has established DBT as an effective,
evidence-based therapy for the treatment of BPD
(Robins & Chapman, 2004). To date the research
investigating the effectiveness of this treatment has
focused on measurable behavioural outcomes such
as incidence and severity of self-harm and length
and frequency of hospitalisations. Lynch, Chapman,
Rosenthal, Kuo, and Linehan (2006), however,
noted that there has been very little research examin-
ing the basic processes or mechanisms underlying
patient change.
Lynch et al. (2006) theorised that learning how to
be mindful (through practising core mindfulness, a
key element of DBT) requires the patients to learn
how to control the focus of their attention ‘‘. . . with-
out attempts to ﬁx, alter, suppress or otherwise
avoid’’ (p. 464) their emotions or experiences.
Likewise, the Linehan (1993a) views were that it is
the therapist’s role to assist the patient towards
increasing levels of self-control and self-direction.
Our primary hypothesis was that DBT brings about
a change in self-regulation, thus allowing for the
expression of a more functional, rather than dysfunc-
tional, personality.
Self-control
The idea that patients’ levels of self-control change
due to DBT is an interesting one. The behaviour
of individuals with BPD does appear to be under-
controlled, or as Linehan (1993a) categorised it,
dysregulated. Under-controlled individuals will
usually ‘‘express affect and impulses relatively im-
mediately and directly even when doing so may be
socially or personally inappropriate’’ (Letzring,
Block, & Funder, 2005; p. 397). Self-control is seen
as a desirable characteristic to possess. High levels of
Correspondence: Dr M. Bore, School of Psychology, University of Newcastle, Newcastle, NSW 2308, Australia. E-mail: miles.bore@newcastle.edu.au
ISSN 0005-0067 print/ISSN 1742-9544 online The Australian Psychological Society Ltd
Published by Taylor & Francis
DOI: 10.1080/00050060903280512
J. Davenport et al.
60
self-control have been associated with achievement,
performance, impulse control, adjustment, interper-
sonal relationships and moral emotions (Tangney,
Baumeister, & Boone, 2004). Rothbaum, Weisz, and
Snyder (1982) reported that self-control is the ability
to change and adapt the self to ensure a better or more
optimal ﬁt between the self and the world. In the
diagnostic criteria for BPD, and in Linehan’s alter-
nate descriptors, it would appear, by deﬁnition, that
lack of self-control is a key feature of BPD.
Baumeister, Heatherton, and Tice (1994) identi-
ﬁed four domains of self-control: controlling
thoughts, emotions, impulses, and performance.
These categorisations appear to be very similar to
the Linehan (1993a) clusters of dysregulation in
BPD. It is the Tangney et al. (2004) position that the
term ‘‘self-control’’ might be better conceptualised
as self-regulation, arguing that individuals who
score highly on measures of self-control can mod-
ulate their behaviour dependent upon both internal
and external cues, and environmental demands.
This is typiﬁed by ‘‘. . . that ability to override or
change one’s inner responses, as well as to interrupt
undesired behavioural tendencies (such as impulses)
and refrain from acting on them’’ (Tangney et al.,
2004; p. 274). In light of this information it would be
hypothesised that individuals with BPD would be
low on self-control.
Tangney et al. (2004) considered self-control as
being a component of an individual’s personality.
Tangney et al. (2004) found that self-control
correlated strongly with the trait of Conscientious-
ness from the ﬁve factor model of personality.
proﬁles of individuals with BPD and the ﬁve factor
model and report that there is a positive correlation
between Neuroticism and BPD and a negative
correlation between both Agreeableness and Con-
scientiousness and BPD. That is, this is a population
who are extremely neurotic, disagreeable and not
conscientious. Additionally, Wilberg et al. (1999)
found that BPD subjects produced low Extraversion
but average Openness scores. These ﬁndings com-
plement the Linehan (1993a) belief that emotional
dysregulation is at the core of the difﬁculties for
the individual with BPD. The description of Neuro-
ticism in the ﬁve factor model includes the idea that
high scores on Neuroticism indicate that people have
a chronically high level of emotional
instability
(Costa & Widiger, 2005).
