387

Psychology and Psychotherapy: Theory, Research and Practice (2013), 86, 387–400 © 2012 The British Psychological Society www.wileyonlinelibrary.com

An exploration of group compassion-focused therapy for personality disorder ∗

Katherine M. Lucre1 and Naomi Corten2 1

Specialist Psychotherapies Service, Birmingham and Solihull Mental Health Foundation Trust, Birmingham, UK 2 Leicestershire Partnership Trust, Leicester, UK Background. People with personality disorders, especially those who also experience high self-criticism and shame, are known to be a therapeutic challenge and there is a high dropout rate from therapy. Compassion-focused therapy (CFT) was designed to address shame and self-criticism specifically, and to develop people’s ability to be self-reassuring and more compassionate to themselves and others. Aims. This study explored how CFT affected self-criticism and self-attacking thoughts, feelings, and behaviours, as well as the general symptoms of anxiety, stress, and depression of a personality disordered group within an outpatient group setting, and evaluated the extent of maintenance at a 1-year follow-up. A secondary objective was to identify some of the key characteristics that such an intervention would require. This was a pilot study exploring the feasibility, acceptability, and potential value of CFT in treating this difficult population and, as such, was designed as a pre-randomized controlled trial (RCT) to provide evidence to support applications for funding for an RCT. Methods and design. This study utilized a mixed method combining qualitative and quantitative methods to support a programme evaluation. Eight participants were introduced to the evolutionary-based CFT model and taken through explorations of the nature of self-criticism and shame. In subsequent sessions, participants were taught the main compassion-focused exercises, and any difficulties were addressed. The group was asked to share their personal stories and experiences of practicing self-compassion and to develop compassionate encouragement for each other. Self-report measures were administered at the beginning, end, and at a 1-year follow-up. Results. This 16-week group therapy was associated with significant reductions in shame measured by the Others as Shamer Scale (OAS), social comparison on the Social Comparison Scale (SCS) feelings of hating oneself, and an increase in abilities to be self-reassuring on the Self-Attacking and Self-Reassuring Scale (FSCRS), depression and stress measured by the Depression Anxiety and Stress Scale (DASS). There were significant changes on all CORE variables, well-being, risk, functioning, and problems. Also interesting was that all variables showed a trend for continued improvement at 1year follow-up, albeit statistically non-significant. A content analysis revealed that patients ∗ Correspondence should be addressed to Katherine M. Lucre, Callum Lodge, 24, Lodge Road, Winson Green, Birmingham B18 5SJ, UK (e-mail: [email protected]).

DOI:10.1111/j.2044-8341.2012.02068.x

388

Katherine M. Lucre and Naomi Corten

had found it a moving and very significant process in their efforts to develop emotional regulation and self-understanding. Conclusion. CFT, delivered in a routine psychotherapy department for personality disorders, revealed a beneficial impact on a range of outcome measures. These improvements were maintained and further changes noted at 1-year follow-up. Further research is needed to explore the benefits of CFT using more detailed analysis and RCTs.

Compassion-focused therapy (CFT) was designed for people who have high shame and self-criticism (Gilbert, 2000, 2010a). These are individuals who can achieve an intellectual understanding yet lack an emotional experience of the value of the therapy a phenomenon known as cognitive emotional mismatch (Linehan, 1993; Stott, 2007). The essence of CFT is that in order for us to feel reassured by our thoughts or behaviours we have to be able to access a particular emotion system. This system evolved as a threat calming and soothing systems in the context of attachment and affiliative relationships (Depue & Morrone-Strupinsky, 2005). CFT is based on an evolutionary (Gilbert, 1989; 2000, 2010a, 2010b) and neuroscience model of emotional regulation (Depue & Morrone-Strupinsky, 2005; Panksepp, 1998). It suggests that our social motives, such as desiring, seeking and utilizing attachments and friendships, developing sexual and reproductive relationships, and belonging to groups and seeking status, evolved over millions of years and are regulated by three specific affect regulation systems. First are those that detect and respond to threats (e.g., with defensive emotions such as anxiety and anger and behaviours such fight, flight, avoidance, and submission (LeDoux, 1998). Second those that detect and respond to rewards (e.g., with feelings of pleasure, excitement drive, and motivated behaviours). Third those that detect sufficiency and safeness and give rise to feelings of contentment, soothing, and affiliation. In CFT, this way of understanding affect regulation is called the ‘three circles or three systems model’. CFT suggests that during early development these three systems become patterned and organized in different ways. Research based on the attachment model (Bowlby, 1980; Mikulincer & Shaver, 2007) shows that children who have secure and caring relationships, have received soothing and affiliation in the context of their distress. This lays down internal models of self as capable and lovable, others as caring and distress as manageable and tolerable (Mikulincer & Shaver, 2007). Therefore, key to the regulation of the threat system, and to some extent also the drive system, is the activation of soothing and safeness via interpersonal interactions (Cozolino, 2008). Humans have evolved to be emotionally regulated within relationships and have particular neurophysiological systems, especially those linked to oxtocin, that enable affiliation to regulate threat (Carter, 1998; Depue & Morrone-Strupinsky, 2005). Given that over 100 million years the evolving mammalian brain has become adapted to regulation through relationships, it makes sense to tap into this important internal regulating process for therapeutic purposes (Gilbert, 2000, 2010a, 2010b; Holmes, 2001; Wallin, 2007; Schore, 1994). CFT posits that some mental health difficulties arise because affect regulation systems get out of balance. In particular, the threat system becomes poorly regulated (Gilbert, 1993, 2010a). This poor regulation can arise, be accentuated and maintained in various ways, such as by neurophysiological consequences of difficult rearing environments, poorly processed traumatic memories, rumination or by living in hostile critical environments. However, CFT also suggests one of the most common ways in which the

