Content uploaded by Andrew M Chanen
Author content
All content in this area was uploaded by Andrew M Chanen
Content may be subject to copyright.
Journal of Personality Disorders, 24(6), 786–799, 2010
2010 The Guilford Press
IMPULSIVITY IN BORDERLINE PERSONALITY
DISORDER: REWARD-BASED DECISION-MAKING
AND ITS RELATIONSHIP TO EMOTIONAL DISTRESS
Katherine A. Lawrence, BAppSci, BA (Hons), DPsych (Clinical),
J. Sabura Allen, BA, MS, PhD,
and Andrew M. Chanen, MBBS, MPM, FRANZCP
Impulsivity in Borderline Personality Disorder (BPD) has been defined
as rapid and unplanned action. However, a preference for immediate
gratification and discounting of delayed rewards might better account
for the impulsive behaviors that appear to regulate emotional distress
in BPD. To investigate this, a delay discounting task was administered
to 30 outpatients diagnosed with BPD and 28 healthy community con-
trols (all aged 15–24) before and after a mood induction. Trait impulsiv-
ity was measured with the Barratt Impulsiveness Scale. The results
showed that the BPD group had a greater preference for immediate
gratification and higher rate of discounting the delayed reward than the
control group. Although the mood induction resulted in increased feel-
ings of rejection and anger in all participants, and the rate of delay dis-
counting changed significantly in the control group, the rate of dis-
counting did not change for the BPD group. There was no evidence of
rapid decision-making in the BPD group as response times were similar
between the two groups during both trials. Finally, greater general im-
pulsiveness and nonplanning impulsiveness were associated with
greater rates of discounting in the BPD group. Together these findings
suggest that BPD is characterized by a preference for immediate gratifi-
cation and tendency to discount longer-term rewards. This characteris-
tic appears to exist independent of feelings of rejection and anger,
rather than being reactive to this, and to be related to trait impulsivity.
From the School of Psychology and Psychiatry, Monash University, Melbourne, Australia
(K. A. L., J. S. A.); and Orygen Youth Health Research, Centre, Centre for Youth Mental
Health, The University of Melbourne, Melbourne and Orygen Youth Health Clinical Program,
Northwestern Mental Health, Melbourne, VIC, Australia (A. M. C.).
The authors would like to thank all participants in the study and the staff of the HYPE Clinic.
Particular thanks go to Ms. Emma McDougall, Dr. Martina Jovev, Dr. Louise McCutcheon,
Ms. Suraya Kamsani and Mr. Andrew Lin for assistance with data collection, A/Prof Suzanne
Mitchell for the delay discounting program and Mr. Brendan Lawrence for assistance with
data analysis.
Orygen Youth Health Research Centre is funded by an unrestricted grant from the Colonial
Foundation, Melbourne, Australia. This study also received support from the School of Psy-
chology, Psychiatry and Psychological Medicine at Monash University with the School of Psy-
chology and Psychiatry at Monash University.
Address correspondence to J. Sabura Allen, School of Psychology and Psychiatry, Building
17, Monash University, VIC 3800; E-mail: sabura.allen@med.monash.edu.au
785
786 LAWRENCE, ALLEN, AND CHANEN
Borderline Personality Disorder (BPD) is a severe mental disorder in which
impulsivity is hypothesized to play a central role (Paris, 2007; Skodol et
al., 2002). A range of potentially self-damaging behaviors, commonly re-
ferred to as impulsive behaviors, are frequently observed in BPD, espe-
cially early in the course of the disorder (Stepp & Pilkonis, 2008; Steven-
son, Meares, & Comerford, 2003). These include gambling, irresponsible
money handling, reckless driving, and unsafe sexual practices (American
Psychiatric Association, 2000), as well as problematic substance use, self-
harm, suicidal behavior, and disordered eating (Dougherty et al., 2004;
Rosval et al., 2006; Soloff, Lynch, Kelly, Malone, & Mann, 2000; Trull,
Sher, Minks-Brown, Durbin, & Burr, 2000). These behavior patterns seri-
ously interfere with the individual’s health and quality of life and have
been found to be the strongest predictor of borderline psychopathology at
7-year follow-up (Links, Heslegrave, & van Reekum, 1999). Understanding
what factors might contribute to impulsive behavior in BPD is important
for targeting treatments, yet these factors have not been clearly elucidated.
It is generally assumed that impulsive behaviors are the product of being
an impulsive individual. High-trait impulsivity is frequently found in BPD
samples (Dougherty, Bjork, Huckabee, Moeller, & Swann, 1999; Fossati et
al., 2004; Henry et al., 2001; van Reekum, Links, Mitton, Fedorov, & Pat-
rick, 1996), and has been defined as a tendency for rapid and unplanned
behaviors (Critchfield, Levy, & Clarkin, 2004) and a tendency to act on
urges without regard for the possible consequences (Oquendo & Mann,
2000). These definitions point to the role of behavioral disinhibition in fa-
cilitating impulsive behaviors in BPD, and this has received empirical sup-
port (Hochhausen, Lorenz, & Newman, 2002; Rentrop et al., 2008). Impul-
sive behaviors in BPD have also been conceptualized as attempts to
manage negative emotions (Brown, Comtois, & Linehan, 2002; Linehan,
1993; Trull et al., 2000). In this way, impulsive behaviors are chosen to
provide immediate distraction or relief from intense negative affect. There-
fore, decision-making processes (in addition to processes related to rapid
action) might also be important in conceptualizing why individuals with
BPD engage in impulsive behaviors. Reward-based decision-making in
BPD has been described as a bias for short-term immediate gratification,
combined with a tendency to devalue delayed rewards (Bornovalova, Lejuez,
Daughters, Rosenthal, & Lynch, 2005). This style of decision-making is
referred to as delay discounting (Reynolds, 2006).