Given that extreme scores on measures of the Big
Five have been found to differentiate between
personality disorders (Wilberg et al., 1999), and that
individuals with BPD have been found to score
highly on Neuroticism, and low on Agreeableness,
Conscientiousness and Extraversion, we predicted
that participants in the present study would produce
this big ﬁve personality proﬁle. A related prediction
was that after undergoing DBT the participants
would then have personality proﬁles on the ﬁve factor
model that are within the normal score range.
The aim of this study was to extend knowledge
of DBT and its impact on BPD by investigating
the underlying changes that occur for people when
they undergo DBT. Although it has been found to be
effective in randomised controlled trials (Elwood
et al., 2002; Linehan et al., 1991; van den Bosch
et al., 2005) there remains the question of what
changes for this population as a consequence of
therapy. Linehan (1993a) theorised that DBT
teaches patients better methods of self-control, thus
decreasing the dysfunction in the individual. Utilis-
ing existing psychometric tools in the area of
Self-control and the ﬁve factor model of personality,
the Linehan (1993a) theory can be investigated. If
Linehan is correct then pre-treatment participants
should show signiﬁcantly different results on our
research measures than post-treatment participants.
Our speciﬁc hypotheses were that pre-DBT parti-
cipants will rate as under-controlled compared to
post-treatment participants, who will score as more
self-controlled; pre-DBT participants will score high
on Neuroticism and low on Conscientiousness,
Agreeableness, and Extraversion compared to the
normal population; post-DBT participants will be
less Neurotic and more Conscientious Agreeable
and Extraverted compared to pre-DBT patients, and
post-DBT participants’ mean scores on Neuroticism,
Conscientiousness, Agreeableness, Openness and
Extraversion will not be signiﬁcantly different to the
general population norms.
Borderline personality disorder and the ﬁve factor model
of personality
The ﬁve factor model has been developed within the
area of normal personality theory and proposes that
there are ﬁve personality dimensions underlying
the variation of personality traits (Wilberg, Urnes,
Friis, Pederson, & Karterud, 1999). Personality traits
are deﬁned as enduring ‘‘dimensions of individual
differences in tendencies to show consistent patterns
of thoughts, feelings and actions’’ (McCrae & Costa,
2003, p. 25). The ﬁve dimensions of this model are
Neuroticism, Extraversion, Openness to Experience,
Agreeableness, and Conscientiousness.
These ﬁve traits have repeatedly been found in
normal samples and cross-culturally (McCrae &
Costa, 1997). Because these ﬁndings have been so
universal, it is argued that extreme variants on the
ﬁve factor model dimensions can differentiate
individuals who have personality pathology from
individuals with normal personality (Wilberg et al.,
1999). Trull, Widiger, Lynam, and Costa (2003)
reviewed the literature on research investigating
Pre- and post-DBT changes in personality
61
Method
Instruments
Participants
Two self-report questionnaires made up the battery
used in this study: the Self-Control Scale (Tangney
et al., 2004) measuring self-control, and; the Inter-
national Personality Item Pool inventory (IPIP)
(Goldberg, 1999), which is a measure of the big ﬁve
personality traits. (A third questionnaire was included
in the battery, the ER89 measure of Ego-Resilience
[Block & Kremen, 1996] but was excluded from
the ﬁnal analysis due to the low Cronbach alpha
coefﬁcient found in our sample a
¼
.63].)
Tangney et al. (2004) created a 36-item self-
control questionnaire that uses a 5-point Likert scale
(1
¼
not at all
,to5
¼
very much
). Tangney et al.
(2004) reported an alpha reliability coefﬁcient of
.85 from their study of Self-Control in 255 under-
graduate students. Items include ‘‘People would
say that I have iron self-discipline’’ and ‘‘I’d be better
off if I stopped to think before acting’’ (reverse
scored).
