You are on page 1of 5

Acta Psychiatr Scand 2009: 120: 373377  2009 John Wiley & Sons A/S

All rights reserved ACTA PSYCHIATRICA

DOI: 10.1111/j.1600-0447.2009.01448.x SCANDINAVICA

Clinical overview
Psychotherapy of borderline personality
Zanarini MC. Psychotherapy of borderline personality disorder. M. C. Zanarini
Professor of Psychology, Harvard Medical School,
Objective: Psychotherapy is considered the primary treatment for Boston, MA, USA and Director, Laboratory for the Study
borderline personality disorder (BPD). Currently, there are four of Adult Development, McLean Hospital, Belmont, MA,
comprehensive psychosocial treatments for BPD. Two of these USA
treatments are considered psychodynamic in nature: mentalization-
based treatment and transference-focused psychotherapy. The other
two are considered to be cognitive-behavioral in nature: dialectical
behavioral therapy and schema-focused therapy.
Method: A review of the relevant literature was conducted. Key words: borderline personality disorder;
Results: Each of these lengthy and complex psychotherapies psychotherapy; efficacy
signicantly reduces the severity of borderline psychopathology or at Mary C. Zanarini, Professor of Psychology, Harvard
least some aspects of it, particularly physically self-destructive acts. Medical School, Boston, MA, USA.
Conclusion: Comprehensive, long-term psychotherapy can be a useful E-mail:
form of treatment for those with BPD. However, less intensive and less
costly forms of treatment need to be developed. Invited paper

Clinical recommendations
Comprehensive psychosocial treatments for BPD have proven to be effective.
However, their training requirements and cost make their widespread use problematic.

Additional comments
Treatments that deal with a wider range of borderline symptoms, including those most associated
with psychosocial impairment, are needed.
Such treatments may be adaptations of currently available comprehensive treatments or newer
treatments developed specically for this purpose.

ecacy of pharmacotherapy for BPD (10); with

most types of psychotropic medication being
Borderline personality disorder (BPD) is both a associated with modest levels of symptom reduc-
common and serious psychiatric disorder. It aects tion across a number of sectors of borderline
about 26% of American adults (1, 2). It is also psychopathology.
associated with high levels of psychiatric care (3) as Currently, four comprehensive forms of psycho-
well as high levels of psychosocial impairment (4). therapy have been found to be eective in treating
In addition, treating borderline patients can be those with BPD (1114). Two of these manualized
challenging because of their tendency to prema- forms of treatment are viewed as psychodynamic in
turely terminate psychotherapy (5), regress in nature (11, 12) and two are viewed as more
treatment (6), and engender strong counter-trans- cognitive behavioral in nature (13, 14).
ference reactions in those attempting to help them
Aims of the study
Psychotherapy is considered to be the primary
treatment for BPD (8). This is so because of the The aim of this study is to review and compare the
stably unstable nature of BPD (9). The primacy of results of the controlled trials of the four compre-
psychotherapy is also because of the limited hensive manualized psychosocial treatments for