Compassion-focused therapy for personality disorder

389

threat system can become overly sensitive, accentuated and be maintained in a state of activation is when the internal, self-evaluative relationship is critical and shame prone. This can result in the individual developing a low social rank in relation to others. Gilbert (2009) describes a process of selective self-focused and attention to the power of others coupled with a need to suppress any upward rank expression of anger. Self-criticism is a common automatic response when individuals experience setbacks. A functional magnetic resonance imaging (fMRI) study found that self-criticism, in contrast to self-reassurance, activates quite different (more threat focused) brain systems (Longe et al., 2010). Vulnerability to shame-based self-criticism is commonly rooted in feeling memories of the self-being rejected, criticized, shamed, and abused (Andrews, 1998; Kaufman, 1989; Schore, 1998). Shame and self-criticism are major pathogenic processes for a wide range of psychopathologies (Gilbert & Irons, 2005; Zuroff, Santor, & Mongrain, 2005). It is also recognized that high levels of self-criticism complicate various psychotherapies and that self-critics tend to do less well in controlled trials (Rector, Bagley, Zegal, Joffe, & Levitt, 2000). Clients who meet the criteria for a diagnosis of personality disorder (PD) are especially prone to high levels of shame and self-criticism thought to be associated with toxic early life experiences (Bateman & Fonagy, 2004). Furthermore, it is possible that if shame is not addressed specifically in the therapy, it could correlate with dropout. R¨ usch et al. (2007) found that women with borderline PD (BPD) reported higher levels of guilt and shame proneness than socially phobic and healthy subjects. This shame proneness correlated positively with a tendency towards self-criticism and poorer quality of life. Liotti (2000) formulated a further link between early ruptures in the primary attachment relationship, increased vulnerability to complex trauma and BPD. Allen et al. (2008) similarly identifies abuse and neglect as significant in undermining the capacity for regulating emotions, a common trait in BPD. There appears to be a link between a dearth in nurturance and a tendency to withdraw in adult relationships, therefore increasing the susceptibility to be guided by early schematic representations of the self as ‘bad, unloveable and unworthy’ (Allen, Fonagy, & Bateman, 2008). Intrusive memories of being shamed and abused are often internalized and it becomes safer to blame the self rather than to view primary care givers as flawed (Gilbert & Irons, 2005). Similarly, Gilbert and Irons (2004) found that in a group of students self-reports of experiencing parents as rejecting was significantly associated with self-criticism. There appears to be a general trend within the literature, however, to focus explicitly on BPD, despite the evidence which suggest that service users with a diagnosis of PD are likely meet the criteria for more than one diagnostic category (Bornstein, 1998; Lilienfeld, Waldman, & Israel, 1994). It likely that once a primary diagnosis of PD has been made, limited attention is paid to assessing for further traits (Herkov & Blashfield, 1995). One way of managing difficulties associated with self-criticism is to give clear deshaming explanations of why people can have difficulties with emotional regulation and traumatic memory. Locating these difficulties within an evolutionary and ‘safety strategies model’ can be helpful in addressing the implicit self-blame that is often a key aspect of shame prone clients. The safety strategies model is an adaptation of Cognitive Behavioural ‘safety behaviours’ concept, but which moves away from using terms such as maladaptive or distorted to describe cognitive and behavioural responses to threat. Instead, the emphasis is placed on the notion that given the toxic early life experiences