The IPIP measure of the big ﬁve is freely available
in the public domain (Goldberg, 1999) and is based
on the ﬁve factor model. This scale correlates
highly with the Costa and McCrae (1992) revised
NEO Personality Inventory (NEO-PI) (Buchanan,
Johnson, & Goldberg, 2005). It is a 100-item
questionnaire with participants rating their response
on a 4-point scale (1
¼
deﬁnitely true
,2
¼
true on the
whole
,3
¼
false on the whole
, and 4
¼
deﬁnitely false
).
Each of the ﬁve factors has 20 items, half of which are
reverse scored. Goldberg (1999) reported alpha
reliability coefﬁcients of .85 for Agreeableness, .90
for Conscientiousness, .91 for Extraversion, .91 for
Neuroticism and .89 for Openness to Experience.
Items include ‘‘I have a good word for everyone’’
(Agreeableness), ‘‘I am always prepared’’ (Conscien-
tiousness), ‘‘I feel comfortable around people’’
(Extraversion), ‘‘Often feel blue’’ (Neuroticism),
and ‘‘I Believe art
Participants were drawn from a metropolitan DBT
program provided by a therapy team attached to
a private hospital. This program is based on the
model developed by Marsha Linehan (1993a,b)
and incorporates the four key elements of therapy:
individual psychotherapy, skills training group, tele-
phone counselling/coaching and the therapist con-
sultation group. The inclusion criterion for this study
was that all participants had a primary diagnosis of
BPD.
Two groups were targeted for this study: the
ﬁrst group were individuals who were either on a
waiting list for therapy, or who had started, but not
completed, their ﬁrst 8-week skill-building module.
This group served as the control condition. The
decision to include individuals who had started
therapy in the control condition was made to maxi-
mise the likelihood of reaching sample sizes large
enough to support statistical analysis. Because the
therapy program runs over 14 months the likelihood
of signiﬁcant changes in the individuals who have
yet to complete their ﬁrst module of skills training
is unlikely, and therefore these individuals present
with characteristics and traits more consistent with
their pre-treatment state than individuals who have
successfully graduated from the program.
The second group consisted of individuals who
had successfully graduated from the DBT program
in the past 3 years. These participants represented
the treatment condition. The decision to place
parameters on how long ago people had ﬁnished
therapy was twofold: ﬁrst, to reduce the likelihood
that change was as a result of something other than
therapy; and second, to increase the likelihood that
participant numbers would be large enough to
support analysis.
Research into BPD and DBT has traditionally
been typiﬁed by small sample sizes (e.g., Linehan,
1993,
N
¼
44; Nee & Farman, 2005,
N
¼
19).
Therefore for the current research, consideration
needed to be given to ways that the sample size could
be maximised.
Questionnaires were sent out to 32 people (17
before and 15 after). In this study we had 17
participants (14 female, one male and two who did
not identify their gender); an overall response rate of
56%. The pre-treatment group consisted of seven
individuals: ﬁve women and two who did not identify
their gender (response rate 29%). The mean age was
28.6 years and the standard deviation was 12.9 years.
In the post-treatment group there were 10 partici-
pants (one man and nine women; response rate
65%). The mean age was 31.6 years with a standard
deviation of 8.7 years.
is important’’ (Openness to
Experience).
Procedure
Participants were invited to participate through a
mail-out. The mail-out consisted of a covering letter
(explaining the purpose of the study, consent,
anonymity, and contact numbers for any questions),
the questionnaires, and a pre-addressed reply paid
envelope. Participants were asked to answer ques-
tions as they are now and not to reﬂect on either how
they were in the past or how they would like to be in
the future. A follow-up letter was sent to participants
approximately 8 weeks later. This letter thanked
those who had responded and informed those who
still wished to participate that they could still do so if
they so wished.
J. Davenport et al.
62
The names and addresses of potential participants
were obtained through the database held by the DBT
program. Status in treatment was accessed to allow
allocation to either the control or treatment groups.