BPD. Two of these treatments are psychodynamic signicantly fewer suicide attempts, spent signi-
in nature and the other two are cognitive behav- cantly fewer days in the hospital, and were signi-
ioral in nature. cantly more likely not to be taking psychotropic
medications during the 18 months of follow-up. In
addition, psychosocial functioning improved signif-
Material and methods icantly more for those in the partial hospital than
those in the standard treatment group.
A review of the relevant literature was conducted.
In the ve-year follow-up, the partial hospital-
Various features of these randomized controlled
ization group showed statistical superiority to the
trials are noted and compared (e.g., length of trial,
control group on suicidality (23% vs. 74%), BPD
number of subjects, outcome measures, attrition).
diagnostic status (13% vs. 87%), service use (2 vs.
A synthesis of the current state of the eld is
3.5 years of outpatient psychiatric treatment), use
presented as well as suggestions for future treat-
of medication (0.02 vs. 1.90 years taking three or
ment research.
more medications), GAF above 60 (45% vs. 10%),
and vocational status (3.2 vs. 1.2 years). Bateman
Results and Fonagy concluded that those in MBT-focused
partial hospital care followed by MBT-focused
Psychodynamic therapies for BPD
outpatient group psychotherapy did substantially
Mentalization-based treatment. Bateman and better over time than those in TAU. However,
Fonagy developed mentalization-based treatment their vocational functioning in particular was less
(MBT) for patients with BPD (11). This treatment than optimal.
aims to increase a patients curiosity about and More recently, Bateman and Fonagy (18) have
skill in identifying his or her feelings and thoughts been assessing MBT in an outpatient psychother-
and those of other people as well. They speculate apy setting. Preliminary results indicate that this
that this diculty in mentalization arouse because treatment appears to be superior to TAU in some
of diculties in early attachment. aspects but both structured treatments are useful
This manualized treatment was rst imple- for those with BPD.
mented in a partial hospital setting, which served
a relatively poor catchment area of London and Transference-focused psychotherapy. Transference-
which was part of Englands National Health focused psychotherapy (TFP) is based on Kern-
Service (15). Thirty-eight borderline patients were bergs conceptualization of the core problem of
randomized to either psychoanalytically oriented BPD (12). Kernberg suggests that excessive early
individual and group treatment focusing on men- aggression has led the young child to split his or
talization principals (n = 19) or standard psychi- her positive and negative images of him or herself
atric care (n = 19). Both treatments lasted up to and his or her mother (19). This excess aggression
18 months. It was found that those treated with may have been inborn or it may have been caused
MBT-informed partial hospital care had signi- by real frustrations. In either case, the preborder-
cantly better results in the areas of reduced self- line child is unable to merge his or her positive and
mutilatory and suicidal acts, days in hospital, use negative images and attendant aects to achieve a
of psychotropic medications, anxiety and depres- more realistic and ambivalent view of him or
sive symptoms, and psychosocial functioning. herself and others.
These same patients were followed-up in two The primary goal of TFP is to reduce symptom-
separate reports (16, 17). The rst detailed the rst atology and self-destructive behavior through the
18 months after partial hospital treatment or modication of representations of self and others
standard psychiatric care (16). The second fol- as they are enacted in the here and now transfer-
lowed these subjects prospectively for ve years ence. Clarications, confrontations, and transfer-
after both phases of treatment ended (17). ence interpretations are the primary techniques of
In the 18th month follow-up, most of those this twice-weekly psychotherapy.
initially treated with MBT-informed partial hospital Clarkin et al. (20) have conducted a trial of TFP
care were treated with MBT-informed twice weekly in outpatients with BPD. Ninety patients with
group psychotherapy, where as those in the standard BPD were randomized to one of three year-long
care group continued to receive treatment as usual outpatient treatments for BPD: TFP (n = 31),
(TAU), which typically included medication visits dialectical behavioral therapy (DBT) (n = 29), or
with a psychiatrist and case management visits with psychodynamically oriented supportive treatment
a nurse. Those in the partial hospital group had (n = 30). Analyses were only conducted on the 62
signicantly fewer episodes of self-mutilation, made patients (69%) who had completed three waves of