390

Katherine M. Lucre and Naomi Corten

threat-focused strategies have developed that may have been useful in early life, but are no longer viable, that is, avoidance, overcompensation, or surrender (Gilbert, 2011; Young, Klosko, & Weishaar, 2003). These considerations underpin CFT because the therapeutic process directs attention to the affect system that gives rise to the feelings of emotional safeness and being able to feel self-reassured. CFT helps clients begin to activate and stimulate their soothing system with a series of therapeutic interactions and specific exercises that are compassion focused. These are linked to helping people become more motivated to be caring of themselves, sensitive to and tolerant of their distress and empathic and validating of the difficulties. Clients are taught how to become more aware of self-criticism and then to switch to a more affiliative, supportive, and compassionate position within themselves. Sometimes this can be achieved by imagining a compassionate person or just a compassion voice in their minds that is oriented to be helpful and understanding to them. At other times, it can involve imagining oneself to be a compassionate being and trying to engage with the particular difficulties through one’s compassionate self. There is growing evidence of the value of compassionate focusing on well-being (e.g., Fredrickson, Cohn, Coffey, & Pek, 2008; Hutcherson, Seppala, & Gross, 2008). Also, there is increasing, but as yet limited, research on the effectiveness of group-based CFT for people with chronic mental health difficulties and PDs. In an early study, Gilbert and Procter (2006) found that within a day hospital setting, CFT produced significant changes in self-criticism, shame, depression, and anxiety and improved the capacity of participants to be compassionate to themselves. In a study of group-based CFT for 19 clients in a high security psychiatric setting, Laithwaite et al. (2009) found ’ . . . a large magnitude of change for levels of depression and self-esteem . . . .. A moderate magnitude of change was found for the social comparison scale and general psychopathology, with a small magnitude of change for shame, . . . .. These changes were maintained at 6-week follow-up’ (p. 521).

Methods This study utilized a mixed design combining qualitative and quantitative methods to support a programme evaluation whose purpose was to attempt to assess the worth or value of some innovation, intervention, service, or approach (Robson, 2002, p. 202). In this case, it was the evaluation of a newly developed CFT groupwork programme for people with PD. The content analysis offered an opportunity to explore the emerging themes relating to what participants found useful about the therapy, whilst the quantitative data charted the progress of the group with regard to the specific aims of the therapy – to reduce self-criticism and build capacity for self-soothing. Participants This study involved client volunteers who regarded themselves as having tendencies to be ‘hard on themselves’ or ‘self-critical’, and who had experienced long-term complex trauma consistent with the diagnostic criteria for PD. The study commenced with 10 subjects, one dropped out due to problems getting to the group, one refused to participate in the research, the data therefore refer to eight subjects who took part. All subjects were White British with an age range of 18–54 years. There were seven women and two men.

Compassion-focused therapy for personality disorder

391

All clients were assessed by the senior clinician and considered to have a diagnosis of PD according to ICD 10. The senior clinician is a Cognitive Behavioural Psychotherapist, also trained in administration of the International PD Examination (a diagnostic instrument for PD). The diagnostic criteria for these clients included emotionally unstable, anxious (avoidant), anakastic, paranoid, and histrionic. All subjects had been known to secondary care services for at least 2 years and many had previous experience of therapy including individual Psychodynamic Psychotherapy, Social Problem Solving Therapy, and Day Therapeutic Community. For the duration of the group, however, participants were not receiving any other form of therapeutic input and for those in receipt of secondary care services, contact was generally reduced. Clients had been referred to the Therapy Services for People with PD in Leicestershire Partnership NHS Trust, United Kingdom. The Service provides a wide range of groupbased therapies varying from psychodynamic to cognitive behavioural interventions. It follows therapeutic community principles, which is modelled on the idea that staff and group members all share a significant involvement in the decision making and have a collective responsibility that encourages a sense of belonging, empowerment, safeness, consistency, and personal responsibility (Campling, 2004). Measures All measures were administered by co-author, who also facilitated the group. All measures were a necessary component of the therapeutic process and as such were given prior to the commencement of the group and during the final and follow-up groups. Feedback was offered to group members in the form of individual and group histograms charting the changes. Social Comparison Scale (SCS) This scale was developed by Allan and Gilbert (1995) to measure self-perceptions of social rank and relative social standing. This scale uses a semantic differential methodology and consists of 11 bipolar constructs. Participants make a global comparison of themselves in relation to other people and rate themselves on a 10-point scale. The items cover judgements concerned with rank, attractiveness, and how well the person thinks they ‘fit in’ with others in society. Low scores indicate feelings of inferiority and general low rank self-perceptions. The SCS has good reliability, with Cronbach’s alphas of .88 and .96 with clinical populations and .91 and .90 with student populations (Allan & Gilbert, 1995, 1997). Submissive Behaviour Scale (SBS) The SBS consists of 16 examples of submissive behaviour (e.g. ‘I agree that I am wrong even though I know I’m not’) that people rate as a behavioural frequency (from 0 = never to 4 = Always). The scale has good reliability, with a Cronbach’s alpha of .89, and 4-month test–retest reliability of .84 with a student population (Gilbert, Allan, & Goss, 1996).