While no study has examined delay discounting directly in BPD, several
studies have demonstrated the influence of reward on decision-making in
BPD. Dougherty and colleagues (1999) found that inpatients diagnosed
with BPD (n=14) were less willing to wait for the larger reward during the
second administration of a delay gratification task compared to a group of
nonclinical controls (n=17), suggesting that a reluctance to delay gratifi-
cation is associated with BPD. Other research has used gambling tasks to
assess reward-based decision-making in BPD. For example, Haaland and
Landro (2007) found that 20 inpatients diagnosed with BPD made less
IMPULSIVITY IN BORDERLINE PERSONALITY DISORDER 787
advantageous choices that favored high gain, high-risk options during the
Iowa Gambling Task, in contrast to 15 nonclinical controls. Similarly,
Kirkpatrick and colleagues (2007) found that, in comparison to incarcer-
ated prisoners diagnosed with other personality disorders (n=17), incar-
cerated prisoners with BPD (n=17) made riskier decisions during a com-
puterized gambling task, but only when the probability of winning was
high. Moreover, the BPD group continued to choose risky options even
when they were presented with a risk-free option. While these studies indi-
cate the influence of reward on decision-making in BPD, the interpretation
of this literature is complicated because performance on gambling tasks
relies on the ability to understand and manipulate information relating to
probabilistic outcomes. A more direct test of reward-based decision-making
is required to clarify whether BPD is associated with preference for imme-
diate reward and discounting of delayed rewards.
The relationship between decision-making, rapid responding and trait
impulsivity in BPD is not well understood, partly because these factors are
rarely examined together (Reynolds, Ortengren, Richards, & de Wit, 2006).
For example, Bazanis and colleagues (2002) examined decision-making in
BPD and found rapid decision-making differentiated BPD participants
from healthy controls, yet the decision-making task did not involve re-
ward-based cues. In another study, discounting the value of delayed re-
wards was related to nonplanning impulsiveness (the tendency to be pres-
ent-oriented), however response times were not assessed (Mobini, Grant,
Kass, & Yeomans, 2007). Furthermore, despite the theoretical link be-
tween emotional distress and impulsive behaviors, it is unclear whether
emotional distress might precipitate or accentuate a decision-making bias
toward immediate gratification and discounting of delayed rewards in
BPD. In nonclinical samples, individuals have been shown to prioritize im-
mediate gratification regardless of future consequences, in the context of
emotional distress (Cooper, Wood, Orcutt, & Albino, 2003; Tice, Bratslav-
sky, & Baumeister, 2001), however this has not been tested in a clinical
sample.
In order to clarify the nature of impulsivity in BPD, the current study
investigated reward-based decision-making, in relation to response rates,
trait impulsivity, and emotional distress, in a clinical sample presenting
for treatment early in the course of BPD. The current research focused on
young people recently diagnosed with BPD for two reasons. First, impul-
sivity is reportedly most problematic early in the course of the disorder,
yet the factors hypothesized to contribute to impulsive behaviors have not
been examined in this population. Second, investigating BPD in the early
stages of development reduces the influence of factors associated with the
course of the disorder, such as duration of BPD, treatment, recurrent, or
chronic common mental disorders, cumulative traumatic events and asso-
ciated lifestyle factors (Chanen et al., 2008). The BPD group was compared
to a healthy control group since there is a normative rise in impulsivity
during adolescence (Steinberg, 2005). It was predicted that, compared to
788 LAWRENCE, ALLEN, AND CHANEN
the healthy control group, decision-making in the BPD group would be
associated with (1) a preference for immediate reward and greater dis-
counting of the value of the delayed reward, (2) more rapid responding,
and (3) higher trait impulsivity. We also predicted that a change in mood
would exacerbate the preference for immediate reward, rate of discounting
the delayed reward, and rapid responding more strongly in the BPD group
than in the control group.
METHOD
SAMPLE
Thirty young people were recruited from a specialized early intervention
service for BPD at Orygen Youth Health (Chanen et al., 2009). Participants
were included if they had subsyndromal BPD (4 DSM-IV BPD criteria) or
full syndrome BPD (5 or more DSM-IV BPD criteria) and did not meet
DSM-IV criteria for mental retardation, psychotic disorder, or psychiatric
disorder due to general medical condition. This is consistent with the ad-
mission criteria and early intervention policy at Orygen. Also, four criteria
were found as efficient as five criteria in identifying BPD patients (Nurn-
berg, Hurt, Feldman, & Suh, 1988). Notably, all participants with sub-
syndromal BPD were diagnosed with at least one and most often more
than one other BPD criterion at the sub-threshold level. Twenty-eight
young people, screened for the absence of Axis I disorders and any DSM-
IV features of BPD or Antisocial Personality Disorder (APD), were recruited
from the same community to form the healthy control group.
The BPD group included 27 females (Mage =18.56, SD =3.45) and 3
males (Mage =19.33, SD =1.53) and the control group included 19 fe-
males (Mage =19.16, SD =3.00) and 9 males (Mage =19.22, SD =2.92).
