Mentalization based treatment for borderline personality disorder

Abstract

Mentalizing is the process by which we make sense of each other and ourselves, implicitly and explicitly, in terms of subjective states and mental processes. It is a profoundly social construct in the sense that we are attentive to the mental states of those we are with, physically or psychologically. Given the generality of this definition, most mental disorders will inevitably involve some difficulties with mentalization, but it is the application of the concept to the treatment of borderline personality disorder (BPD), a common psychiatric condition with important implications for public health, that has received the most attention. Patients with BPD show reduced capacities to mentalize, which leads to problems with emotional regulation and difficulties in managing impulsivity, especially in the context of interpersonal interactions. Mentalization based treatment (MBT) is a time-limited treatment which structures interventions that promote the further development of mentalizing. It has been tested in research trials and found to be an effective treatment for BPD when delivered by mental health professionals given limited additional training and with moderate levels of supervision. This supports the general utility of MBT in the treatment of BPD within generic mental health services.

Borderline personality disorder (BPD) is a complex and serious mental disorder characterized by a pervasive pattern of difficulties with emotion regulation and impulse control, and instability both in relationships and in selfimage 1. It represents a serious public health problem, because it is associated with suicide attempts and self harm, both of which are consistent targets of mental health services. Recurrent suicidal behaviour is reported in 69-80% of patients with BPD, and suicide rates are estimated to be up to 10% 2.

BPD is a common condition that is thought to occur globally with a prevalence of 0.2-1.8% in the general population 3. Higher prevalence rates are found in clinical populations. Moran et al 4 found a prevalence rate of 4-6% among primary care attenders, suggesting that people with BPD are more likely to visit their general practitioner. Chanen et al 5 reported a prevalence rate of 11% in adolescent outpatients and 49% in adolescent inpatients. The highest prevalence has been found in people requiring the most intensive level of care, with a rate of 60-80% among patients in forensic services 6-7.

The high prevalence and increased suicide rate in patients with BPD make an unassailable argument that effective treatment needs to be developed and that treatment has to be widely available. Whilst a number of treatments for BPD have been shown to be moderately effective in randomized controlled trials, it remains of considerable concern that most of them require extensive training, making them unavailable to most patients. Mentalization based treatment (MBT) was developed with this in mind. It requires relatively little additional training on top of general mental health training, and has been implemented in research studies by community mental health professionals, primarily nurses, with limited training given modest levels of supervision.

WHAT IS MENTALIZATION?

The term mentalization grew out of the Ecole Psychosomatique de Paris and to some extent was operationalized by developmental researchers investigating theory of mind 8. It was first used by Fonagy in 1989 9 in a broader way and has since been developed in relation to understanding a number of mental disorders.

Mentalization, or better mentalizing, is the process by which we make sense of each other and ourselves, implicitly and explicitly, in terms of subjective states and mental processes. It is a profoundly social construct in the sense that we are attentive to the mental states of those we are with, physically or psychologically. Given the generality of this definition, most mental disorders will inevitably involve some difficulties with mentalization. In fact, we can conceive of most mental disorder as the mind misinterpreting its own experience of itself, thus ultimately a disorder of mentalization. However, the key issue is whether the dysfunction is core to the disorder and/or a focus on mentalization is heuristically valid, i.e. provides an appropriate domain for therapeutic intervention.

While mentalizing theory is being applied to a number of disorders (e.g., posttraumatic stress disorder 10, eating disorders 11 and depression 12), in a number of contexts (e.g., inpatient, partial hospital, and outpatient facilities), and in different groups of patients (e.g., adolescents, families, substance abusers), the treatment method is most clearly organized as a therapy for BPD 13. It is only in this condition that clear empirical support with randomized controlled trials 14-15 is available.

In BPD, a fragile mentalizing capacity vulnerable to social and interpersonal interaction is considered a core feature of the disorder. If a treatment is to be successful, it must either have mentalization as its focus or at the very least stimulate development of mentalizing as an epiphenomenon.

The failure of adult mental processing in borderline states had been apparent to most clinicians, but none had identified the primary difficulty as a loss of mentalizing arising from early development. The simple basic suggestion we made was that representing self and others as thinking, believing, wishing or desiring did not arrive at age 4 as a consequence of maturation, but rather was a developmental achievement that was profoundly rooted in the quality of early object relations. Its predictable vulnerability to disappearance under stress in borderline conditions was seen as an appropriate focus for psychodynamically oriented psychological intervention, even though concerns had been expressed over many decades about the use of psychodynamic therapy in the treatment of BPD. These began as long ago as 1938, when an American psychoanalyst, Adolph Stern, identified a group of patients, now considered to have had BPD, who did not respond to classical psychoanalytic treatment 16. He later described modifications of psychotherapy for his borderline group that remain relevant today 17.