The Other as Shamer Scale (OAS) Shame can have an external focus (thinking that others look down on and negatively evaluate the self) and an internal (self-evaluative) focus. The OAS was devised to measure

392

Katherine M. Lucre and Naomi Corten

‘external shame’ (Allan, Gilbert, & Goss, 1994; Goss, Gilbert, & Allan, 1994). Participants are asked to rate the 18 items on a 5-point scale according to the frequency with which they make certain evaluations about how others judge them (0 = never to 4 = almost always). Items include ’I feel other people look down on me’, ’other people see me as somehow defective as a person’, and ’other people always remember my mistakes’. In the original study, the scale showed good reliability with a Cronbach’s alpha of .92 (Goss et al., 1994). Self-Attacking and Self-Reassuring Scale (FSCRS) This was developed by Gilbert and Irons (2004). It is made up of 22 items to measure the form and style of people’s critical and self-reassuring, self-evaluative responses to a disappointment. An example of an item is ‘I think I deserve my self-criticism’. Participants are asked to estimate how like them each statement is on a Likert scale, ranging from 0 (not at all like me) to 4 (extremely like me). Depression Anxiety and Stress Scale (DASS21) This is a shortened version of the DASS42 (Lovibond & Lovibond, 1995). It consists of 21 items; three subscales measure levels of depression (e.g., ’I couldn’t seem to experience any positive feelings at all’), anxiety (e.g., ‘I was aware of the dryness of my mouth’), and stress (e.g., ’I found it hard to wind down’). Participants are asked to rate how much each statement applied to them over the past week on a 4-point scale (0 = did not apply to me at all, 3 = applied to me very much, or most of the time). The DASS21 subscales have good reliability, with Cronbach’s alphas of .94 for Depression, .87 for Anxiety, and .91 for Stress (Antony, Bieling, Cox, Enns, & Swinson, 1998). Clinical Outcomes in Routine Evaluation (CORE) The CORE was developed by the Psychological Research Centre at the University of Leeds (1998) and designed for use in psychotherapy, psychological therapies, and counselling. It is the first standardized public domain approach to audit, evaluation, and outcome measure for psychological therapies including psychotherapy. Participants are asked to respond to 34 questions (such as ‘I have felt criticised by other people’) about how they have been feeling over the last week on a 5-point scale, ranging from ‘not at all’ to ‘most or all of the time’. The CORE measures participants’ levels of distress in comparison with national ‘cut-off’ scores. The CORE is separated into four subscales; risk, problems, well-being, functioning. Procedure All clients who were accepted for the CFT group were advised about the research project by group co-facilitator and invited to participate. All participants provided informed consent. Intervention The senior clinician is an accredited Cognitive Behavioural Psychotherapist and both therapists had attended Professor Paul Gilbert’s 3 day Compassionate Mind Training.

Compassion-focused therapy for personality disorder

393

The co-therapist is a band four Group Facilitator. Both attended Paul Gilbert’s CFT bimonthly supervision group for the duration of the 16-week group. Specific guidance and supervision was also provided on an individual basis by Paul Gilbert as required. There were three main components to the groupwork process, which commenced at assessment: formulation and psychoeducation, compassionate mind training, and planning for practice. Formulation and psychoeducation The formulation process involved documenting the individual’s history and presenting problems within a diagrammatic structure to illustrate how the client’s safety strategies result in unintended consequences that exacerbate the problem. This process also elicits the ‘key fears’ of the client and links back to early life experiences. The emphasis for this aspect of the assessment was also to introduce the idea that the person’s difficulties are not their fault and simply a consequence of our shared brain design and for most their toxic childhood experiences. The psychoeducation component informed the early weeks of the group and was used to underpin the later more exploratory phase of the process, that is, clients difficulties were linked back to the three circles, formulation, and brain design. The functions of self-criticism were explored during this early phase of treatment (e.g., to keep me on my toes, make sure I know when I have done wrong), and fears associated with giving it up (I might become lazy or arrogant). CFT does not encourage clients to spend a lot of time engaging with or challenging self-criticism directly. Rather, the focus is on developing the compassionate attention, thinking, feeling, and behaviour that is linked to the development of soothing affiliative system. CFT suggests that activation of this system is a naturally evolved regulation of the threat system, and as standard behavioural therapy suggests, it is difficult to feel critical and compassionate for oneself at the same time. Hence, the group focuses explicitly on developing soothing rather than understanding more about the critical side. Compassionate mind training Specific exercises were taught to the group to develop the capacity for self-soothing. The purpose of these exercises were to encourage the development of a different and more compassionate relationship with the self that could pervade all aspects of the individual’s life, for example, relationships with others, responses to distress. Each group finished with a compassion-focused imagery exercise where members spent 5–10 min feeling compassion for each person in the group. This was done as a mindfulness meditation task incorporating the development of a compassionate image that formed the basis of ’between group’ practice. This exercise was supported by the use of group-specific CDs that were given to the group towards the end of the programme to facilitate integration and internalization of the therapeutic work. Planning for practice Planning for practice was also incorporated into the group structure with time spent anticipating the blocks to practice. Semi-precious stones were given to the group to support the mindfulness component of the practice ‘something to focus on’ and also a transitional object to connect to the groupwork process. The between session

394

Katherine M. Lucre and Naomi Corten

Table 1. Mean ranks and p values (Friedman’s ANOVA)