Psychotherapy and BPD

assessment, indicating continuation into the nine to Linehan has also conducted a randomized trial
12 month period of the trial. of DBT vs. community treatment by experts
Patients in all three treatments showed signi- (CTBE) (24). The trial consisted of one-year of
cant positive change in depression, anxiety, global DBT and one year of follow-up. Subjects being
functioning, and social functioning during one year treated with DBT were half as likely to make a
of treatment. Both TFP and DBT were each suicide attempt, required less hospitalization for
associated with improvement in suicidality. Both suicide ideation, and had lower medical risk for all
TFP and supportive treatment were each associ- suicide and self-mutilatory acts together. Subjects
ated with improvement in anger and facets of treated with DBT were also less likely to drop out
impulsivity. Only TFP was signicantly associated of treatment, had fewer psychiatric hospitaliza-
with improvement in irritability, verbal assault, tions, and had fewer emergency room visits. The
and direct assault. Where as TFP was signicantly authors conclude that the results of this study
associated with more outcome measures, there suggest that DBTs eectiveness is not due solely to
were no signicant between-group dierences. general factors associated with expert treatment.
However in a separate report, it was found that Verheul and his colleagues have also conducted
TFP was associated with signicantly greater a study of the ecacy of DBT (25). He compared
improvement than comparison treatments in two DBT to treatment as usual in Holland. DBT
psychodynamically important areas: more secure resulted in a signicantly better retention rate and
attachment and greater reective capacity (21). signicantly greater reductions of self-mutilating
and self-damaging impulsive behaviors, particu-
larly among those with a history of frequent self-
Cognitive behavioral therapies for BPD
mutilation. Taken together, the results of this study
Dialectical behavioral therapy. Linehan (13) has suggest that clinicians independent of Linehans
suggested that the core feature of BPD is emotional Seattle group can successfully provide DBT.
dysregulation. She suggests that this lability may
be because of both inborn biological vulnerabilities Schema-focused therapy. Schema-focused therapy
and an invalidating childhood environment. In any (SFT) is based on the work of Jerey Young (14).
case, the person with BPD is easily upset, becomes Borderline patients are thought to have four
extremely upset very rapidly, and takes a good deal dysfunctional life schemas that maintain their
of time to calm down. Linehan also suggests that psycho-pathology and dysfunction: detached pro-
emotional dysregulation then fuels both the impul- tector, punitive parent, abandoned abused child,
sivity and interpersonal turbulence that is charac- and angry impulsive child. Change is achieved
teristic of those with BPD. through a range of behavioral, cognitive, and
Linehan rst published results of a randomized experiential techniques that focus on the therapeu-
trial of DBT in 1991 (22). This treatment consists tic relationship, daily life outside therapy, and past
of skills groups, individual therapy as well as experiences (including traumatic experiences).
phone coaching for patients and a consultation Recovery in SFT is achieved when dysfunctional
team for clinicians treating them. She found that schemas no longer control the patients life.
DBT was superior to TAU, both of which lasted SFT was compared with TFP at four community
for a year, in reducing the number of episodes of mental health centers in Holland (26). Each ther-
parasuicidal behavior (self-mutilation and suicide apy was conducted for up to three years and each
attempts) and their medical severity. Those treated involved two sessions per week. Forty-ve patients
with DBT were also signicantly more likely to were randomized to SFT and 43 to TFP. Thirty-
stay in individual therapy and had signicantly three completed SFT and 27 were still in treatment.
fewer days in the hospital for psychiatric reasons. In contrast, 21 completed TFP and 19 were still in
In a second study, these subjects were followed treatment.
for another year (23). It was found that during the Statistically and clinically signicant improve-
rst six months of the follow-up, those treated with ments were found in both treatment groups on all
DBT had signicantly less parasuicidal behavior, four study outcomes. These outcomes were: bor-
less anger, and better self-reported social adjust- derline psychopathology, general psychopathol-
ment. During the second six months of follow-up, ogy, quality of life, and SFT TFP personality
those treated with DBT had signicantly fewer concepts. Those treated with SFT also were found
psychiatric inpatient days and better interviewer- to have performed signicantly better than those
rated social adjustment. The authors conclude that treated with TFP on each of these outcomes. In
DBT retains its superiority to TAU over a one-year addition, they were found to have signicantly
follow-up period. greater reduction in symptom severity on six of the


