An Overview of Dialectical Behaviour Therapy in the Treatment of Borderline Personality Disorder

Barry Kiehn and Michaela Swales

Patients showing the features of Borderline Personality Disorder as defined in DSM-IV are notoriously difficult to treat (Linehan 1993a). They are difficult to keep in therapy, frequently fail to respond to our therapeutic efforts and make considerable demands on the emotional resources of the therapist, particular when suicidal and parasuicidal behaviours are prominent.

Dialectical Behaviour Therapy is an innovative method of treatment that has been developed specifically to treat this difficult group of patients in a way which is optimistic and which preserves the morale of the therapist.

The technique has been devised by Marsha Linehan at the University of Washington in Seattle and its effectiveness has been demonstrated in a controlled study, the results of which will be summarised later in this paper.


Dialectical Behaviour Therapy is based on a bio-social theory of borderline personality disorder. Linehan hypothesises that the disorder is a consequence of an emotionally vulnerable individual growing up within a particular set of environmental circumstances which she refers to as the ‘Invalidating Environment’.

An ’emotionally vulnerable’ person in this sense is someone whose autonomic nervous system reacts excessively to relatively low levels of stress and takes longer than normal to return to baseline once the stress is removed. It is proposed that this is the consequence of a biological diathesis.

The term ‘Invalidating Environment’ refers essentially to a situation in which the personal experiences and responses of the growing child are disqualified or “invalidated” by the significant others in her life. The child’s personal communications are not accepted as an accurate indication of her true feelings and it is implied that, if they were accurate, then such feelings would not be a valid response to circumstances. Furthermore, an Invalidating Environment is characterised by a tendency to place a high value on self-control and self-reliance. Possible difficulties in these areas are not acknowledged and it is implied that problem solving should be easy given proper motivation. Any failure on the part of the child to perform to the expected standard is therefore ascribed to lack of motivation or some other negative characteristic of her character. (The feminine pronoun will be used throughout this paper when referring to the patient since the majority of BPD patients are female and Linehan’s work has focused on this subgroup).

Linehan suggests that an emotionally vulnerable child can be expected to experience particular problems in such an environment. She will neither have the opportunity accurately to label and understand her feelings nor will she learn to trust her own responses to events. Neither is she helped to cope with situations that she may find difficult or stressful, since such problems are not acknowledged. It may be expected then that she will look to other people for indications of how she should be feeling and to solve her problems for her. However, it is in the nature of such an environment that the demands that she is allowed to make on others will tend to be severely restricted. The child’s behaviour may then oscillate between opposite poles of emotional inhibition in an attempt to gain acceptance and extreme displays of emotion in order to have her feelings acknowledged. Erratic response to this pattern of behaviour by those in the environment may then create a situation of intermittent reinforcement resulting in the behaviour pattern becoming persistent.

Linehan suggests that a particular consequence of this state of affairs will be a failure to understand and control emotions; a failure to learn the skills required for ’emotion modulation’. Given the emotional vulnerability of these individuals this is postulated to result in a state of ’emotional dysregulation’ which combines in a transactional manner with the Invalidating Environment to produce the typical symptoms of Borderline Personality Disorder.

Patients with BPD frequently describe a history of childhood sexual abuse and this is regarded within the model as representing a particularly extreme form of invalidation.

Linehan emphasises that this theory is not yet supported by empirical evidence but the value of the technique does not depend on the theory being correct since the clinical effectiveness of DBT does have empirical support.


Linehan groups the features of BPD in a particular way, describing the patients as showing dysregulation in the sphere of emotions, relationships, behaviour, cognition and the sense of self. She suggests that, as a consequence of the situation that has been described, they show six typical patterns of behaviour, the term ‘behaviour’ referring to emotional, cognitive and autonomic activity as well as external behaviour in the narrow sense.

Firstly, they show evidence of ’emotional vulnerability’ as already described. They are aware of their difficulty coping with stress and may blame others for having unrealistic expectations and making unreasonable demands.

