119

10

Psychotherapies

Psychotherapy for personality disorders (PDs), like any other form of

treatment, should be evidence based. For many years, if you wanted to

read about treatment, the only choices you had were to read books describ-

ing the clinical experience of an “expert.” You could also go to workshops

to learn about these ideas. Recommendations were not based on empirical

evidence, however, because there wasn’t any.

Almost all research on the treatment of PDs has studied patients

meeting criteria for borderline PD (BPD), which is also the condition

that most interests clinicians. Starting with the seminal work of Linehan

(1993), a number of innovative methods of treatment have been tested

in clinical trials and shown to be effective. There are now half a dozen

therapies for patients with BPD, each described by an acronym. I am not

convinced, however, that they work in different ways and have different

effects.

http://dx.doi.org/10.1037/14642-011 A Concise Guide to Personality Disorders, by J. Paris Copyright © 2015 by the American Psychological Association. All rights reserved.

13924-11_Ch10_3rdPgs.indd 119 2/17/15 10:08 AM

Co py

ri gh

t Am

er ic

an P

sy ch ol

og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

A CONCISE GUIDE TO PERSONALITY DISORDERS

120

Three caveats should be considered. First, even if one follows a tested

method of psychotherapy, many decisions still need to be made that

derive from experience and skill. Second, research on the therapy of BPD

does not have the heft of literature on other mental disorders, such as

depression or anxiety. Third, there is no evidence that any one method

that has been tested is better than any other. Any therapy that is well

planned will be better than unstructured treatment as usual (TAU)

because it provides patients with external structures that make up for their

inner chaos.

A further limitation, one that applies to any psychotherapy, is that not

every patient can be expected to benefit from treatment. Even in the most

seriously ill, however, rehabilitation can have partial effects. Yet some are

sicker than others, and those who do best usually have “ego strengths”—

a job, an intimate relationship, or both. These areas of positive function-

ing provide patients with a base on which to build skills in other areas.

Although there are always surprises and some patients who are seriously

ill may recover, the observation that better functioning is predictive of

outcome in psychotherapy is a well-known and consistent finding in

research (Bohart & Greaves-Wade, 2013). If you are treating patients

who have no job, no relationship, and no life to speak of, what is there to

work on? If, on the other hand, patients have a life, then therapy benefits

from a laboratory setting in which people can practice what they learn

in treatment sessions.

Although specialized therapies may not be different from each other,

not all psychotherapies applied in practice are equal. What researchers

call TAU tends to be a mess in which patients talk about their problems

to a sympathetic professional but are not given specific guidance in over-

coming dysfunctional emotions, thoughts, and behaviors. This is why

clinical trials always find that specific methods do better than TAU: It

is not hard to do better. Yet when comparisons are made between two

well-structured approaches, differences usually disappear (McMain et al.,

2009; Zanarini, 2009). Patients need planned and structured forms of

therapy, but the brand name may make no difference.

13924-11_Ch10_3rdPgs.indd 120 2/17/15 10:08 AM

Co py

ri gh

t Am

er ic

an P

sy ch ol

og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

PSYCHOTHERAPIES

121

SPECIFIC METHODS OF PSYCHOTHERAPY FOR BPD

Dialectical Behavior Therapy

Dialectical behavior therapy (DBT) was the first evidence-based treatment

for BPD. Developed by Marsha Linehan (1993), this was the first psycho-

therapy for BPD to undergo successful clinical trials (Linehan, Armstrong,

Suarez, Allmon, & Heard, 1991; Linehan et al., 2006). The introduction of

DBT was a turning point in the treatment of the disorder, and its principles

lie at the core of all successful therapy in this population. Here, at last, was

a practical approach that targeted the key traits and symptoms of BPD.