While there was no statistically significant difference in age between the
two groups (t=−0.66, p=0.51), the ratio of female to males was signifi-
cantly higher in the BPD group than in the control group, χ
2
(1) =4.07, p<
0.05. The majority of participants were born in Australia (BPD group =
93%, control group =86%) and came from similar socioeconomic (SES)
backgrounds (50% of the BPD group and 40% of the control group lived in
low SES areas). The BPD group included 9 participants with sub-syndro-
mal BPD and 21 with full syndrome BPD. Nine BPD participants also met
criteria for APD and/or Conduct Disorder. The most common comorbid
Axis I disorders were Major Depressive Disorder (n=19), Post Traumatic
Stress Disorder (n=14), Generalized Anxiety Disorder (n=13), Dysthymic
Disorder (n=11), Panic Disorder without Agoraphobia (n=9) and Panic
Disorder with Agoraphobia (n=8).
MATERIALS
Diagnostic Instruments. Diagnostic assessments for BPD participants
were conducted by their treating clinician. All treating clinicians are spe-
IMPULSIVITY IN BORDERLINE PERSONALITY DISORDER 789
cifically trained in BPD assessment using clinical interview and semi-
structured assessment, including the Structured Clinical Interview for
DSM-IV Axis I Disorders—Patient Edition (SCID I/P; First, Gibbon,
Spitzer, & Williams, 1996), the BPD module of the Structured Clinical In-
terview for DSM-IV Personality Disorders (SCID-II; First, Gibbon, Spitzer,
Williams, & Smith, 1997) and the APD Module of the Diagnostic Interview
for Personality Disorders (DIPD-IV; Zanarini, Frankenberg, Sickel, & Yong,
1996). Control participants were screened using the SCID-I/NP (First,
Spitzer, Gibbon, & Williams, 1996), the BPD module of the SCID-II, and
the APD module of the DIPD-IV. Using all available data, diagnoses for
each individual are then derived through a consensus process with a se-
nior clinician in order to maintain reliability of diagnoses.
Delay Discounting. The preference for immediate rewards and discount-
ing the value of delayed rewards was measured using a modified version
of Mitchell’s (1999) computerized delay discounting task and scoring pro-
cedure. This task also measures response times for every decision made,
providing an opportunity to test for the influence of rapid responding in
reward-based decision-making. The task comprized 138 hypothetical
questions. Each question was presented as a choice between two amounts
of money (the standard item and an alternative item). The standard item
was $1,000 and was offered after one of six delay periods: no delay, 1 day,
1 week, 2 months, 6 months, 1 year. The alternative item varied between
$0 and $1,050 in $50 increments and was presented as available now. The
questions were presented in random order during both administrations.
Participants indicated which of the two items they preferred by clicking on
the Money Now or Money Later button. The time between the presentation
of the question and the participant clicking on the Next Question button
was recorded as the response time for that item.
Trait Impulsivity. Trait impulsivity was measured using the Barratt Im-
pulsiveness Scale, version 11 (BIS-11; Patton, Stanford, & Barratt, 1995).
The BIS-11 is a 30-item questionnaire, designed to assess general impul-
siveness and three facets of impulsivity (motor, nonplanning, and atten-
tional impulsiveness). Motor impulsiveness refers to acting without think-
ing; nonplanning impulsiveness refers to being present-oriented or lacking
in being future oriented; attentional impulsiveness refers to impulsive
thinking and difficulty with concentration. All items were rated on a 4-
point scale (1 =Rarely/Never, 2 =Occasionally, 3 =Often, and 4 =Almost
always/Always). The internal consistency of the BIS-11 within normal and
psychiatric samples has been found to be acceptable (αranged from 0.79–
0.83; Patton et al., 1995).
Mood Induction. Mood induction techniques traditionally induce de-
pressed mood. However, rejection is likely to be a more ecologically rele-
vant mood state to induce in BPD since real or imagined abandonment
is believed to facilitate impulsive behaviors in this population (American
Psychiatric Association, 2000). A mild rejection experience was simulated
by the exclusion condition of Cyberball—a virtual ball-toss computer game
790 LAWRENCE, ALLEN, AND CHANEN
(Williams & Jarvis, 2006). The game appeared as an Internet web page and
depicted four animated ball-tossers standing in a circle. A picture of a
young person was attached to three of the animated ball-tossers to repre-
sent the other players ostensibly involved in the game. The fourth anima-
tion represented the participant and did not have an attached photo. Each
time the ball was thrown to the participant, they were required to click on
one of the other players in order to throw the ball to them. The game was
preprogrammed for all participants to receive the ball twice only at the
beginning of the game. The game concluded after 4 minutes. Mood ratings
of rejection, anger, and sadness were used to test whether Cyberball in-
duced emotional distress. Participants rated “how much” they were feeling
each emotion on a 10cm visual analogue scale from “not at all (rejected)”
to “extremely (rejected).”
PROCEDURE
The study was approved by the North Western Mental Health Research
and Ethics Committee and the Monash University Standing Committee on
Ethics in Research Involving Humans. Following written-informed con-
sent, participants completed the BIS-11. Before commencing the com-
puter-based tasks, a cover story was told to participants, as stipulated
in the standard instructions for the general administration of Cyberball.
Participants were informed that one of the study’s research questions re-
lated to whether mental visualization improved decision-making (by imag-
ining the consequences of a decision). They were told that they would play
a ball toss game over the Internet with three university students known to
the experimenter, as a way of engaging their mental visualization skills in
between the two decision-making tasks. Participants then completed the
delay discounting task twice, before and after the mood induction. Mood
ratings were collected before and after the mood induction. All participants
were debriefed at the end of the procedure regarding the purpose of the
mood induction. There were no adverse reactions to the procedure.