On the other hand they have internalised the characteristics of the Invalidating Environment and tend to show ‘self-invalidation’. They invalidate their own responses and have unrealistic goals and expectations, feeling ashamed and angry with themselves when they experience difficulty or fail to achieve their goals.

These two features constitute the first pair of so-called ‘dialectical dilemmas’, the patient’s position tending to swing between the opposing poles since each extreme is experienced as being distressing.

Next, they tend to experience frequent traumatic environmental events, in part related to their own dysfunctional lifestyle and exacerbated by their extreme emotional reactions with delayed return to baseline. This results in what Linehan refers to as a pattern of ‘unrelenting crisis’, one crisis following another before the previous one has been resolved. On the other hand, because of their difficulties with emotion modulation, they are unable to face, and therefore tend to inhibit, negative affect and particularly feelings associated with loss or grief. This ‘inhibited grieving’ and the ‘unrelenting crisis’ constitute the second ‘dialectical dilemma’.

The opposite poles of the final dilemma are referred to as ‘active passivity’ and ‘apparent competence’. Patients with BPD are active in finding other people who will solve their problems for them but are passive in relation to solving their own problems. On the other hand, they have learned to give the impression of being competent in response to the Invalidating Environment. In some situations they may indeed be competent but their skills do not generalise across different situations and are dependent on the mood state of the moment. This extreme mood dependency is seen as being a typical feature of patients with BPD.

A pattern of self-mutilation tends to develop as a means of coping with the intense and painful feelings experienced by these patients and suicide attempts may be seen as an expression of the fact that life is at times simply does not seem worth living. These behaviours in particular tend to result in frequent episodes of admission to psychiatric hospitals. Dialectical Behaviour Therapy, which will now be described, focuses specifically on this pattern of problem behaviours and in particular, the parasuicidal behaviour.


The term ‘dialectical’ is derived from classical philosophy. It refers to a form of argument in which an assertion is first made about a particular issue (the ‘thesis’), the opposing position is then formulated (the ‘antithesis’ ) and finally a ‘synthesis’ is sought between the two extremes, embodying the valuable features of each position and resolving any contradictions between the two. This synthesis then acts as the thesis for the next cycle. In this way truth is seen as a process which develops over time in transactions between people. From this perspective there can be no statement representing absolute truth. Truth is approached as the middle way between extremes.

The dialectical approach to understanding and treatment of human problems is therefore non-dogmatic, open and has a systemic and transactional orientation. The dialectical viewpoint underlies the entire structure of therapy, the key dialectic being ‘acceptance’ on the one hand and ‘change’ on the other. Thus DBT includes specific techniques of acceptance and validation designed to counter the self-invalidation of the patient. These are balanced by techniques of problem solving to help her learn more adaptive ways of dealing with her difficulties and acquire the skills to do so. Dialectical strategies underlie all aspects of treatment to counter the extreme and rigid thinking encountered in these patients. The dialectical world view is apparent in the three pairs of ‘dialectical dilemmas’ already described, in the goals of therapy and in the attitudes and communication styles of the therapist which are to be described.The therapy is behavioural in that, without ignoring the past, it focuses on present behaviour and the current factors which are controlling that behaviour.


The success of treatment is dependant on the quality of the relationship between the patient and therapist. The emphasis is on this being a real human relationship in which both members matter and in which the needs of both have to be considered. Linehan is particularly alert to the risks of burnout to therapists treating these patients and therapist support and consultation is an integral and essential part of the treatment. In DBT support is not regarded as an optional extra. The basic idea is that the therapist gives DBT to the patient and receives DBT from his or her colleagues. The approach is a team approach.

The therapist is asked to accept a number of working assumptions about the patient that will establish the required attitude for therapy:

1. The patient wants to change and, in spite of appearances, is trying her best at any particular time.

2. Her behaviour pattern is understandable given her background and present circumstances. Her life may currently not be worth living (however, the therapist will never agree that suicide is the appropriate solution but always stays on the side of life. The solution is rather to try and make life more worth living).