Today, DBT remains the leading evidence-based method of therapy

for patients with BPD. It is an adaptation of cognitive behavioral therapy

(CBT), combined with interventions common to other approaches, but

specifically designed to target the emotion dysregulation that characterizes

BPD, and to reduce impulsive behaviors. It applies chain analysis to inci-

dents leading to self-injury and overdoses—that is, showing patients what

emotions lead up to impulsive behaviors and teaching them alternative

ways of handling dysphoric emotions. DBT also emphasizes empathic

responses to distress that provide validation for the inner experience

of patients. The program consists of weekly individual therapy, group

psychoeducation, telephone availability for coaching, as well as support

through consultation for therapists undertaking these procedures. The

method is an eclectic mix of behavior therapy, CBT, mindfulness based

on Zen Buddhism, and original ideas such as radical acceptance (Linehan,

1993). These techniques have been described in some detail (Linehan,

2014; Linehan & Koerner, 2012).

The first published trial (Linehan et al., 1991) compared 1 year of DBT

with TAU and found DBT to be superior, especially in regard to reductions

in self-harm, overdose, and hospitalization. The question was whether it

was too easy to do better than TAU. For this reason, Linehan et al. (2006)

conducted a second clinical trial in which the comparison group was

“treatment by community experts”—therapists who identified themselves

as interested in BPD and experienced in its treatment. The results again

13924-11_Ch10_3rdPgs.indd 121 2/17/15 10:08 AM

Co py

ri gh

t Am

er ic

an P

sy ch ol

og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

A CONCISE GUIDE TO PERSONALITY DISORDERS

122

favored DBT, with reductions in overdoses and subsequent hospitaliza-

tions within a year, although this time there were no differences between

the groups in the frequency of self-harm. Replication studies in other cen-

ters produced similar results, albeit with higher rates of attrition (Linehan

& Koerner, 2012). A meta-analysis (Kliem, Kröger, & Kosfelder, 2010) sup-

ported the conclusion that DBT is an effective and specific method that is

superior to traditional ways of treating BPD patients.

Although several specific methods of therapy designed for BPD

symptoms have been supported by randomized controlled trials (Paris,

2010b), the strongest evidence supports DBT. The method is a clinical

application of psychological research on emotion regulation (Gross, 2013).

The dysregulation in BPD leads to unstable emotions that are abnormal

responses to interpersonal conflict (Koenigsberg, 2010). That conclusion

has been confirmed by studies of BPD patients using ecological momen-

tary assessment, a technology that allows researchers to track emotional

instability more closely by immediate recording of affective and behav-

ioral responses to life events (Russell et al., 2007; Trull et al., 2008). In

DBT, patients are taught better ways of calming down, during and after

emotional storms, which then reduces the frequency of self-harm and

overdoses.

There are some important unanswered questions about DBT. Although

the original cohort received therapy 20 years ago, it has never been fol-

lowed up, so we do not know whether treated samples maintain their

gains and continue to improve beyond a 1-year posttreatment follow-up.

Also, given the resources required to conduct DBT, it needs to be deter-

mined whether this complex program can be dismantled or streamlined

for greater clinical impact. One report found that a 6-month version of

the therapy can also be effective (Stanley, Brodsky, Nelson, & Dulit,

2007). A treatment lasting for a year (and often more) becomes quickly

inaccessible as waiting lists grow and most patients and their families

cannot afford the expense.

Finally, there is the question of whether DBT is a uniquely effica-

cious treatment for BPD or whether other well-structured approaches

can produce the same results. To address this issue, McMain et al. (2009)

13924-11_Ch10_3rdPgs.indd 122 2/17/15 10:08 AM

Co py

ri gh

t Am

er ic

an P

sy ch ol

og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

PSYCHOTHERAPIES

123

administered DBT for 1 year, with random assignment to a comparison

condition called general psychiatric management, a manualized version of

the American Psychiatric Association (2001) guidelines for the treatment

of BPD. The results of this comparative trial found no differences between

the groups in overdoses, hospitalization, or self-harm. This negative find-

ing had important clinical implications. It suggests that although DBT is

better than most treatments, it can be matched by other therapies that are

designed for this population and that are equally well-structured. Further,

because results were good in both groups, the treatment package used for

the comparison has now been studied on its own, under the name of good

psychiatric management (Gunderson & Links, 2014).