DATA ANALYSIS
The point at which the immediate amount (alternative item) is preferred
over the delayed reward (standard item) is known as the “indifference point.”
Each indifference point was calculated as the midway point between the
last consistently preferred alternative item and the alternative item when
the standard item is consistently preferred. Nonparametric tests were used
for comparisons of indifference points between groups and over time be-
cause some of the indifference points were not normally distributed (skew-
ness >1.5), and the variance between the groups at various delays was
significantly different (Ferguson & Takane, 1989). Pearson product-
moment correlations were used to examine whether trait impulsivity was
associated with rate of discounting. One case was excluded from the corre-
IMPULSIVITY IN BORDERLINE PERSONALITY DISORDER 791
lation analyses as this participant did not complete the self report mea-
sure. kvalues were transformed into natural logarithms to reduce the im-
pact of the outliers on the correlation calculation (Tabachnick & Fidell,
2001).
To assess the rate of discounting the delayed reward, a hyperbolic equa-
tion was fitted to each participant’s indifference point data using the
Solver subroutine in Microsoft Excel (Mitchell, 1999):
V=M
1+k*X
where Vrepresents the value of the standard item indexed by the indiffer-
ence point, Mrepresents the amount of money available from the standard
item ($1,000), kis the fitted parameter indexing the rate of discounting,
and Xrepresents the length of the delay.
RESULTS
A power analysis, based upon the conventional α=0.05, power of 0.80,
and an anticipated large effect size, confirmed that at least 26 people were
required for each group to conduct the following analyses (Cohen, 1992).
Table 1 presents the median indifference points and the difference in dis-
tribution of responses at each delay for the two groups during the first
administration of the delay discounting task. This shows that the distribu-
tion of the indifference point data significantly differed between the two
groups on all the delays except for the 1 week delay. The figures also indi-
cate that BPD participants were prepared to accept less money than the
control participants when the standard item was delayed by 2 months, 6
months, and 1 year.
1
TABLE 1. The Difference in Median Indifference Points and Distribution
of Responses Between the Two Groups at Each Delay
Median
Group Difference in Distribution
Indifference Points
of Responses
BPD Control
Delay (n=30) (n=28) Zscore Mann-Whitney
No Delay 1000 1000 −0.012 U(56) =419.5, p=0.99
1 Day 975 975 −2.689 U(56) =269.5, p<0.01
1 Week 850 950 −1.676 U(56) =314.5, p=0.09
2 Months 400 875 −4.002 U(56) =163.5, p<0.001
6 Months 175 700 −4.318 U(56) =143.0, p<0.001
1 Year 87.5 625 −4.330 U(56) =142.5, p<0.001
Note. Although Day 1 medians are the same, the distributions of the scores
differ significantly between the groups.
1. As there was a significant difference in gender ratio between the two groups, the influence
of gender on delay discounting was assessed by comparing the distribution of indifference
points at each delay and rate of discounting between males and females. Preference for imme-
diate reward and rate of discounting did not significantly differ between males and females
792 LAWRENCE, ALLEN, AND CHANEN
Participants’ experiences of anger, sadness and rejection before and
after the mood induction are presented in Table 2. A repeated measures
analysis of variance revealed that playing Cyberball significantly increased
feelings of rejection and anger within the sample, rejection F(56) =20.66,
p<0.001 and anger F(56) =23.00, p< 0.001. Level of sadness was not
affected by the mood induction. A main effect of group was also found
indicating that rejection, anger and sadness were significantly higher in
the BPD group than the control group before and after the mood induction,
rejection F(56) =41.30, p<0.001, anger F(56) =38.11, p<0.001, and
sadness F(56) =59.20, p<0.001. There were no interaction effects.
To examine the effect of mood on decision-making, the hyperbolic func-
tion was fitted to the indifference points recorded before and after the
mood induction for each group. These results are presented in Figure 1.
Both groups exhibited discounting of the delayed reward, however the BPD
group discounted the delayed reward more steeply (as indicated by the
higher kvalues) than the controls before and after the mood induction,
U(56) =162.00, p<0.001, U(56) =145.00, p<0.001 respectively. A Wilco-
xon-signed ranks test revealed that the rate of discounting (k) among the
BPD group did not significantly alter following the mood induction (z=
−0.34, p=0.73). In contrast, the test indicated a significant reduction in
kfor the control group indicating less discounting of the delayed rewards
following the mood induction (z=−2.10, p<0.05).
There was no significant difference in mean response time between the
groups during either delay discounting administrations (first—BPD group
M=3.16 sec, SD =1.35, Control group M=3.20 sec, SD =0.82; second—
BPD group M=2.15 sec, SD =0.99, Control group M=2.19 sec, SD =
0.62). A repeated measures analysis of variance confirmed that speed of
responding significantly increased during the second administration of the
task, F(56) =272.25, p<0.001, and that this was similar between the two
groups, F(56) =445.36, p=0.89.