THE DEVELOPMENT OF MENTALIZING

Mentalizing theory is rooted in Bowlby’s attachment theory and its elaboration by contemporary developmental psychologists, whilst paying attention to constitutional vulnerabilities. There is suggestive evidence that borderline patients have a history of disorganized attachment, which leads to problems in affect regulation, attention and self control 18-19. It is our suggestion that these problems are mediated through a failure to develop a robust mentalizing capacity.

Our understanding of others critically depends on whether as infants our own mental states were adequately understood by caring, attentive, non-threatening adults. The most important cause of disruption in mentalizing is psychological trauma early or late in childhood, which undermines the capacity to think about mental states or the ability to give narrative accounts of one’s past relationships. Building on the accumulating evidence from developmental psychopathology, the mentalization theory of BPD first suggests that individuals are constitutionally vulnerable and/or exposed to psychological trauma; second, that both these factors can undermine the development of social/cognitive capacities necessary for mentalization via neglect in early relationships 20, especially where the contingency between their emotional experience and the caregiver’s mirroring is non-congruent 21; third, that this results in an hypersensitive attachment system within interpersonal contexts; and fourth, that this leads to the development of an enfeebled ability to represent affect and effortfully control attentional capacity 22.

Given the known continuity of attachment styles over time, residues of attachment problems of childhood might be expected to be apparent in adulthood. The adult attachment literature in relation to BPD has been reviewed by Levy 23. While the relationship between BPD diagnosis and a specific attachment category is not obvious, there is little doubt that BPD is strongly associated with insecure attachment (only 6-8% of BPD patients are coded as secure). It appears that early attachment insecurity is a relatively stable characteristic of BPD patients, particularly in conjunction with subsequent negative life events 24.

MENTALIZATION BASED TREATMENT

The focus in treatment of BPD needs to be on stabilizing the sense of self and helping the patient maintain an optimal level of arousal in the context of a well-managed, i.e. not too intense and yet not too detached, attachment relationship between patient and therapist. The patient with BPD is exquisitely sensitive to all interpersonal interactions. So, the therapist needs to be aware that therapy, an interpersonal interaction, inevitably will provoke anxiety related to loss of a sense of self and that the ensuing emotional experiences will rapidly threaten to overwhelm the patient’s mental capacities, leading to escalating emotions and inability to accurately understand others’ motives. Psychiatrists and other mental health professionals also need to be aware of this sensitivity if they are to avoid iatrogenic interactions with patients with BPD. Inpatient hospital admission, for example, is an intense emotional experience for all patients and, unless carefully managed, will make patients with BPD worse by overstimulating their attachment processes. This overstimulation in treatment may account for the poor long-term outcomes of patients with BPD when unmodified intensive treatments were offered 25(25).

Patients with BPD have a vulnerability in regulating emotional responses and generating effective strategies for controlling their thoughts and feelings, which challenges their capacity for thinking about their own actions in terms of subtle understandings of their thoughts and feelings. They slip into what superficially could be described as a kind of mindless state, both in relation to others and to themselves. Of course, the story turns out to be more complicated than this, because these incapacities, palpable at certain times, are not always evident. But, at moments of emotional distress, particularly distress triggered by actual or threatened loss, the capacity for mentalization is most likely to apparently evaporate. The question is how this understanding and the clinical observations can usefully be translated into a therapeutic approach that could be helpful given the prevalence and severity of this clinical problem within a public healthcare system.

To this end, we defined some core underpinning techniques to be used in the context of group and individual therapy and labeled them MBT 13-26. Only three important aspects of treatment will be considered here, namely the aim of interventions, the therapeutic stance, and mentalizing the transference.

Aims of interventions in MBT

The initial task in MBT is to stabilize emotional expression, because without improved control of affect there can be no serious consideration of internal representations. Although the converse is also true, identification and expression of affect are targeted first because they represent an immediate threat to continuity of therapy as well as potentially to the patient’s life. Uncontrolled affect leads to impulsivity, and only once this affect is under control is it possible to focus on internal representations and to strengthen the patient’s sense of self.