The questionnaires were mailed out by staff from the
DBT program in envelopes that had the program’s
return address. This ensured that the researchers had
no access to sensitive patient details and any letters
that were ‘‘returned to sender’’ would not be sent to
the researchers.
reported in the literature. The second author,
however, has used the IPIP big ﬁve scale with several
samples of psychology and medicine students
(
N
¼
1189) and these data were used to provide the
norms for the IPIP big ﬁve scores. The norms for the
Self-Control scale are from a sample of 255 North
American undergraduate psychology students as
reported in Tangney et al. (2004).
Due to the small sample size, the pre-treatment and
post-treatment groups were compared non-parame-
trically using the Kruskal–Wallis test. Although this
test utilises medians for analysis, the means and
medians have been reported in Table 1. The pre-
treatment group produced signiﬁcantly lower
Self-Control, Agreeableness and Conscientiousness
scores than the post-treatment group (
p
.05). No
signiﬁcant differences between the two groups were
found for Extraversion, Neuroticism or Openness to
Experience scores.
The one-sample Wilcoxon signed-rank test, an-
other non-parametric test, was used to compare the
pre-treatment and post-treatment results to the
norms for each construct (Table 1). This analysis
found that, compared to the norm, the pre-treatment
group produced signiﬁcantly lower scores on the
traits of Self-Control, Agreeableness, Conscientious-
ness and Neuroticism. The post-treatment group,
however, signiﬁcantly differed from the norm only on
the measure of Neuroticism. No other signiﬁcant
differences were found.
To further demonstrate the differences observed
between pre- and post-treatment groups and scale
norms, the Z scores for both groups were calculated
(based on the norm means and standard deviations).
The mean Z scores for each group are shown in
Figure 1, which can be viewed as a personality
proﬁle of each group compared to the norm. The
Results
The data from each questionnaire were entered into
a spreadsheet and statistical analysis under-
taken using Minitab version 13 (Minitab Inc.,
Pennsylvania, USA). The questionnaires were scored
by reverse scoring negatively worded items as
indicated in the scoring protocol of each test and
then summing items to produce a score for each con-
struct measured. With regard to unanswered items,
three participants left one question unanswered,
one participant left two questions and another
participant did not answer three questions. For these
participants their relevant trait scores were divided
by the number of items answered and then multiplied
by the number of items presented for that trait.
A Cronbach’s alpha reliability coefﬁcient was
produced for each questionnaire subtest (Table 1).
The Self-Control scale and each of the IPIP big ﬁve
traits all demonstrated acceptable reliability, with
coefﬁcients ranging from .88 to .94. The results here
are consistent with published reliability coefﬁcients
for these scales.
Table 1 lists the means, standard deviations,
medians and norms for each scale. No Australian
norms for the IPIP big ﬁve scale have as yet been
Table 1. Scores vs. stage of treatment for borderline personality disorder
Before treatment
After treatment
Norms
M
M
M
SD
SD
SD
Mdn
Mdn
Mdn
Alpha reliability
Self-Control
26.0
9.4
35.9
11.8
39.2
8.6
.88
23.0
a
37.5
b
39
.
b
Extraversion
55.0
11.1
53.1
9.5
57
10.7
.88
56.0
52.5
58
Agreeableness
49.3
5.5
60.7
11.2
60
7.6
.89
50.0
a
63.0
b
61
.
b
Conscientiousness
43.1
12.9
55.6
12.1
57
8.5
.94
40.0
a
58.5
b
58
.
b
Neuroticism
68.3
5.3
64.3
12.2
45
10.1
.91
68.0
a
67.5
a
44
.
b
Openness to Experience
68.1
8.5
64.2
7.5
61
8.1
.82
73.0
65.0
62
a,b
Different superscripts indicate signiﬁcant differences at
p
.05.
Note
.