Submissive Behaviour Other as Shamer Social comparison FSCRS: Inadequate self Hated self Reassured self DASS: Depression Anxiety Stress CORE: Well-being Symptoms Functioning Risk

N

Pre (A) Mean rank

Post (B) Mean rank

Follow-up (C) Mean rank

Friedman’s Significance (p)

Wilcoxon post hoc

8 8 8

2.44 2.75 1.06

1.81 2.00 2.19

1.75 1.25 2.75

.303 .011∗ .002∗∗

A=B=C A⬎B⬎C A⬍B=C

8 8 8

2.63 3.00 1.00

1.69 1.63 2.50

1.69 1.38 2.50

.062 .001∗∗ .002∗∗

A⬎B=C A⬎B=C A⬍B=C

8 8 8

2.75 2.50 2.63

1.88 2.06 2.06

1.38 1.44 1.31

.012∗ .081 .025∗

A⬎B=C A=B=C A⬎C

7 7 7 7

2.86 2.93 3.00 2.39

1.86 1.57 1.79 2.29

1.29 1.50 1.21 1.36

.008∗∗ .006∗∗ .002∗∗ .070†

A⬎B=C A⬎B=C A⬎B=C A=B=C

Note. † Non-significant in Friedman’s ANOVA; Wilcoxon post hoc test revealed a significant change between pre and follow-up. ∗ p ⬍ .05; ∗∗ p ⬍ .01.

practice related to building capacity through imagery, mindful attention to breathing, and engagement in experimental acts of self-compassion. The groupwork was also supported by handouts of the diagrammatic/pictorial representations of the group process and exploration. This resulted in each group member having a ‘compassion toolkit’ that could be utilized following the end of the programme.

Results Data for all the measures at pre, post, and 1-year follow-up evaluation are given in Table 1. The analysis used a Friedman ANOVA with SPSS 14 (IBM) due to the small size of the study. With regard to the social rank variables, feelings of shame significantly reduced during the therapeutic sessions and actually further improved at follow-up. Social comparison also improved during therapeutic sessions and improvement was maintained with a nonsignificant further improvement at follow-up. There was a small improvement in the submissive behaviour measure but it was non-significant. In regard to self-criticism, there was a highly significant reduction in self-hatred but inadequate self-criticism just failed to reach significance. There was also a highly significant increase in self-reassurance that taps into ways of being more positive and compassionate to oneself. With regard to the symptom measures, there were significant changes in depression and stress (at 1 year), but although the level of reported anxiety decreased, it did not reach significance. Again the data show a non-significant trend for further improvements post treatment over the year.

Compassion-focused therapy for personality disorder

395

Significant change occurred on all CORE measures coupled with a non-significant trend for further improvement over the year. This indicated a reduction in symptoms of emotional distress that many participants scored highly on at the beginning of therapy. There was also a significant increase in member’s perception of their general social functioning and experience of well-being. The reduction in the level of risk that group members perceived that they posed to themselves and/or others significantly reduced, but not to the same extent as the other measures. At 1-year follow-up, mean CORE scores across all dimensions, with the exception of risk had reduced to within a subclinical range. Qualitative data Clients were asked to write their reflections of the therapy both at the end of the therapy group and also when they were contacted with an invitation to the 1-year follow-up session. This was undertaken to try to identify their own personal experiences but also to find ways of improving the service. The content analysis was undertaken by the senior clinician, who reviewed the letters received by clients in addition to the documented feedback from the groups and the follow-up session. An analysis of the material revealed a number of emerging themes, relating to the experience of the group, and key learning points (Robson, 2002). Taking responsibility for one’s thoughts and reassurance I have learned that no-one else can do this for me and that sucks sometimes, but I have to reassure myself because I wouldn’t believe someone else anyway. My angry thoughts and words that I believed protected me from attack from others, have kept me trapped and now I don’t need them anymore . . . I have realized now that the only one who is truly in charge of my recovery is me.. I used to tell myself to snap out of it and it didn’t work, now I have a soothing voice which tells me that it is OK.

The comfort of shared group experiences Many groups members echoed the sentiment I am not alone: knowing that others struggle as they do seems to have been a key component of the group. The use of semiprecious stones as a transitional object supported the imagery in the early stages of the therapy: I still have my stone and use it often to ground my thoughts, sometimes I use it just to remind me that there are others out there who understand.

Fear of compassion Fear of compassion featured throughout the group, with many associating warmth and kindness to self with inactivity and self-indulgent/self-destructive behaviour. I used to be scared that I would be stuck on a compassionate sofa, going nowhere . . . I never realized that being kind to myself could help me DO things.. I always thought this compassion stuff would make me weak.. pathetic.. but now I feel that it has made me stronger.