3. In spite of this she needs to try harder if things are ever to improve. She may not be entirely to blame for the way things are but it is her personal responsibility to make them different.

4. Patients can not fail in DBT. If things are not improving it is the treatment that is failing.

In particular the therapist must avoid at all times viewing the patient, or talking about her, in pejorative terms since such an attitude will be antagonistic to successful therapeutic intervention and likely to feed into the problems that have led to the development of BPD in the first place.

Linehan has a particular dislike for the word “manipulative” as commonly applied to these patients. She points out that this implies that they are skilled at managing other people when it is precisely the opposite that is true. Also the fact that the therapist may feel manipulated does not necessarily imply that this was the intention of the patient. It is more probable that the patient did not have the skills to deal with the situation more effectively.

The therapist relates to the patient in two dialectically opposed styles. The primary style of relationship and communication is referred to as ‘reciprocal communication’, a style involving responsiveness, warmth and genuineness on the part of the therapist. Appropriate self-disclosure is encouraged but always with the interests of the patient in mind. The alternative style is referred to as ‘irreverent communication’. This is a more confrontational and challenging style aimed at bringing the patient up with a jolt in order to deal with situations where therapy seems to be stuck or moving in an unhelpful direction. It will be observed that these two communication styles form the opposite ends of another dialectic and should be used in a balanced way as therapy proceeds.

The therapist should try to interact with the patient in a way that is:

1. accepting of the patient as she is but which encourages change.

2. centred and firm yet flexible when the circumstances require it.

3. nurturing but benevolently demanding.

The dialectical approach is here again apparent.

There is a clear and open emphasis on the limits of behaviour acceptable to the therapist and these are dealt with in a very direct way. The therapist should be clear about his or her personal limits in relations to a particular patient and should as far as possible make these clear to her from the start. It is openly acknowledged that an unconditional relationship between therapist and patient is not humanly possible and it is always possible for the patient to cause the therapist to reject her if she tries hard enough. It is in the patient’s interests therefore to learn to treat her therapist in a way that encourages the therapist to want to continue helping her. It is not in her interests to burn him or her out. This issue is confronted directly and openly in therapy. The therapist helps therapy to survive by consistently bringing it to the patient’s attention when limits have been overstepped and then teaching her the skills to deal with the situation more effectively and acceptably.

It is made quite clear that the issue is immediately concerned with the legitimate needs of the therapist and only indirectly with the needs of the patient who clearly stands to lose if she manages to burn out the therapist.

The therapist is asked to adopt a non-defensive posture towards the patient, to accept that therapists are fallible and that mistakes will at times inevitably be made. Perfect therapy is simply not possible. It needs to be accepted as a working hypothesis that (to use Linehan’s words) “all therapists are jerks”.


This form of therapy must be entirely voluntary and depends for its success on having the co-operation of the patient. From the start, therefore, attention is given to orienting the patient to the nature of DBT and obtaining a commitment to undertake the work. A variety of specific strategies are described in the Linehan’s book (Linehan 1993a) to facilitate this process.

Before a patient will be taken on for DBT she will be required to give a number of undertakings:

1. To work in therapy for a specified period of time (Linehan initially contracts for one year). and, within reason, to attend all scheduled therapy sessions.

2. If suicidal or parasuicidal behaviours are present, she must agree to work on reducing these.

3. To work on any behaviours that interfere with the course of therapy (‘therapy interfering behaviours’).

4. To attend skills training.

The strength of these agreements may be variable and a “take what you can get approach” is advocated. Nevertheless a definite commitment at some level is required since reminding the patient about her commitment and re-establishing such commitment throughout the course of therapy are important strategies in DBT.