Table 3 displays the means and standard deviations for the BIS scores,
TABLE 2. Mean Intensity of Mood Before and After the Mood
Induction for BPD and Control Participants
BPD (n=30) Control (n=28)
Emotion rating Before M(SD) After M(SD ) Before M(SD) After M(SD)
Rejection 2.41 (3.44) 4.15 (3.79) 0.27 (0.75) 1.57 (1.91)
Anger 1.53 (2.16) 2.71 (2.62) 0.46 (0.81) 1.10 (1.51)
Sadness 3.29 (2.73) 3.29 (2.69) 0.37 (0.56) 0.60 (1.51)
in this sample and, therefore, are not reported in detail here. Similarly, as one third of the
BPD group were also diagnosed with Conduct Disorder or Antisocial Personality Disorder,
another disorder known for high levels of impulsivity, the influence of presence of CD or
APD on delay discounting was also assessed. Preference for immediate reward and rate of
discounting did not significantly differ between the two groups. Therefore, the data analysis
reported here is limited to the BPD group as a whole, as compared to the healthy control
group.
IMPULSIVITY IN BORDERLINE PERSONALITY DISORDER 793
FIGURE 1. Median indifference points for the BPD group and control group before and after
the mood induction. The lines represent the hyperbolic functions fitted to these data points
for each group. krepresents the discounting parameter for each function for each group
before and after the mood induction. r
2
represents the variance accounted for by each func-
tion.
and the Pearson correlation coefficients (r) between the rate of discounting
(k) before and after the mood induction and the impulsivity scores for each
group. The results show that the BPD group had significantly higher
scores on general impulsiveness and all BIS subscales than the control
group, and that higher levels of general impulsivity and nonplanning im-
pulsiveness were associated with higher rates of discounting in the BPD
group. This finding remained stable across time regardless of the mood
induction. No correlations were found between trait impulsivity and rate
of discounting for the control group.
TABLE 3. Mean Impulsiveness Scores and Correlations Between Trait
Impulsivity and Rates of Discounting for Both Groups
BPD (n=29) Control (n=28)
Before kAfter kBefore kAfter k
BIS-11 Impulsiveness M(SD)rrM(SD)rr
General 82.6 (10.8) 0.49* 0.50* 62.8 (9.7) −0.12 0.004
Attentional 22.3 (2.8) 0.23 0.22 16.8 (3.0) −0.05 0.06
Motor 28.4 (5.7) 0.34 0.36 21.0 (4.0) −0.01 −0.06
Nonplanning 31.9 (25.0) 0.55* 0.55* 25.0 (4.8) −0.12 0.02
Notes. *p<0.01 Difference between groups on BIS scores: General impulsiveness t(55) =
7.27, p<0.001; Attentional impulsiveness t(55) =7.25, p<0.001; Motor impulsiveness t(55) =
5.70, p<0.001; Nonplanning impulsiveness t(55) =5.20, p<0.001.
794 LAWRENCE, ALLEN, AND CHANEN
DISCUSSION
The main findings from this study clarify the relationships among prefer-
ence for immediate rewards, impulsivity, and emotional distress in BPD.
The preference for immediate rewards and the tendency to discount de-
layed rewards were more prominent in the BPD group than in the healthy
control group. High levels of trait impulsivity clearly differentiated the BPD
group from the control group, and general impulsiveness and nonplanning
impulsiveness were positively correlated with rate of discounting within
the BPD group. However, differences in impulsivity did not extend to the
rate of responding in the delayed discounting task as group response
times were similar. The findings following the mood induction were unex-
pected in two ways. First, a change in mood did not alter reward-based
decision-making in the BPD group, and, second, the rate of discounting
reduced instead of increased in the control group.
While our delay discounting findings are consistent with research in
other impulsive populations (Mitchell, 1999; Petry, 2001, 2002; Petry &
Casarella, 1999), the extent of the preference for immediate rewards and
the discounting of delayed rewards in our BPD sample, as compared to
the healthy controls, is remarkable. Even when the delay was 1 day, BPD
participants were inclined to accept a lesser amount of money because it
was available immediately instead of waiting 24 hours for the full amount.
In contrast, while control participants discounted the value of $1,000 over
long delays, the lowest amounts they were prepared to accept were much
larger than the BPD group’s choices, suggesting a greater degree of self-
control and ability to delay gratification. Many of the BPD participants
joked about how they wished the money was real so they could pay out-
standing mobile phone bills or buy more cigarettes. Such comments reflect
how this style of decision-making is manifested in the day-to-day lives of
individuals with BPD, where immediate gratification often takes prece-
dence over pursuing longer term options such as saving money or quitting
smoking. It is possible that the level of decision-making bias might be ac-
centuated in our sample since younger age is associated with greater im-
pulsivity in BPD (Stevenson et al., 2003). Comparing the rate of discount-
ing within BPD samples on the basis of age might be a valid way of testing
this assertion.
The association between BPD and trait impulsivity is well documented
(Dougherty et al., 1999; Fossati et al., 2004; van Reekum et al., 1996) and
is supported by our study. However, it is the relationship between trait
impulsivity and delay discounting that helps to further clarify the nature
of impulsivity in BPD. In the BPD group, rate of discounting was correlated
with nonplanning impulsiveness but not attentional or motor impulsive-
ness. This is consistent with findings from an undergraduate sample
(Mobini et al., 2007) and suggests that the tendency to be present-oriented
with limited regard for future planning leads to a style of decision-making
that is biased toward immediate gratification. Furthermore, since re-
sponse times did not differ between the two groups, the tendency to seek
IMPULSIVITY IN BORDERLINE PERSONALITY DISORDER 795
immediate rewards over delayed rewards in BPD appears to be a deliberate
and intentional process, not rapid and unplanned as some theorists have
suggested (Critchfield et al., 2004; Hochhausen et al., 2002). Several stud-
ies support a rapid-response model of BPD impulsivity (e.g., Hochhausen
et al., 2002; Rentrop et al., 2008) and the present findings do not neces-
sarily refute this hypothesis. However, in relation to engaging in risky be-
haviors where decision-making is inherent, BPD impulsivity might be bet-
ter characterized as a preference for immediate gratification driven by
nonplanning impulsiveness rather than a tendency for quick action based
on disinhibition.