nine DSM-IV criteria for BPD (each of the treatment. One such treatment with proven ecacy
symptoms of a cognitive, impulsive, and interper- is STEPPS a group treatment lasting 20 weeks
sonal nature): identity disturbance, dissocia- (27).
tion paranoia, physically self-destructive acts, It is also clear that only SFT has proven ecacy
other impulsivity, abandonment fears, and stormy with a broad range of BPD symptoms. This may be
relationships. However, no signicant between- because of it was the only study to measure the
group dierences were found on any of the three severity of symptoms from all four sectors of
DSM-IV aective criteria for BPD: anger, empti- borderline psychopathology and this, in turn, may
ness, or moodiness. In addition, SFT was associ- have been as a result of the fact that the other
ated with a signicantly higher retention rate. treatment developers did not have access to such a
measure. However, it is clear that both DBT and
MBT are primarily aimed at reducing the fre-
quency of physically self-destructive acts and their
Taken together, the results of these studies indicate sequelae.
that there are now four manualized psychosocial However, recent research has shown that about
treatments for BPD that have proven to be half of the symptoms of BPD are acute in nature
somewhat eective in decreasing borderline psy- and half are temperamental in nature (28). Accord-
chopathology or at least, selected aspects of it. ing to this research, acute symptoms resolve
These symptomatic improvements have been most relatively rapidly, are specic to BPD, and are
consistent in the areas of self-mutilation and often the reasons for costly forms of treatment,
suicide attempts. This treatment focus makes a such as psychiatric hospitalizations. In contrast,
good deal of sense as these acts are among the most temperamental symptoms are relatively slow to
common reasons for costly psychiatric hospitaliza- resolve, are not specic to BPD, and are associated
tions. They are also a leading indicator of a with ongoing psychosocial impairment. Self-muti-
deteriorating clinical state that can lead to broken lation and help-seeking suicide attempts are exam-
relationships and a substantial decrement in psy- ples of acute symptoms, where as intense anger and
chosocial functioning. profound fears of abandonment are examples of
It is a real advance in the eld that both temperamental symptoms.
clinicians and patients now have four treatments A fully successful comprehensive treatment for
from which to choose. This is important as one BPD must address both acute and temperamental
treatment approach may make more sense to a symptoms. It may be that these comprehensive
particular clinician or a particular patient than the treatments will need to be revamped to address
others. both types of symptoms. Or they might be
In addition, the evidence is not particularly followed by other treatments aimed to these
strong that one treatment is better than another. It more long-lasting symptoms that have such a
is denitely true that DBT is the most thoroughly negative impact on both social and vocational
studied of these treatments and the one most functioning.
widely used, particularly in the US. But it is not Going forward, treatment researchers need to
clear that it is better than these other treatments. establish a standardized set of outcomes for BPD
Rather, it seems that borderline patients are quite and a standardized set of measures to assess these
responsive to a variety of treatments. The take outcomes. Without this, it will remain extremely
home message may be that any reasonable treat- dicult to assess the relative ecacy of these
ment provided by reasonable people in a reason- treatments.
able manner may be benecial to these patients. In conclusion, comprehensive, long-term psy-
Despite the advance that these treatments rep- chotherapy can be a useful form of treatment for
resent, there remains much to be accomplished. It those with BPD. However, less intensive and less
is clear that all of these treatments are both costly forms of treatment need to be developed. In
intensive and relatively long in duration. These addition, psychosocial treatments that are aimed at
also require special training. For these reasons, the quieter, but psychosocially detrimental symp-
they may be beyond the ability of most private toms of BPD are needed.
practitioners, mental health clinics, and even major
medical centers to provide. Less intensive and less
costly forms of psychosocial treatment for BPD are
needed. In order to treat more borderline patients 1. Grant BF, Chou SP, Goldstein RB et al. Prevalence, cor-
relates, disability, and comorbidity of DSM-IV borderline
with a proven manualized psychotherapy for BPD, personality disorder: results from the Wave 2 National
it may be necessary to rely on adjunctive forms of