A key question about DBT is whether the results of the treatment are

specific to the method or to the structure. By and large, psychotherapy

research supports a common factors model in which all well-structured

treatments yield similar outcomes (Wampold, 2001). The positive results

of DBT could be due to its high level of structure rather than to its specific

interventions. This supposition was supported by the study by McMain

et al. (2009). So although DBT is clearly better than TAU and somewhat

better than treatment by therapists with experience in treating BPD,

it is not necessarily better than a well-thought-out program of clinical

management.

The popularity of DBT depends on its comprehensiveness, as well as

on its commitment to conducting research to demonstrate its efficacy.

DBT is not the only evidence-based therapy on the market, but it is the

only method that has been tested in multiple clinical trials outside the cen-

ter where it was developed, showing that its efficacy cannot be accounted

for by allegiance effects. The ideas behind DBT are fundamental for any

therapist seeing patients with BPD. It is not a narrowly focused form of

treatment that only deals with cognitive schemata; it also provides vali-

dating responses to current emotional upsets and offers education about

emotion regulation.

However, there is a serious problem with DBT: Its expense makes it

inaccessible. This is mainly because of the length of therapy. DBT has been

tested for a year, but even that length of time is beyond most insurance

13924-11_Ch10_3rdPgs.indd 123 2/17/15 10:08 AM

Co py

ri gh

t Am

er ic

an P

sy ch ol

og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

A CONCISE GUIDE TO PERSONALITY DISORDERS

124

policies or the financial resources of most families. Of even more concern,

Linehan (1993) suggested that even this lengthy period may be only the

first phase of a treatment that could go on for several years. I am reminded

of the story of psychoanalysis in which inevitably incomplete results led to

an interminable course of therapy. DBT is effective but is accessible only

to those who can pay for it. Even if it were properly insured, its length

would still make access a problem: Clinics offering the treatment, even for

12 months, often have extensive waiting lists. It is important to shorten

DBT or to make it intermittent (or do both). This is the only way to provide

service to more patients.

We all owe a debt to Marsha Linehan. I have learned an enormous

amount from her and have applied her principles in all the clinics I lead

that treat BPD. Moreover, Linehan’s recent public statements, acknowl-

edging that she herself once suffered from BPD but recovered, were coura-

geous and have done a great deal to reduce the stigma associated with this

disorder. Nonetheless, treatment for BPD suffers from the perception that

DBT is the only brand that works. Therapists should not feel badly if they

are not in a position to provide DBT in a formal way or to refer patients to

a DBT clinic. In the next chapter, I show that its principles can be incor-

porated into normal clinical practice. In my view, brand names are bad for

therapy. As cognitive theory evolves, it has become a more general term for

what might be called simply psychotherapy (Beck & Haigh, 2014). Livesley

(2012) recommended that DBT give up its brand name and incorporate

its best ideas into a general model of treatment for BPD.

Other Evidence-Based Psychotherapies for BPD

Although other methods have been devised, they do not differ from DBT in

any essential way (Paris, in press). We now examine those that have under-

gone clinical trials.

Mentalization-based treatment (MBT; Bateman & Fonagy, 2006) is

rooted in attachment theory, but the method also has a strong cognitive

component. Its assumption is that BPD patients have trouble recognizing

emotions (their own and those of other people), that is, mentalization.

13924-11_Ch10_3rdPgs.indd 124 2/17/15 10:08 AM

Co py

ri gh

t Am

er ic

an P

sy ch ol

og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

PSYCHOTHERAPIES

125

MBT teaches patients how to do that better. Like most effective programs,

it uses a combination of group and individual therapy. Although devel-

oped by psychoanalysts, it uses a number of cognitive methods similar to

DBT in that patients are taught to recognize their emotions, learn how to

tolerate them, and manage them in more adaptive ways.