The aim and the actual outcome of an intervention are more important in MBT than the type of intervention itself. The primary aim of any intervention has to be to reinstate mentalizing when it is lost or to help to maintain it in circumstances when it might be lost or is being lost. Any intervention that succeeds in these aims may be used in MBT. As a result of this, MBT takes a more permissive approach to interventions than most other therapies, giving it a plurality in terms of technique which might account for its popularity and appeal to practitioners from different schools as well as the limited amount of training required before practitioners begin using it in their everyday practice. We do not ask that practitioners learn a new model of therapy from the beginning, but that they modify their current practice focussing on mentalizing rather than behaviours, cognitions, or insight. We do, however, ask that they undertake to develop a particular therapeutic stance and implement a series of steps to try to engage the patient in a process of mentalizing, firstly using some generic psychotherapy techniques such as empathy, support and clarification, and then moving on to other interventions specifically designed to “stress” the attachment relationship within controlled conditions, which includes a focus on the patienttherapist relationship through “mentalizing the transference”.

Therapeutic stance

The therapist’s mentalizing therapeutic stance should include: a) humility deriving from a sense of “notknowing”; b) patience in taking time to identify differences in perspectives; c) legitimizing and accepting different perspectives; d) actively questioning the patient about his/her experience – asking for detailed descriptions of experience (“what questions”) rather than explanations (“why questions”); e) careful eschewing of the need to understand what makes no sense (i.e., saying explicitly that something is unclear). An important component of this stance is monitoring one’s own mentalizing failures as a therapist. In this context, it is important to be aware that the therapist is constantly at risk of losing his/her capacity to mentalize in the face of a nonmentalizing patient. Consequently, we consider therapists’ occasional enactments as an acceptable concomitant of the therapeutic alliance, something that simply has to be owned up to. As with other instances of breaks in mentalizing, such incidents require that the process is “rewound and the incident explored”. Hence, in this collaborative patient-therapist relationship, the two partners involved have a joint responsibility to understand mental processes underpinning events both within and without therapy.

Mentalizing the transference

We caution about the use of transference interpretation in the treatment of BPD because it assumes a level of mentalizing capacity of the patient that he/she often does not possess. This may have led to the suggestion that we “specifically eschew transference interpretation” 27(27). We do not. In fact we specifically employ transference interpretation, give indicators about when it can be used and carefully define six essential components. But equally we caution practitioners firstly about the commonly stated aim of transference interpretation, namely to provide insight, and secondly about genetic aspects, such as linking current experience to the past, because of their potential iatrogenic effects.

Our first step is the validation of the transference feeling, that is establishing the patient’s perspective. Of course this is not the same as agreeing with the patient, but it must be evident to the patient that the therapist has at least understood his/her point of view. The danger of the genetic approach to the transference is that it might implicitly invalidate the patient’s experience. The second step is exploration. The events which generated the transference feelings must be identified. The behaviours that the thoughts or feelings are tied to need to be made explicit, sometimes in painful detail. The third step is accepting enactment on the part of the therapist. Most experiences of the patient in the transference are likely to be based on reality, even if on a very partial connection to it. Mostly this means that the therapist has been drawn into the transference and acted in some way consistent with the patient’s perception of him/her. It may be easy to attribute this to the patient, but this would be completely unhelpful. On the contrary, the therapist should initially explicitly acknowledge even partial enactments of the transference as inexplicable voluntary actions that he/she accepts agency for, rather than identifying them as a distortion of the patient. Drawing attention to such therapist components may be particularly significant in modeling to the patient that one can accept agency for involuntary acts and that such acts do not invalidate the general attitude which the therapist tries to convey. Only then can distortions be explored. Step four is collaboration in arriving at an interpretation. Transference interpretations must be arrived at in the same spirit of collaboration as any other form of interpretive mentalizing. The metaphor we use in training is that the therapist must imagine sitting side-by-side with the patient, not opposite. They sit side-by-side looking at the patient’s thoughts and feelings, where possible both adopting the inquisitive stance. The fifth step is for the therapist to present an alternative perspective and the final step is to monitor carefully the patient’s reaction as well as one’s own.

We suggest these steps are taken in sequence and we talk about mentalizing the transference to distinguish the process from transference interpretation, which is commonly viewed as a technique to provide insight. Mentalizing the transference is a shorthand term for encouraging patients to think about the relationship they are in at the current moment (the therapist relationship) with the aim to focus their attention on another mind, the mind of a therapist, and to assist them in the task of contrasting their own perception of themselves with how they are perceived by another, by the therapist or indeed by members of a therapeutic group.