The therapist agrees to make every reasonable effort to help the patient and to treat her with respect, as well as to keep to the usual expectations of reliability and professional ethics. The therapist does not however give any undertaking to stop the patient from harming herself. On the contrary, it should be make quite clear that the therapist is simply not able to prevent her from doing so. The therapist will try rather to help her find ways of making her life more worth living. DBT is offered as a life-enhancement treatment and not as a suicide prevention treatment, although it is hoped that it may indeed achieve the latter.


There are four primary modes of treatment in DBT :

1. Individual therapy

2. Group skills training

3. Telephone contact

4. Therapist consultation

Whilst keeping within the overall model, group therapy and other modes of treatment may be added at the discretion of the therapist, providing the targets for that mode are clear and prioritised.

The individual therapist is the primary therapist. The main work of therapy is carried out in the INDIVIDUAL THERAPY sessions. The structure of individual therapy and some of the strategies used will be described shortly. The characteristics of the therapeutic alliance have already been described.

Between sessions the patient should be offered TELEPHONE CONTACT with the therapist, including out of hours telephone contact. This tends to be an aspect of DBT balked at by many prospective therapists. However, each therapist has the right to set clear limits on such contact and the purpose of telephone contact is also quite clearly defined. In particular, telephone contact is not for the purpose of psychotherapy. Rather it is to give the patient help and support in applying the skills that she is learning to her real life situation between sessions and to help her find ways of avoiding self-injury. Calls are also accepted for the purpose of relationship repair where the patient feels that she has damaged her relationship with her therapist and wants to put this right before the next session. Calls after the patient has injured herself are not acceptable and, after ensuring her immediate safety, no further calls are allowed for the next twenty four hours. This is to avoid reinforcing self-injury.

SKILLS TRAINING is usually carried out in a group context, ideally by someone other that the individual therapist. In the skills training groups patients are taught skills considered relevant to the particular problems experienced by people with borderline personality disorder. There are four modules focusing in turn on four groups of skills:

1. Core mindfulness skills.

2. Interpersonal effectiveness skills.

3. Emotion modulation skills.

4. Distress tolerance skills.

The ‘core mindfulness skills’ are derived from certain techniques of Buddhist meditation, although they are essentially psychological techniques and no religious allegiance is involved in their application. Essentially they are techniques to enable one to become more clearly aware of the contents of experience and to develop the ability to stay with that experience in the present moment.

The ‘interpersonal effectiveness skills’ which are taught focus on effective ways of achieving one’s objectives with other people: to ask for what one wants effectively, to say no and have it taken seriously, to maintain relationships and to maintain self-esteem in interactions with other people.

‘Emotion modulation skills’ are ways of changing distressing emotional states and ‘distress tolerance skills’ include techniques for putting up with these emotional states if they can not be changed for the time being.

The skills are too many and varied to be described here in detail. They are fully described in a teaching format in the DBT skills training manual (Linehan, 1993b).

The therapists receive DBT from each other at the regular THERAPIST CONSULTATION GROUPS and, as already mentioned, this is regarded as an essential aspect of therapy. The members of the group are required to keep each other in the DBT mode and (among other things) are required to give a formal undertaking to remain dialectical in their interaction with each other, to avoid any pejorative descriptions of patient or therapist behaviour, to respect therapists’ individual limits and generally are expected to treat each other at least as well as they treat their patients. Part of the session may be used for ongoing training purposes.


Patients with BPD present multiple problems and this can pose problems for the therapist in deciding what to focus on and when. This problem is directly addressed in DBT. The course of therapy over time is organised into a number of stages and structured in terms of hierarchies of targets at each stage.

The PRE-TREATMENT STAGE focuses on assessment, commitment and orientation to therapy.

STAGE 1 focuses on suicidal behaviours, therapy interfering behaviours and behaviours that interfere with the quality of life, together with developing the necessary skills to resolve these problems.

STAGE 2 deals with post-traumatic stress related problems (PTSD)

STAGE 3 focuses on self-esteem and individual treatment goals.