396

Katherine M. Lucre and Naomi Corten

Many of the group observed at the follow-up that using self-compassion as a precursor to increased activity had undermined and eroded this key fear. In particular, the shared ideas around gardening as a medium for compassionate activity were maintained 1-year on. Awareness of self-criticism and addressing it with assertive action Participants demonstrated an increased ability to identify when they were engaging their internal critic. Even more importantly, they reported new found strategies for dealing with this self-criticism rather than blaming themselves for its existence. I catch myself out daily using the ‘I should have / could have/ ought to have..’ type phrases and swiftly remind myself to ‘be compassionate!’ and often catch myself telling others to do the same thing. The voice in head telling me ‘I’m useless, wrong unwanted’(I used to tell it to go away or who do you think you are?) But now I have stopped listening to it, I don’t hear it at all. I really feel like I have gained some control back for myself, not through medication or negative actions but real self control over my feelings. I have learned things that I am sure I will keep with me for the rest of my life. Of course I’m not saying that I am now healed ..but I can feel positive about things and believe that I can improve. I left each group feeling a little better in myself, a little scared but feeling for once I had somewhere to go and what ‘moving on’ really meant.

One-year follow-up Six of the eight group members attended the follow-up group with apologies and letters received from the other two members. This group was offered as a further therapy session, agreed at the outset of the programme, to explore the extent of integration of the CFT model, address any difficulties with ongoing practice, and plan for further development of capacity for self-soothing. It is of note that, with one exception, none of the group members had received any form of psychiatric input since the end of the programme. One group member had completed a degree achieving first class honours, another had started full-time vocational employment, one had passed their driving test, and one had got two kittens to replace her beloved pet. Many had become keen gardeners, having discovered through the group that gardening could assist in accessing the soothing and drive systems interchangeably. Other group members had managed difficult situations in their personal lives and relationships with a degree of compassion and kindness for themselves that they recognized had altered the outcomes significantly.

Discussion This study explored the acceptability and value of CFT in a standard PD service in the United Kingdom. The results showed a generally significant improvement on all measures with a continuation of non-significant improvement at 1-year follow-up. It is of note when considering the follow-up data that seven of the eight group members had been discharged from mental health services. It is interesting to note that although feelings of inadequacy did reduce it was nonsignificant. This mirrors Gilbert and Procter (2006) and indicates that these clients have

Compassion-focused therapy for personality disorder

397

a belief that by being critical they can ‘drive themselves to improve’. However, there was a significant reduction in the more pathogenic ‘hated self’ measure and a significant improvement in the ‘reassured self’ measure. This seems to indicate that the shift in the level of self-loathing and hatred correlates with the increased capacity to soothe and reassure the self. It may be that a sense of self as adequate may require more behavioural than emotional evidence. Future therapy may need to be more focused on this element of self-criticism and the need for behavioural activation to address this particular measure. It is also of note that a significant change in submissive behaviour was not found as in Gilbert and Procter (2006). This is an interesting finding in light of the qualitative data that clearly indicates that the group had made connections between self-compassion and assertive action, ’real control over my feelings’. The authors suggest that perhaps this group, whilst recognizing the behavioural changes they had made, continued to struggle with asserting themselves with others, or in recognizing this. This tendency towards subjugation is likely to link with the established patterns of behaviour often associated with early experience of neglect, criticism, and abuse (Young et al., 2003). The significant reduction in the experience of external shame is encouraging as the group process and content focused on addressing this issue specifically, through increasing awareness of how much our negative self-perception ‘colours’ how we imagine that others view us. Especially important for them was the focus on shaming and using an evolutionary model to help them realize that much of what goes on in our minds is not our fault but is related to the way that our brains are built and how we have been socially shaped. It is the authors’ view that this de-shaming process sets the foundation for compassion because these individuals have a deep sense of being unacceptable or simply bad in some way. Once patients have a deep understanding of the evolutionary model, this seems to help them let go of a lot of self-blaming. It is possible that this focus on addressing the internal experience of shame could have an ameliorating impact on external shame. It is, of course, therapeutically problematic to encourage the group to see others as kind, compassionate, and supportive, given the propensity for this group to recreate damaging relationships. The group focused instead on improving the relationship with the self whilst accepting that they may not get what they need from others. It is possible that the same process facilitated the improvement in the way group members viewed themselves in relation to others, evidenced by the significant improvement on the SCS and significant reductions in the experience of others as shaming.

Limitations Although the data show some encouraging results, it is acknowledged that the sample size was very small and only related to one group intervention. Also, it is difficult to identify the specific aspects of the programme that may have accounted for the significant changes in the self-report measures. Although some cautious suggestions are made by the authors, without more thorough analysis of the programme, these suggestions remain unproven. Although seven of the eight group members had been discharged from the mental health services at the point of follow-up, one group member had continued with therapeutic input that could have influenced the scoring of the follow-up questionnaires, thereby introducing some bias.

398

Katherine M. Lucre and Naomi Corten

As this study took place in the context of routine service evaluation, limitations in the gathering of qualitative data about the group have emerged. Information about preand post-service usage, use of medication, and other relevant issues that could have strengthened the claims made about recovery and improvement were not gathered.