The nature of the relationship between a change in mood and the rate
of discounting in our sample is surprising. Despite feelings of anger and
rejection increasing at the same rate in both groups, the rate of discount-
ing was only affected in the control group, and contrary to the expected
direction. The control group valued the delayed reward more after their
mood changed, suggesting a shift toward greater self-control rather than
immediate gratification which has been shown in other research (Tice et
al., 2001). Method variance might explain this difference, as the other
study asked participants to imagine themselves causing a car accident in
which a child was killed to induce emotional distress. It is possible that
an experience of social rejection affects decision-making differently.
The finding that the rate of discounting was not affected by a change in
mood in the BPD is unexpected. We predicted that a change in mood
would lead to increased discounting of the delayed reward. It might be
argued that since the BPD group exhibited such a steep rate of discounting
during the first administration of the task, a floor effect was created and
there was no room for further discounting. While this might be true for the
6-month-and-1 year-delay periods, indifference points could have altered
at 1 week and 2 months which would have changed the rate of discount-
ing. Alternatively, given that the rate of discounting decreased in the con-
trol group, and this might be considered an adaptive response to social
rejection, it is interesting that discounting the delayed reward in the BPD
group did not change in this direction either. In this context, lack of re-
sponding to social rejection might be considered maladaptive. Thus, the
problem with BPD might not be rejection-induced impulsivity, but a failure
to reduce impulsivity after rejection.
One caveat to the interpretation of these results is whether the degree
of mood change induced by the experience of rejection was equivalent to
emotional distress. It is possible that the degree of mood change was not
significant enough to alter decision-making as theoretically expected
(Brown et al., 2002; Linehan, 1993; Trull et al., 2000). As increasing the
intensity of the rejection experience was not an ethical option in the cur-
rent research, future studies might monitor the relationship between neg-
ative mood states and impulsive decision-making in naturalistic settings
to complement the current findings. Nonetheless, it is noteworthy that in
a clinical sample renowned for high sensitivity and reactivity to emotional
796 LAWRENCE, ALLEN, AND CHANEN
triggers (such as perceived rejection or abandonment, Gunderson, 2007),
peer rejection did not alter decision-making in our BPD sample. While a
full discussion of this issue is beyond the scope of the current article, it is
an interesting area for further research. It is also possible that decision-
making regarding monetary rewards is not directly related to interpersonal
experience as simulated by Cyberball. Decision-making during a task that
requires cooperation to obtain maximal monetary rewards (cf. King-Casas
et al., 2008) might be a more ecologically-valid measure of impulsive be-
havior within an interpersonal context.
Other limitations to the study include the small clinical sample size, al-
though it is comparable to other published laboratory-based studies in
BPD. Our participants were also diagnosed with a range of comorbid disor-
ders and the lack of a clinical control group does not allow us to comment
further on the specificity of the findings to BPD. However, comorbidity is
the norm in BPD at all ages (Chanen, Jovev, & Jackson, 2007) and exclud-
ing participants with comorbid disorders would have created a highly
atypical sample. The next step in clarifying the nature of impulsivity as
specific to BPD should be to replicate this investigation within a larger
clinical sample and contrast BPD to other impulse control disorders such
as substance related disorders and APD. Depression and level of suicidal-
ity might also influence preference for immediate rewards and rates of dis-
counting in BPD and future studies should control for these variables.
While gender did not influence the current findings, further examination
of gender differences in delay discounting might be another interesting
avenue for future research. The delay discounting paradigm provided a
useful method of examining reward-based decision-making in BPD, how-
ever it also has its limitations. The task is limited to testing preference
for immediate reward which represents positive reinforcement. Negative
reinforcement, such as the reduction of distress, is pertinent in the study
of impulsive behaviors in BPD and should be included in future studies of
BPD impulsivity. Moreover, given the hypothetical nature of the task, the
findings of this study should be tested with other behavioral measures
of impulsivity and in naturalistic settings to further clarify the nature of
impulsivity in BPD.
In summary, these findings suggest that the early phase of BPD is char-
acterized by a tendency to be present-oriented and that this aspect of trait
impulsivity is manifested in reward-based decision making. This implies
that individuals diagnosed with BPD engage in behaviors that are defined
as impulsive in a deliberate manner, rather than rapidly and thought-
lessly, because of the immediate benefits associated with such activities.
In this way, longer-term goals or solutions, that might well be more re-
warding, are devalued or disregarded when immediate gratification is
available. These findings typify why some individuals with BPD might have
difficulty engaging in treatment. Basic requirements of therapy, such as
waiting for appointments and working toward therapeutic goals, do not fit
well with a population that prefers immediate gratification and discounts
IMPULSIVITY IN BORDERLINE PERSONALITY DISORDER 797
the value of delayed rewards. Unexpectedly, an experience of rejection did
not exacerbate this style of decision making in our BPD sample. However,
more intense mood states are likely to increase impulsive behaviors in
some circumstances and future studies are encouraged to further investi-
gate the relationship between mood and impulsivity in this population.