The targeted behaviours of each stage are brought under control before moving on to the next phase. In particular post-traumatic stress related problems such as those related to childhood sexual abuse are not dealt with directly until stage 1 has been successfully completed. To do so would risk an increase in serious self injury. Problems of this type (flashbacks for instance) emerging whilst the patient is still in stages 1 or 2 are dealt with using ‘distress tolerance’ techniques. The treatment of PTSD in stage 2 involves exposure to memories of the past trauma.

Therapy at each stage is focused on the specific targets for that stage which are arranged in a definite hierarchy of relative importance. The hierarchy of targets varies between the different modes of therapy but it is essential for therapists working in each mode to be clear what the targets are. An overall goal in every mode of therapy is to increase dialectical thinking.

The hierarchy of targets in individual therapy for example is as follows:

1. Decreasing suicidal behaviours.

2. Decreasing therapy interfering behaviours.

3. Decreasing behaviours that interfere with the quality of life.

4. Increasing behavioural skills.

5. Decreasing behaviours related to post-traumatic stress.

6. Improving self esteem.

7. Individual targets negotiated with the patient.

In any individual session these targets must be dealt with in that order. In particular, any incident of self harm that may have occurred since the last session must be dealt with first and the therapist must not allow him or herself to be distracted from this goal.

The importance given to ‘therapy interfering behaviours’ is a particular characteristic of DBT and reflects the difficulty of working with these patients. It is second only to suicidal behaviours in importance. These are any behaviours by the patient or therapist that interfere in any way with the proper conduct of therapy and risk preventing the patient from getting the help she needs. They include, for example, failure to attend sessions reliably, failure to keep to contracted agreements, or behaviours that overstep therapist limits.

Behaviours that interfere with the quality of life are such things as drug or alcohol abuse, sexual promiscuity, high risk behaviour and the like. What is or is not a quality of life interfering behaviour may be a matter for negotiation between patient and therapist.

The patient is required to record instances of targeted behaviours on the weekly diary cards. Failure to do so is regarded as therapy interfering behaviour.


Within this framework of stages, target hierarchies and modes of therapy a wide variety of therapeutic strategies and specific techniques is applied.

The core strategies in DBT are ‘validation’ and ‘problem solving’. Attempts to facilitate change are surrounded by interventions that validate the patient’s behaviour and responses as understandable in relation to her current life situation, and that show an understanding of her difficulties and suffering.

Problem solving focuses on the establishment of necessary skills. If the patient is not dealing with her problems effectively then it is to be anticipated either that she does not have the necessary skills to do so, or does have the skills but is prevented from using them. If she does not have the skills then she will need to learn them. This is the purpose of the skills training.

Having the skills, she may be prevented from using them in particular situations either because of environmental factors or because of emotional or cognitive problems getting in the way. To deal with these difficulties the following techniques may be applied in the course of therapy:

1. Contingency management

2. Cognitive therapy

3. Exposure based therapies

4. Pharmacotherapy

The principles of using these techniques are precisely those applying to their use in other contexts and will not be described in any detail. In DBT however they are used in a relatively informal way and interwoven into therapy. Linehan recommends that medication be prescribed by someone other than the primary therapist although this may not be practical.

Particular note should be made of the pervading application of contingency management throughout therapy, using the relationship with the therapist as the main reinforcer. In the session by session course of therapy care is taken to systematically reinforce targeted adaptive behaviours and to avoid reinforcing targeted maladaptive behaviours. This process is made quite overt to the patient, explaining that behaviour which reinforced can be expected to increase. A clear distinction is made between the observed effect of reinforcement and the motivation of the behaviour, pointing out that such a relationship between cause and effect does not imply that the behaviour is being carried out deliberately in order to obtain the reinforcement. Didactic teaching and insight strategies may also be used to help the patient achieve an understanding of the factors that may be controlling her behaviour.

The same contingency management approach is taken in dealing with behaviours that overstep the therapist’s personal limits in which case they are referred to as ‘observing limits procedures’.