Some closing thoughts Taken as a whole, this therapy looks to be highly beneficial for this group of patients and the non-significant trends for further improvements over the year are of particular interest as they indicate that patients maintain their improvements well with no evidence of drop-off. It is our belief that in a larger group the continuing change and improvement would have been significant. This is important because CFT seeks to teach people skills that, when practiced, are designed to have continued benefits. The authors found that the capacity to ‘practice what you preach’ was fundamental to this process and therefore therapist self-compassion practice should be integrated into future training and protocols for this intervention. The significance of the behavioural component of this group is supported by the maintenance and continuation of change at 1-year follow-up. This, of course, indicates the cost-effectiveness of a comparatively short-term intervention for clients with high levels of psychological disturbance. Future studies will be required to gather more extensive data with the inclusion of randomized controlled trial exploring the benefits of CFT compared with other time-limited group therapies for people with PD. However, it is the view of the authors that this is an encouraging starting point for exploring structured, time limited, cognitive behavioural informed groupwork interventions for clients with complex trauma and PD.

References Allan, S., & Gilbert, P. (1995). A Social Comparisons Scale: Psychometric properties and relationship to psychopathology. Personality and Individual Difference, 19, 293–299. Allan, S., & Gilbert, P. (1997). Submissive behaviour and psychopathology. British Journal of Clinical Psychology, 36, 467–488. Allan, S., Gilbert, P., & Goss, K. (1994). An exploration of shame measures: II: Psychopathology. Personality and Individual Differences, 17, 719–722. Allen, J. H., Fonagy, P., & Bateman, A. W. (2008). Mentalizing in clinical practice. Washington, DC: American Psychiatric Association. Andrews, B. (1998). Shame and childhood abuse. In P. Gilbert & B. Andrews (Eds.), Shame: Interpersonal behavior, psychopathology and culture (pp. 176–90). New York: Oxford University Press. Antony, M. M., Bieling, P. J., Cox, B. J., Enns, M. W., & Swinson, R. P. (1998). Psychometric properties of the 42-item and 21-item versions of the depression anxiety stress scales in clinical groups and a community sample. Psychological Assessment, 10, 176–181. Bateman, A., & Fonagy, P. (2004). Psychotherapy for borderline personality disorder: Mentalization-based treatment. New York: Oxford Press. Bornstein, R. F. (1998). Reconceptualizing personality disorder diagnosis in the DSM-V: The discriminant validity challenge. Clinical Psychology: Science and Practice, 5, 333–343. Bowlby, J. (1980). Loss: Sadness and depression. Attachment and loss, Vol. 3. London: Hogarth Press. Campling, P. (2004). Therapeutic communities. Advances in Psychiatric Treatment, 7, 365–372.

Compassion-focused therapy for personality disorder

399

Carter, C. S. (1998). Neuroendocrine perspectives on social attachment and love. Psychoneuroendorinlogy, 23, 779–818. Cozolino, L. (2008). The healthy aging brain: Sustaining attachment, attaining wisdom. New York: Norton. Depue, R. A., & Morrone-Strupinsky, J. V. (2005). A neurobehavioral model of affiliative bonding. Behavioral and Brain Sciences, 28, 313–395. Fredrickson, B. L., Cohn, M. A., Coffey, K. A., Pek, J., & Finkel, S. A. (2008). Open hearts build lives: Positive emotions, induced through loving-kindness mediation, build consequential personal resources. Journal of Personality and Social Psychology, 95, 1045–1062. Gilbert, P. (1989). Human nature and suffering. Hove: Lawrence Erlbaum Associates. Gilbert, P. (1993). Defence and safety: Their function in social behaviour and psychopathology. British Journal of Clinical Psychology, 32, 131–153. Gilbert, P. (2000). Social mentalities: Internal ‘social’ conflicts and the role of inner warmth and compassion in cognitive therapy. In P. Gilbert & K. G. Bailey (Eds.), Genes on the couch: Explorations in evolutionary psychotherapy (pp. 118–150). Hove: Brenner-Routledge. Gilbert, P. (2005). Compassion and cruelty: A biopsychosocial approach. In P. Gilbert (ed). Compassion: conceptualisations, research and use in psychotherapy (pp. 9–74). London: Routledge. Gilbert, P. (2009). The compassionate mind. London: Constable-Robinson. Oaklands, CA: New Harbinger. Gilbert, P. (2010a). Compassion focused therapy: Distinctive features. London: Routledge. Gilbert P. (2010b). Compassion focused therapy: Special issue. International Journal of Cognitive Therapy, 3, 95–201. Gilbert, P. (2011). Compassion focused therapy in groups. Unpublished manuscript. Gilbert, P., Allan, S., & Goss, K. (1996). Parental representations, shame interpersonal problems and vulnerability to psychopathology. Clinical Psychology and Psychotherapy, 3, 23–34. Gilbert, P., & Irons, C. (2004). A pilot exploration of the use of compassionate imagery in a group of self critical people. Memory, 12(4), 507–516. Gilbert, P., McEwan, K., Matos, M., & Rivis, A. (2011). Fear of compassion: Development of a self-report measure. Psychology and Psychotherapy, 83(3), 239–255. Gilbert, P., & Procter, S. (2006). Compassionate mind training for people with high shame and selfcriticism: A pilot study of a group therapy approach. Clinical Psychology and Psychotherapy, 13, 353–379. Goss, K., Gilbert, P., & Allan, S. (1994). An exploration of shame measures: I: The ‘other as shamer’ scale. Personality and Individual Differences, 17, 713–717. Herkov, M. J., & Blashfield, R. (1995). Clinician diagnoses of personality disorder: Evidence of a hierarchical structure. Journal Personality Assessment, 65, 313–321. Holmes, J. (2001). The search for a secure base: Attachment and psychotherapy. Sussex: Routledge. Hutcherson, C. A., Seppala, E. M., & Gross, J. J. (2008). Loving-kindness meditation increases social connectedness. Emotion, 8, 720–724. Irons, C., Gilbert, P., Baldwin, M., Braccus, J., & Palmer, M. (2006). Parental recall, attachment, relating and self attacking/self reassurance. British Journal of Clinical Psychology, 45, 297– 308. Kaufman, G. (1989). The psychology of shame. New York: Springer. Laithwaite, H., Gumley, A., O’Hanlon, M., Collins, P., Doyle, P., Abraham, L., & Porter, S. (2009). Recovery after psychosis (RAP): A compassion focused programme for individuals residing in high security settings. Behavioural and Cognitive Psychotherapy, 37, 511–526. LeDoux, J. (1998). The emotional brain. London: Weidenfeld and Nicolson. Lilienfeld, S. O., Waldman, I. D., & Israel, A. C. (1994). A critical examination of the use of the term ‘comorbidity’ in psychopathology research. Clinical Psychology: Science and Practice, 1, 71–83.