REFERENCES
American Psychiatric Association. (2000). Di- Cooper, M. L., Wood, P. K., Orcutt, H. K., &
Albino, A. (2003). Personality and theagnostic and statistical manual of men-
tal disorders—text revision (4th ed.). predisposition to engage in risky or
problem behaviors during adoles-Washington, DC: American Psychiat-
ric Association. cence. Journal of Personality and So-
cial Psychology, 84, 390–410.Bazanis, E., Rogers, R. D., Dowson, J. H.,
Taylor, P., Meux, C., Staley, C., et al. Critchfield, K. L., Levy, K. N., & Clarkin, J. F.
(2004). The relationship between im-(2002). Neurocognitive deficits in deci-
sion-making and planning of patients pulsivity, aggression, and impulsive-
aggression in borderline personalitywith DSM-III-R borderline personality
disorder. Psychological Medicine, 32, disorder: An empirical analysis of self-
report measures. Journal of Personal-1395–1405.
Bornovalova, M. A., Lejuez, C. W., Daugh- ity Disorders, 18, 555–570.
Dougherty, D. M., Bjork, J. M., Huckabee,ters, S. B., Rosenthal, M. Z., & Lynch,
T. R. (2005). Impulsivity as a common H.C.G., Moeller, G., & Swann, A. C.
(1999). Laboratory measures of ag-process across borderline personality
and substance use disorders. Clinical gression and impulsivity in women
with borderline personality disorder.Psychology Review, 25, 790–812.
Brown, M. Z., Comtois, K. A., & Linehan, Psychiatry Research, 85, 315–326.
Dougherty, D. M., Mathias, C. W., Marsh,M. M. (2002). Reasons for suicide at-
tempts and nonsuicidal self-injury in D. M., Papageorgiou, T. D., Swann,
A. C., & Moeller, F. G. (2004). Labora-women with borderline personality
disorder. Journal of Abnormal Psychol- tory measured behavioural impulsivity
relates to suicide attempt history. Sui-ogy, 111, 198–202.
Chanen, A. M., Jovev, M., & Jackson, H. J. cide and Life-Threatening Behavior, 34,
374–385.(2007). Adaptive functioning and psy-
chiatric symptoms in adolescents with Ferguson, G. A., & Takane, Y. (1989). Statis-
tical Analysis in Psychology and Edu-borderline personality disorder. Jour-
nal of Clinical Psychiatry, 68, 297– cation (6th ed.). New York: McGraw-
Hill Book Company.306.
Chanen, A. M., McCutcheon, L. K., Germano, First, M. B., Gibbon, M., Spitzer, R. L., &
Williams, J.B.W. (1996a). StructuredD., Nistico, H., Jackson, H. J., &
McGorry, P. D. (2009). The HYPE clinical interview for DSM-IV axis I dis-
orders-patient version (SCID-I/P). Newclinic: An early intervention service for
borderline personality disorder. Jour- York: Biometrics Research Depart-
ment, New York State Psychiatric In-nal of Psychiatric Practice, 15, 163–
172. stitute.
First, M. B., Gibbon, M., Spitzer, R. L.,Chanen, A. M., Velakoulis, D., Carison, K.,
Gaunson, K., Wood, S. J., Yuen, H. P., Williams, J.B.W., & Smith, B. L.
(1997). Structured clinical interview foret al. (2008). Orbitofrontal, amygdala
and hippocampal volumes in teenag- DSM-IV personality disorders. New
York: American Psychiatric Press.ers with first-presentation borderline
personality disorder. Psychiatry Re- First, M. B., Spitzer, R. L., Gibbon, M., &
Williams, J.B.W. (1996a). Structuredsearch: Neuroimaging, 163, 116–125.
Cohen, J. (1992). A power primer. Psycholog- clinical interview for DSM-IV axis I dis-
orders, research version, non-patientical Bulletin, 112, 155–159.
798 LAWRENCE, ALLEN, AND CHANEN
edition. (SCID-I/NP). New York: Bio- sivity, delay discounting and cognitive
distortions. Personality and Individualmetrics Research, New York State Psy-
chiatric Institute. Differences, 43, 1517–1528.
Nurnberg, H. G., Hurt, S. W., Feldman, A., &Fossati, A., Barratt, E. S., Carretta, I., Leonar-
di, B., Grazioli, F., & Makkei, C. Suh, R. (1988). Evaluation of diagnos-
tic criteria for borderline personality(2004). Predicting borderline and anti-
social personality disorder features in disorder. The American Journal of Psy-
chiatry, 145, 1280–1283.nonclinical subjects using measures of
impulsivity and aggressiveness. Psy- Oquendo, M. A., & Mann, J. J. (2000). The
biology of impulsivity and suicidality.chiatry Research, 125, 161–170.
Gunderson, J. G. (2007). Disturbed relation- Psychiatric Clinics of North America,
23, 11–25.ships as a phenotype for borderline
personality disorder. American Journal Paris, J. (2007). The nature of borderline
personality disorder: Multiple dimen-of Psychiatry, 164, 1637–1640.
Haaland, V. O., & Landro, N. I. (2007). Deci- sions, multiple symptoms, but one
category. Journal of Personality Disor-sion making as measured with the
Iowa gambling task in patients with ders, 21, 457 –473.
Patton, J. H., Stanford, M. S., & Barratt, E. S.borderline personality disorder. Jour-
nal of International Neuropsychological (1995). Factor structure of the Barratt
Impulsiveness Scale. Journal of Clini-Society, 13, 699–703.