Problem solving and change strategies are again balanced dialectically by the use of validation strategies. It is important at every stage to convey to the patient that her behaviour, including thoughts feelings and actions are understandable, even though they may be maladaptive or unhelpful.

Significant instances of targeted maladaptive behaviour occurring since the last session (which should have been recorded on the diary card) are initially dealt with by carrying out a detailed ‘behavioural analysis’. In particular every single instance of suicidal or parasuicidal behaviour is dealt with in this way. Such behavioural analysis is an important aspect of DBT and may take up a large proportion of therapy time.

In the course of a typical behavioural analysis a particular instance of behaviour is first clearly defined in specific terms and then a ‘chain analysis’ is conducted, looking in detail at the sequence of events and attempting to link these events one to another. In the course of this process hypotheses are generated about the factors that may be controlling the behaviour. This is followed by, or interwoven with, a ‘solution analysis’ in which alternative ways of dealing with the situation at each stage are considered and evaluated. Finally one solution should be chosen for future implementation. Difficulties that may be experienced in carrying out this solution are considered and strategies of dealing with these can be worked out.

It is frequently the case that patients will attempt to avoid this behavioural analysis since they may experience the process of looking in such detail at their behaviour as aversive. However it is essential that the therapist should not be side tracked until the process is completed. In addition to achieving an understanding of the factors controlling behaviour, behavioural analysis can be seen as part of contingency management strategy, applying a somewhat aversive consequence to an episode of targeted maladaptive behaviour. The process can also be seen as an exposure technique helping to desensitise the patient to painful feelings and behaviours. Having completed the behavioural analysis the patient can then be rewarded with a ‘heart to heart’ conversation about the things she likes to discuss.

Behavioural analysis can be seen as a way of responding to maladaptive behaviour, and in particular to parasuicide, in a way that shows interest and concern but which avoids reinforcing the behaviour.

In DBT a particular approach is taken in dealing with the network of people with whom the patient is involved personally and professionally. These are referred to as ‘case management strategies’. The basic idea is that the patient should be encouraged, with appropriate help and support, to deal with her own problems in the environment in which they occur. Therefore, as far as possible, the therapist does not do things for the patient but encourages the patient to do things for herself. This includes dealing with other professionals who may be involved with the patient. The therapist does not try to tell these other professionals how to deal with the patient but helps the patient learn how to deal with the other professionals. Inconsistencies between professionals are seen as inevitable and not necessarily something to be avoided. Such inconsistencies are rather seen as opportunities for the patient to practice her interpersonal effectiveness skills. If she grumbles about the help she is receiving from another professional she is helped to sort this out herself with the person involved. This is referred to as the ‘consultation-to-the-patient strategy’ which, among other things, serves to minimise the so-called “staff splitting” which tends to occur between professionals dealing with these patients.

Environmental intervention is acceptable but only in very specific situations where a particular outcome seems essential and the patient does not have the power or capability to produce this outcome. Such intervention should be the exception rather than the rule.


The effectiveness of DBT has been assessed in two major trials. The first (Linehan et al, 1991) compared the effectiveness of DBT relative to treatment as usual (TAU). The second (Linehan et al, in press) examined the effectiveness of DBT skills training when added to standard community psychotherapy.

In the first randomised controlled trial, there were three main goals:

Firstly, to reduce the frequency of parasuicidal behaviours. This is clearly of importance because of the distressing nature of the behaviour but also because of the increased risk of completed suicide in this group (Stone, 1987).

Secondly, to reduce behaviours that interfere with the progress of therapy (‘therapy interfering behaviours’), as the attrition rate from therapy in borderline women with a history of parasuicidal behaviours is high.

Finally, to reduce behaviours that interfere with the patients’ quality of life. In this study this latter goal was interpreted more specifically as a reduction in in-patient psychiatric days, which is hypothesised to interfere with the patient’s quality of life.

Participants all met DSM-IIIR criteria for BPD, and were matched for number of lifetime parasuicide episodes, number of lifetime admissions to hospital, age and anticipated good or poor prognosis.