400

Katherine M. Lucre and Naomi Corten

Linehan, M. M. (1993). Skills training manual for treating borderline personality disorder. New York: Guilford Press. Liotti, G. (2000). Disorganised attachment, models of borderline states and evolutionary psychotherapy. In P. Gilbert & B. Bailey (Eds.), Genes on the couch: Explorations in evolutionary psychotherapy (pp. 232–256). Hove: Brunner-Routledge. Longe, O., Maratos, F. A., Gilbert, P., Evans, G., Volker, F., Rockliffe, H., & Rippon G. (2010). Having a word with yourself: Neural correlates of self-criticism and self-reassurance. NeuroImage, 49, 1849–1856. Lovibond, S. H., & Lovibond, P. F. (1995). Manual for the depression anxiety stress scales (2nd ed.). Sydney: Psychology Foundation. Mikulincer, M., & Shaver, P.R. (2007). Attachment in adulthood: Structure, dynamics, and change. New York: Guilford. Panksepp, J. (1998). Affective neuroscience. New York: Oxford University Press. Rector, N. A., Bagley, R. M., Zegal, S. V., Joffe, R. T., & Levitt, A. (2000). Self criticism and dependency in depressed patients treated with cognitive therapy or pharmacology. Cognitive Therapy Research, 24, 571–584. Robson, C. (2002). Real world research (2nd ed.). Oxford: Blackwell Publishing. R¨ usch, N., Weber, M., Il’yasov, K., Lieb, K., Ebert, D., Hennig, J., & Tebartz van Elst, I. (2007). Inferior frontal white matter microstructure and patterns of psychopathology in women with borderline personality disorder and comorbid attention deficit hyperactivity disorder. Neuroimage, 35, 738–747. Schore, A. N. (1994). Affect regulation and the origin of the self: The neurobiology of emotional development. Mahwah, NJ: Erlbaum. Schore, A. N. (1998). Early shame experience and infant brain development. In P. Gilbert & B. Andrews (Eds.), Shame: Interpersonal behavior, psychopathology and culture (pp. 176–90). New York: Oxford University Press. Stott, R. (2007). When the head and heart do not agree: A theoretical and clinical analysis of rationalemotional dissociation (RED) in cognitive therapy. Journal of Cognitive Psychotherapy: An International Quarterly, 21, 37–50. van der Kolk, B., Alexander, C. M., & Lars W. (Eds.). (1996). Traumatic stress. The effects of overwhelming experience on mind, body and society. London: Guildford Press. Wallin, D (2007). Attachment in psychotherapy. New York: Guilford. Young, J., Klosko, J. S., & Weishaar, M. E. (2003). Schema therapy. New York: Guildford Publications. Zuroff, D. C., Santor, D., & Mongrain, M. (2005). Dependency, self-criticism, and maladjustment. In J. S. Auerbach, K. N. Levy, & C. E. Schaffer (Eds.), Relatedness, self-definition and mental representation. essays in honour of Sidney J. Blatt (pp. 75–90). London: Routledge. Received 1 December 2010; revised version received 21 February 2012