MBT was first tested in a randomized controlled trial (RCT) in a day

program lasting 18 months (Bateman & Fonagy, 2001) and found to be

superior to TAU. A second study in a larger sample of outpatients given

18 months of treatment found a decline of both self-reported symptoms

and clinically significant problems, including suicide attempts and hospi-

talization (Bateman & Fonagy, 2009). This is the only method for which

researchers have followed up a cohort for 8 years to determine if the effects

of treatment remain stable, which turned out to be the case (Bateman &

Fonagy, 2004).

MBT needs successful clinical trials in centers outside the hospital where

it originated. Thus far, the only attempt at replication outside the United

Kingdom reported few differences from standard therapy (Jørgensen et al.,

2013). However, Bateman and Fonagy (2008) do not consider MBT as a

“one-and-only” approach but encourage mental health workers to learn

its principles and then apply them in their own clinical settings, without

necessarily following a strict protocol. One can only applaud such open-

mindedness and flexibility. Finally, Bateman and Fonagy (2008) have stated

that the results of their research are not specific to their method but sup-

port any structured approach to psychotherapy. This may be the most con-

sistent finding in this literature (Choi-Kain & Gunderson, 2008).

Transference-focused psychotherapy (TFP; Clarkin, Levy, Lenzenweger,

& Kernberg, 2007) is based on the theories of the psychoanalyst Otto

Kernberg. It differs from other methods in that its focus is on distortions

between therapist and patient in the session, used to illustrate inter-

personal problems elsewhere in the patient’s life. It has thus far under-

gone two clinical trials, one comparing it with DBT, with only minimal

differences (Clarkin, Levy, Lenzenweger, & Kernberg, 2007), and one

comparing it with TAU, to which it was superior (Doering et al., 2010).

TFP aims to generalize what happens in therapy to outside relationships.

13924-11_Ch10_3rdPgs.indd 125 2/17/15 10:08 AM

Co py

ri gh

t Am

er ic

an P

sy ch ol

og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

A CONCISE GUIDE TO PERSONALITY DISORDERS

126

Given the long record of failure for psychodynamic therapy in BPD, one

might consider this approach with caution. Nonetheless, at this point TFP

has about as much support as MBT. It shares the advantage of being struc-

tured and well thought out.

Cognitive analytic therapy is based on similar concepts and can be

considered as another psychodynamic–cognitive hybrid. It applies object

relations theory to establish a firmer sense of self in patients. It has been

tested in a population of adolescents (Chanen & McCutcheon, 2013),

where it was effective, albeit not superior, to a manualized version of

“good clinical care.”

Schema-focused therapy (SFT; Young, Klosko, & Weishaar, 2003) is

another mixture of cognitive and psychodynamic approaches that aims

to modify how patients think about their world (i.e., cognitive schemata),

but it also focuses on the distorting effects of negative childhood experiences.

It has undergone one clinical trial comparing it with transference-focused

psychotherapy (Giesen-Bloo et al., 2006), with only minor differences in

outcome and one trial in which it was superior to TAU (Bamelis, Evers,

Spinhoven, & Arntz, 2014). The problem with schema-focused therapy

is that it is designed to last for 3 years, making it even more inaccessible

than DBT.

Standard CBT has been tested in a study conducted in the United

Kingdom (Davidson, Tyrer, Norrie, Palmer, & Tyrer, 2010). After an aver-

age of 26 sessions, BPD patients did better with cognitive therapy than

with TAU. It also seems likely that CBT for BPD is now being conducted

on broader, more flexible principles. When a Cochrane review (Stoffers,

Völlm, et al., 2012) concluded that the data for cognitive therapy were

“promising,” they were not thinking of standard CBT. Linehan had devel-

oped DBT because of her impression that standard CBT was not effective

for BPD. Yet a large RCT (Davidson et al., 2010) found manualized CBT,

modified to target PD symptoms, was superior to TAU for the treatment

of recurrent deliberate self-harm.