There were 22 patients in each group. The experimental group received standard DBT as outlined above. The experience of the patients in the treatment as usual group was variable; some received regular individual psychotherapy, others dropped out of individual therapy whilst continuing to have access to in-patient and day-patient services. All participants were assessed on number of parasuicidal episodes and a range of questionnaire measures of mood. Patients were blindly assessed at pre-treatment, 4, 8 and 12 months and followed up at 6 and 12 months post-treatment. Measures of treatment compliance and other treatment delivered (e.g. in patient psychiatric days) were also taken. At pre-treatment there were no significant differences on any of the measures between the control and experimental groups including demographic criteria.

With regard to the first aim of the trial (i.e. the reduction of suicidal behaviour), during the year of treatment patients in the control group engaged in more parasuicidal acts than DBT patients at all time points. The medical risk for parasuicidal acts was higher in the control group than in the DBT group.

Patients in the DBT group were more likely to start therapy and were more likely to remain in therapy than those in the control group. The one year attrition rate in the DBT group was 16.7% compared to 50% for those in the control group who commenced the year with a new therapist. The DBT patients reported more individual and group therapy treatment hours per week than the TAU group, which reflects the intensive nature of DBT treatment. However, the control patients reported more day treatment hours per week.

With regard to the third goal of the trial, patients in the control group had significantly more inpatient psychiatric days per person than those receiving DBT (38.6 days per year as compared to 8.46 days per year for the DBT group).

These results were considered to indicate the superiority of DBT over treatment as usual. However, one major criticism of the trial is that the variable and patchy therapeutic experience of the control group may be considered to favour DBT. This criticism can be challenged, however, since one of the treatment aims of DBT is to keep the patient in therapy. This it seems to have succeeded in doing. However, it is still pertinent to enquire how well DBT would compare to a consistent treatment alternative. An attempt was made to explore this by comparing the DBT patients with those in the TAU group who received regular individual therapy. It was found that the gains of the patients in the DBT group over the TAU group remained even using this more rigorous comparison.

Despite the more intensive nature of DBT it remained cheaper than TAU, largely because of the reduction in the number of in-patient and day-treatment days received by the DBT patients.

It is of interest that, although the DBT patients showed significant gains across the three areas of interest (number of parasuicides, treatment compliance and inpatient days), there were no between-group differences on any of the questionnaire measures of mood and suicidal ideation. During the follow-up year, patients in the DBT group had higher Global Assessment Scores and a better work performance than the patients in the TAU group. In the first 6 months, DBT patients had fewer suicidal acts, lower anger scores and better self-reported social adjustment than TAU patients. In the final 6 months, DBT patients had fewer in-patient days treatment and better interviewer rated social adjustment than TAU patients.

The second trial had two parts. Firstly, it compared standard community psychotherapy (SCP) plus the group skills component of DBT with SCP alone without added skills training. Secondly, it compared the SCP group from the first part of the present study with the experimental group in the previously described randomised control trial. In this latter comparison, assignment to conditions was not random. However, all subjects were screened in the same way, during the same time frame and were all subject to blind assessment.

The results of the first part of this study indicated that the addition of DBT skills training to SCP for this group of parasuicidal borderline women did not confer any additional therapeutic benefit. In this part of the study the skills training was truly ancillary in that there were no meetings between the individual therapists and the group therapists, nor were any attempts made to assist the patient to generalise the skills learnt in the group to her everyday life.

In the second part of the study there were some pre-treatment differences between the two groups. The DBT patients were less depressed than the control group and reported higher levels of unemployment. These differences were not considered to be particularly important for three reasons. Firstly, depression was not correlated with any of the outcome variables. Secondly, although the lower depression scores favoured the DBT group, the lower unemployment favoured the SCP group. Finally, the levels of depression did not differ between the two groups after the pre-treament point.