Henry, C., Mitropoulou, V., New, A. S., Koen- cal Psychology, 51, 768–774.
Petry, N. M. (2001). Delay discounting ofigsberg, H. W., Silverman, J., & Siever,
L. J. (2001). Affective instability and money and alcohol in actively using al-
coholics, currently abstinent, and con-impulsivity in borderline personality
and bipolar II disorders: Similarities trols. Psychopharmacology, 154, 243–
250.and differences. Journal of Psychiatric
Research, 35, 307–312. Petry, N. M. (2002). Discounting of delayed
rewards in substance abusers: Rela-Hochhausen, N. M., Lorenz, A. R., & Newman,
J. P. (2002). Specifying the impulsivity tionship to antisocial personality dis-
order. Psychopharmacology, 162, 425–of female inmates with borderline per-
sonality disorder. Journal of Abnormal 432.
Petry, N. M., & Casarella, T. (1999). Exces-Psychology, 111, 495–501.
King-Casas, B., Sharp, C., Lomax-Bream, L., sive discounting of delayed rewards in
substance abusers with gamblingLohrenz, T., Fonagy, P., & Montague,
P. R. (2008). The rupture and repair of problems. Drug and Alcohol Depen-
dence, 56, 25–32.cooperation in borderline personality
disorder. Science, 321, 806–810. Rentrop, M., Backenstrass, M., Jaentsch, B.,
Kaiser, S., Roth, A., Unger, J., et al.Kirkpatrick, T., Joyce, E., Milton, J., Duggan,
C., Tyrer, P., & Rogers, R. D. (2007). (2008). Response inhibition in border-
line personality disorder: PerformanceAltered emotional decision-making in
prisoners with borderline personality in a go/nogo task. Psychopathology,
41, 50–57.disorder. Journal of Personality Disor-
ders, 21, 243–261. Reynolds, B. (2006). A review of delay-
discounting research with humans:Linehan, M. M. (1993). Cognitive-behavioural
treatment of borderline personality dis- Relations to drug use and gambling.
Behavioural Pharmacology, 17, 651–order. New York: The Guilford Press.
Links, P. S., Heslegrave, R., & van Reekum, 667.
Reynolds, B., Ortengren, A., Richards, J. B.,R. (1999). Impulsivity: Core aspect of
borderline personality disorder. Jour- & de Wit, H. (2006). Dimensions of im-
pulsive behavior: Personality and be-nal of Personality Disorders, 13, 1–9.
Mitchell, S. H. (1999). Measures of impulsiv- havioral measures. Personality and In-
dividual Differences, 40, 305–315.ity in cigarette smokers and non-
smokers. Psychopharmacology, 146, Rosval, L., Steiger, H., Bruce, K., Israel, M.,
Richardson, J., & Aubut, M. (2006).455–464.
Mobini, S., Grant, A., Kass, A. E., & Yeomans, Impulsivity in women with eating dis-
orders: Problem of response inhibition,M. R. (2007). Relationships between
functional and dysfunctional impul- planning or attention? International
IMPULSIVITY IN BORDERLINE PERSONALITY DISORDER 799
Journal of Eating Disorders, 39, 590– ing Multivariate Statistics (4th ed.).
Boston: Allyn and Bacon.593.
Skodol, A. E., Siever, L. J., Livesley, W. J., Tice, D. M., Bratslavsky, E., & Baumeister,
R. F. (2001). Emotional distress regu-Gunderson, J. G., Pfohl, B., & Widiger,
T. A. (2002). The borderline diagnosis lation takes precedence over impulse
control: If you feel bad, do it! Journal ofII: Biology, genetics, and clinical
course. Biological Psychiatry, 51, 951– Personality and Social Psychology, 80,
53–67.963.
Soloff, P. H., Lynch, K. G., Kelly, T. M., Trull, T. J., Sher, K. J., Minks-Brown, C.,
Durbin, J., & Burr, R. (2000). Border-Malone, K. M., & Mann, J. J. (2000).
Characteristics of suicide attempts of line personality disorder and sub-
stance use disorders: A review and in-patients with major depressive episode
and borderline personality disorder: A tegration. Clinical Psychology Review,
20, 235–253.comparative study. American Journal
of Psychiatry, 157, 601–608. van Reekum, R., Links, P. S., Mitton, J. E.,
Fedorov, C., & Patrick, J. (1996). Im-Steinberg, L. (2005). Cognitive and affective
development in adolescence. Trends in pulsivity, defensive functioning, and
borderline personality disorder. Cana-Cognitive Sciences, 9, 69–74.
Stepp, S. D., & Pilkonis, P. (2008). Age- dian Journal of Psychiatry, 41, 81–84.
Williams, K. D., & Jarvis, B. (2006). Cyber-related differences in individual DSM
criteria for borderline personality dis- ball: A program for use in research on
interpersonal ostracism and accep-order. Journal of Personality Disorders,
22, 427– 432. tance. Behavior Research Methods, 38,
174–180.Stevenson, J., Meares, R., & Comerford, A.
(2003). Diminished impulsivity in Zanarini, M. C., Frankenberg, F. R., Sickel,
A. E., & Yong, L. (1996). The diagnosticolder patients with borderline person-
ality disorder. American Journal of interview for DSM-IV personality disor-
ders (DIPD-IV). Belmont MA: McLeanPsychiatry, 160, 165–166.
Tabachnick, B. G., & Fidell, L. S. (2001). Us- Hospital.