It is important to know that therapy lasting for a few months can

be effective. The evidence for this conclusion was recently reviewed by

Davidson and Tran (2014). What was most striking about these findings

13924-11_Ch10_3rdPgs.indd 126 2/17/15 10:08 AM

Co py

ri gh

t Am

er ic

an P

sy ch ol

og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

PSYCHOTHERAPIES

127

was that the mean length of treatment was 16 sessions. This suggests that

BPD might be treated more rapidly, and less expensively, than by treat-

ments designed to continue for a year or two. Perhaps the most chronic

and severe patients with BPD require several years of therapy, but it makes

no sense to make a long duration the standard of care.

Systems training for emotional predictability and problem solving (STEPPS;

Blum et al., 2008) is a brief and practical program that closely resembles

DBT in its focus on emotion regulation skills. It is designed to supple-

ment TAU, particularly in settings where specialized individual therapies

are not available. STEPPS, based on psychoeducation in groups, has been

supported by clinical trials, with a 1-year follow-up (Blum et al., 2008).

STEPPS is a short-term intervention with psychoeducation conducted

in groups, designed to supplement standard psychotherapy or manage-

ment conducted elsewhere. It is particularly suitable for populations living

in regions where specialized treatment is not available. It is inexpensive

and offers ready accessibility. STEPPS has been subjected to a successful

clinical trial in BPD (Blum et al., 2008), with one replication (Bos, van

der Wel, Appelo, & Verbraak, 2010). It has also undergone one test in

the treatment of antisocial PD (Black, Gunter, Loveless, Allen, & Sieleni,

2010), although one cannot conclude that this makes antisocial PD as

treatable as BPD.

COMMON FACTORS IN THE TREATMENT OF BPD

Although these individual approaches each have useful ideas and tech-

niques, they work through common mechanisms (Paris, in press). A vast

literature shows that the effective factors for outcome in any form of psy-

chotherapy are common rather than specific (Wampold, 2001). Nor is it

necessarily true that patients with BPD can only be seen in specialized

clinics; most benefit from what Gunderson and Links (2014) termed good

psychiatric management. The ideas behind effective therapy are spread-

ing to the wider therapeutic community, and interventions (e.g., teaching

emotion regulation) are becoming part of the armamentarium of thera-

pists of all persuasions.

13924-11_Ch10_3rdPgs.indd 127 2/17/15 10:08 AM

Co py

ri gh

t Am

er ic

an P

sy ch ol

og ic al A ss oc ia ti on . No t fo r fu

rt he

r di

st ri

bu ti

on .

A CONCISE GUIDE TO PERSONALITY DISORDERS

128

It is unfortunate that psychotherapy as a field continues to be defined

by competing methods, many of which use a three-letter acronym. It is

even more unfortunate that clinicians define themselves as practitioners

of any single method. Research can help us get beyond these unnecessary

divisions.

Thus, I agree with Livesley (2012), who suggests that psychotherapy

should be evidence based, not acronym based. Even if some interventions

are partly specific to BPD, we need a single model of therapy to make use

of the best ideas for all sources. To have multiple methods competing for

market share may be good for book sales, but it is not the way to develop

evidence-based practice.

We do not need so many forms of psychotherapy, most of which

resemble each other more in practice than in theory. If therapies based on

so many different ideas and using many different techniques can produce

the same results, they must have a lot in common. One of the main ingre-

dients is structure. Traditional therapies for PD failed because they rely on

unstructured techniques that leave patients adrift. These are not patients

who get better just by being heard and supported. People with BPD also

need specific instruction about emotion regulation, control of impulsiv-

ity, and life skills that can be used to find a job and build a social network.

Although different methods seem to target different aspects of PD, the

failure of comparative tri