During the treatment year there were no significant differences between the groups with regard to staying in therapy. There were some slight differences in the distribution of therapeutic hours, with DBT patients reporting more group treatment hours than the SCP group. Most importantly, however, there were no significant relationships between number of treatment hours and any of the outcome variables. Over the treatment year, standard DBT patients compared to SCP patients had fewer parasuicidal episodes, fewer episodes leading to medical treatment and fewer psychiatric in-patient days. DBT patients also reported less anger than the SCP patients.

This research then provides some evidence for the therapeutic efficacy of DBT. This evidence is primarily derived from one randomised control trial in which DBT was found to be superior on a number of variables to treatment as usual. Clearly this finding requires replication. There is also some evidence to suggest that DBT is superior to other forms of psychotherapy with this group of patients. However, this result comes from a comparison made using only a sub-sample of patients in the randomised trial (Linehan et al, 1991) and from a further comparison between two groups from different studies (Linehan et al, in press). Consequently, the effectiveness of DBT compared to other alternative treatments awaits further exploration. This will remain a challenge, particularly given the high drop-out rates from treatment of this group of patients.


Dialectical Behaviour Therapy then is a novel method of therapy specifically designed to meet the needs of patients with Borderline Personality Disorder and their therapists. It directly addresses the problem of keeping these patients in therapy and the difficulty of maintaining therapist motivation and professional well-being. It is based on a clear and potentially testable theory of BPD and encourages a positive and validating attitude to these patients in the light of this theory. The approach incorporates what is valuable from other forms of therapy, and is based on a clear acknowledgement of the value of a strong relationship between therapist and patient. Therapy is clearly structured in stages and at each stage a clear hierarchy of targets is defined. The method offers a particularly helpful approach to the management of parasuicide with a clearly defined response to such behaviours. The techniques used in DBT are extensive and varied, addressing essentially every aspect of therapy and they are underpinned by a dialectical philosophy that recommends a balanced, flexible and systemic approach to the work of therapy. Techniques for achieving change are balanced by techniques of acceptance, problem solving is surrounded by validation, confrontation is balanced by understanding. The patient is helped to understand her problem behaviours and then deal with situations more effectively. She is taught the necessary skills to enable her to do so and helped to deal with any problems that she may have in applying them in her natural environment. Generalisation outside therapy is not assumed but encouraged directly. Advice and support available between sessions and the patient is encouraged and helped to take responsibility for dealing with life’s challenges herself. The method is supported by empirical evidence which suggests that it is successful in reducing self-injury and time spent in psychiatric in-patient treatment.


Linehan, M.M. (1993a) Cognitive Behavioural Treatment of Borderline Personality Disorder. The Guilford Press, New York and London.

Linehan, M.M. (1993b) Skills Training Manual for Treating Borderline Personality Disorder. The Guilford Press, New York and London.

Linehan, M.M., Armstrong, H.E., Suarez, A., Allmon, D. & Heard, H.L. (1991) Cognitive-behavioural treatment of chronically parasuicidal borderline patients. Archives of General Psychiatry, 48, 1060-1064.

Linehan, M.M., Heard, H.L. & Armstrong, H.E. (in press) Dialectical behaviour therapy, with and without behavioural skills training, for chronically parasuicidal borderline patients.

Stone, M.H. (1987) The course of borderline personality disorder. In Tasman, A., Hales, R.E. & Frances, A.J. (eds) American Psychiatric Press Review of Psychiatry. Washington DC; American Psychiatric Press inc. 8, 103-122.

Barry Kiehn, Consultant Child and Adolescent Psychiatrist, Gwynfa Adolescent Service, Pen-y-Bryn Road, Upper Colwyn Bay, Clwyd, North Wales, LL29 6AL.


Michaela Swales, Chartered Clinical Psychologist, Gwynfa Adolescent Service and Lecturer in the Psychology of Adolescence, University College of North Wales, Bangor, Gwynedd, LL57 2DG.


From November 1995 Psychiaty On-Line; Copyright © 1995