Pre- and post-DBT changes in personality
63
The ﬁnal hypothesis was that the scores on
Neuroticism, Conscientiousness, Agreeableness,
and Extraversion, for participants who had completed
DBT, would be no different to the norms. The
data analysis found that this was supported for all
traits except for Neuroticism, in that post-treatment
participants remained as high on Neuroticism as pre-
treatment individuals.
The overall ﬁndings were that signiﬁcant person-
ality differences were observed between the pre- and
post-treatment groups. Participants who had not
yet received DBT had low self-control, were less
agreeable and less conscientious compared to the
post-treatment group and the psychology and med-
icine student scores we used as norms. Participants
who had received DBT were just as self-controlled,
agreeable and conscientious as the norm. But both
pre- and post-treatment participants were highly
neurotic compared to the norm. Our ﬁndings have
implications for our understanding of DBT and what
occurs for individuals who have a personality that is
considered to be disordered.
Figure 1. Z scores for Pre-treatment and post-treatment groups for
Agreeableness (A), Conscientiousness (C), Extraversion (E),
Neuroticism (N), Openness (O) and Self-Control (SC).
post-treatment group can be seen to be more
normative than the pre-treatment group, with the
exception of the trait of Neuroticism.
Self-control
Self-control was assessed in this study using the
Tangney et al. (2004) deﬁnition and assessment tool.
In their deﬁnition Tangney et al. (2004) likened self-
control to self-regulation: ‘‘the ability to regulate the
self strategically in response to goals, priorities, and
environmental demands’’ (p. 314). Tangney et al.
found that higher levels of self-control were positively
correlated to better adjustment, less pathology,
better relationships and interpersonal skills and
more optimal emotional responses. These ﬁndings
have clear links to the difﬁculties the BPD population
experiences. As noted earlier, Linehan (1993a)
theorised that the BPD population have, at the core
of their struggle, an emotional regulation system that
is dysfunctional. This then negatively impacts upon
many areas of an individual’s life, including the
ability to make and maintain relationships, and to be
interpersonally effective. Such negative impacts also
include increased use of mental health services.
The present results support the view of both
Linehan (1993a) and Tangney et al. (2004), in that
the hypothesis that pre-treatment participants would
rate signiﬁcantly lower on the self-control measure
compared to post-treatment participants and the
norms was supported. The results also showed that
self-control scores were higher for the post-treatment
group. What this means for the BPD population is
that their low levels of self-control contribute to the
difﬁculties they have in their daily lives and that DBT
appears to help individuals develop strategies and
insight into their behaviours that subsequently assists
them to develop greater levels of self-control.
Discussion
This study was designed to investigate four hypoth-
eses. These were, ﬁrst, that participants prior to
therapy would rate as under-controlled on a measure
of self-control. Data analysis supported the ﬁrst
hypothesis by ﬁnding that pre-treatment participants
were signiﬁcantly under-controlled when compared
to both post-treatment participants and the ﬁndings
of Tangney et al. (2004), which were used as norms
in this instance.
The second hypothesis was that the pre-treatment
participants would score highly on Neuroticism, and
have low Conscientiousness, Agreeableness, and
Extroversion compared to the normal population
(Australian psychology and medicine students in this
instance). Pre-treatment participants did produce
signiﬁcantly higher Neuroticism scores and lower
Conscientiousness and Agreeableness mean scores
compared to the norms. There was no signiﬁcant
difference, however, between the pre-treatment
group scores and the norms for Extraversion or
Openness to Experience.
The third hypothesis was that post-treatment
participants would be less Neurotic and more
Conscientious, Agreeable, and Extraverted when
compared to pre-treatment participants. This hy-
pothesis was partially supported in that the post-
treatment participants produced signiﬁcantly higher
Conscientiousness and Agreeableness scores. Pre-
and post-treatment Extraversion and Neuroticism
scores, however, were not signiﬁcantly different.
[ Pobierz całość w formacie PDF ]

Changes in personality in pre- and post-dialectical behaviour therapy BPD groups - A question of self-control, ♥ ...

Odnośniki