Psychotherapy and BPD

Epidemiologic Survey on Alcohol and Related Conditions. 17. Bateman AW, Fonagy P. 8-year follow-up of patients trea-
J Clin Psychiatry 2008;69:533545. ted for borderline personality disorder: mentalization-
2. Lezenweger MF, Lane MC, Loranger AW et al. DSM-IV based treatment versus treatment as usual. Am J Psychia-
personality disorders in the national comorbidity survey try 2008;165:631638.
replication. Biol Psychiatry 2007;62:553564. 18. Bateman A Mentalization based treatment: a discussion of
3. Bender DS, Dolan RT, Skodal AE et al. Treatment utili- theory and evidence. Annual Meeting of American Psy-
zation by patients with personality disorders. Am J Psy- chiatric Association, May 19, 2009.
chiatry 2001;158:295302. 19. Kernberg O. Borderline conditions and pathological nar-
4. Skodol AE, Gunderson JG, McGlashan TH et al. Functional cissism. New York: Jason Aronson, 1975.
impairment in schizotypal, borderline, avoidant, and 20. Clarkin JF, Levy KN, Lenzenweger MF et al. Evaluating
obsessive-compulsive personality disorders. Am J Psychi- three treatments for borderline personality disorder: a
atry 2002;159:276283. multiwave study. Am J Psychiatry 2007;164:922928.
5. Gunderson JG, Frank AF, Ronningstam EF et al. Early dis- 21. Levy KN, Meehan KB, Kelly KM et al. Change in attach-
continuation of borderline patients from psychotherapy. ment patterns and reective function in a randomized
J Nerv Ment Dis 1989;177:3842. control trial of transference-focused psychotherapy for
6. Gunderson JG, Kolb JE. Discriminating features of bor- borderline personality disorder. J Consult Clin Psychol
derline patients. Am J Psychiatry 1978;135:792796. 2006;74:10271040.
7. Maltsberger JT, Buie DH. Countertransference hate in the 22. Linehan MM, Armstrong HE, Suarez A. Cognitive-behav-
treatment of suicidal patients. Arch Gen Psychiatry ioral treatment of chronically parasuicidal borderline pa-
1974;30:625633. tients. Arch Gen Psychiatry 1991;48:10601064.
8. American Psychiatric Association. Practice guideline for 23. Linehan MM, Heard HL, Armstong HE. Naturalistic follow-
the treatment of patients with borderline personality dis- up of a behavioral treatment for chronically parasuicidal
order. Washington, DC: American Psychiatric Associa- borderline patients. Arch Gen Psychiatry 1993;50:971974.
tion, 2001. 24. Linehan MM, Comtois KA, Murray AM et al. Two-year
9. Schmideberg M. The treatment of psychopaths and bor- randomized controlled trial and follow-up of dialectical
derline patients. Am J Psychother 1947;1:4570. behavioral therapy vs therapy by experts for suicidal
10. Zanarini MC. Update on pharmacotherapy of borderline behaviors and borderline personality disorder. Arch Gen
personality disorder. Curr Psychiatry Rep 2004;6:5570. Psychiatry 2006;63:757766.
11. Bateman A, Fonagy P. Psychotherapy for borderline per- 25. Verheual R, Van Den Bosch LMC, Koeter MWJ et al.
sonality disorder: mentalization-based treatment. New Dialectical behaviour therapy for women with borderline
York: Oxford University Press, 2004. personality disorder: 12-month randomized clinical trial in
12. Kernberg OF, Selzer MA, Koeningsberg HW et al. Psycho- the Netherlands. Br J Psychiatry 2003;182:135140.
dynamic psychotherapy of borderline patients. New York: 26. Giesen-Bloo J, van Dyck R, Spinhoven S et al. Outpatient
Basic Books, 1989. psychotherapy for borderline personality disorder: ran-
13. Linehan MM. Cognitive-behavioral treatment of borderline domized trial of schema-focused therapy vs. transference-
personality disorder. New York: Guilford Press, 1993. focused psychotherapy. Arch Gen Psychiatry 2006;63:649
14. Young JE, Klosko J, Weishaar ME. Schema therapy: a 658.
practitioners guide. New York: Guilford Press, 2003. 27. Blum N, St. John D, Pfohl B et al. Systems Training for
15. Bateman AW, Fonagy P. The eectiveness of partial hos- Emotional Predictability and Problem Solving (STEPPS)
pitalization in the treatment of borderline personality dis- for outpatients with borderline personality disorder: a
order: a randomized controlled trial. Am J Psychiatry randomized controlled trial and 1- year follow-up. Am J
1999;156:15631569. Psychiatry 2008;165:468478.
16. Bateman AW, Fonagy P. Treatment of borderline person- 28. Zanarini MC, Frankenburg FR, Reich DB et al. The sub-
ality disorder with psychoanalytically oriented partial syndromal phenomenology of borderline personality dis-
hospitalization: an 18-month follow-up. Am J Psychiatry order: a 10-year follow-up study. Am J Psychiatry 2007;
2001;158:3642. 164:929935.