Whilst we might point to similarities in patterns of relationships in the therapy and in childhood or currently outside of the therapy, the aim of this is not to provide the patients with an explanation (insight) that they might be able to use to control their behaviour pattern, but far more simply to highlight one other puzzling phenomenon that requires thought and contemplation, part of our general therapeutic stance aimed to facilitate the recovery of mentalization which we see as the overall aim of treatment.

EFFECTIVENESS OF MENTALIZATION BASED TREATMENT

Our initial study of MBT 14 compared its effectiveness in the context of a partial hospital program with routine general psychiatric care for patients with BPD. Treatment took place within a routine clinical service and was implemented by mental health professionals without full psychotherapy training who were offered expert supervision. Results showed that patients in the partial hospital program showed a statistically significant decrease on all measures, in contrast with the control group, which showed limited change or deterioration over the same period. Improvement in depressive symptoms, decrease in suicidal and self-mutilatory acts, reduced inpatient days, and better social and interpersonal function began after 6 months and continued to the end of treatment at 18 months.

The 44 patients who participated in the original study were assessed at 3 month intervals after completion of the trial using the same battery of outcome measures 15. Results demonstrated that patients who had received partial hospital treatment not only maintained their substantial gains, but also showed a statistically significant continued improvement on most measures, in contrast with the control group of patients who showed only limited change during the same period. Because of continued improvement in social and interpersonal function, these findings suggest that longerterm rehabilitative changes were stimulated.

Finally, an attempt was made to assess health care costs associated with partial hospital treatment compared with treatment within general psychiatric services 28. Health care utilization of all patients who participated in the trial was assessed using information from case notes and service providers. Costs were compared 6 months prior to treatment, during 18 months of treatment, and at 18-month followup. No cost differences were found between the groups during pretreatment or treatment. During the treatment period, the costs of partial hospital treatment were offset by less psychiatric inpatient care and reduced emergency department treatment. The trend for costs to decrease in the experimental group during followup was not duplicated in the control group, suggesting that specialist partial hospital treatment for BPD is no more expensive than general psychiatric care and leads to considerable cost savings after the completion of 18-month treatment.

All patients who participated in the partial hospital treatment trial have now been followed up 8 years after initial randomization 29. The primary outcome for this longterm followup study was number of suicide attempts. However, in the light of the limited improvement related to social adjustment in follow-along studies, we were concerned to establish whether the social and interpersonal improvements found at the end of 36 months had been maintained and whether additional gains in the area of vocational achievement had been made in either group. Patients treated in the MBT program remained better than those receiving treatment as usual, but, although maintaining their initial gains at the end of treatment, their general social function remained somewhat impaired. Nevertheless, many more were in employment or full time education than the comparison group, and only 14% still met diagnostic criteria for BPD compared to 87% of the patients in the comparison group who were available for interview.

A further randomized controlled trial of MBT in an outpatient setting (MBT-OP) has recently been completed. One hundred thirty-four patients were randomly allocated to MBT-OP or structured clinical management representing best current practice. Substantial improvements were observed in both conditions across all outcome variables. Patients randomized to MBT-OP showed a steeper decline of both selfreported and clinically significant problems, including suicide attempts and hospitalization 30.

Further research studies are underway, including randomized controlled trials on patients with substance use disorders and patients with eating disorders. A partial replication study of the original partial hospital trial has also been completed by an independent group in the Netherlands, showing that good results are achievable within mental health services away from the instigators of the treatment.

CONCLUSIONS

MBT may not be radically different from other forms of intervention widely practiced by psychotherapists and other mental health professionals in the various contexts in which individuals with BPD are being treated. We claim no originality for the intervention. MBT represents the relatively unadulterated implementation of a combination of developmental processes readily identified in all our histories: a) the establishment of an intense (attachment) relationship based on attempts to engage the patients in a process of understanding their mental states, and b) the coherent representation of their feelings and thoughts, so that patients are able to identify themselves as thinking and feeling in the context of powerful bonds and high levels of emotional arousal. In turn, the recovery of mentalization helps patients regulate their thoughts and feelings, which then makes relationships and selfregulation a realistic possibility.

Although we would claim to have identified a particular method that makes the delivery of this therapeutic process possible, we make no claims of uniqueness. Many situations can likely bring about symptomatic and personality change by this mechanism and hence our permissiveness of technique. The goal of further research is to identify increasingly effective and costeffective methods for generating change in this excessively problematic group.