Table 1.
Summary description of included studies
| Study (country) | n | Design | Population | Intervention | Control | Number of sessions | Measure | Main outcomes |
|---|---|---|---|---|---|---|---|---|
| Braehler et al. 2013 (UK) | 40 | RCT | Schizophrenia-spectrum disorder with psychotic features | Group CFT for psychosis + TAU. Focused on reduction of shame, stigma, self-blame, development of compassion. Mindfulness, appreciation, compassion imagery and attention and reflection through writing | TAU. Free to vary post-randomization but could include: psychotropic medication, occupational therapy, day centre support or psychological therapies | 16 weekly 2-h sessions | NRSS, BDI, PANAS, FORSE, PBIQ-R, CGI-I | CFT showed significantly more clinical global improvement and more compassion in narratives than TAU. Increase in compassion in CFT group associated with decrease in BDI depression, PBIQ shame, entrapment and social marginalization, FORSE intrusiveness and fear of relapse |
| Kelly et al. 2010 (Canada) | 119 | RCT | Smokers | Self-compassion intervention based on CMT. Imagery-based self-talk exercises, designed to stimulate soothing-affiliation system | 1 × baseline control involving daily self-monitoring exercises. 2 × experimental control (self-energizing and self-controlling imagery exercises) | Daily exercises for 3 weeks. Two laboratory sessions 3 weeks apart, baseline control 20 min + experimental imagery 25 min | Cigarettes smoked per day, SSC-SF, Trait Self-criticism Scale of the DEQ, Imagery Vividness Rating Scale, SeCS | No differences in rate of smoking reduction between three experimental imagery groups but all reduced smoking more than baseline control. Effects moderated by trait self-criticism, readiness to change and vividness of imagery |
| Shapira & Mongrain, 2010 (Canada) | 1002 | RCT | Non-clinical sample recruited through the internet | Self-compassion intervention. Letter-writing exercises about distressing event and providing compassion to themselves | 1 × experimental control: optimistic thinking intervention, letter-writing visualizing future with resolved issues. 1 × control condition wrote freely about early memory | Daily exercises over 7-day period | DEQ, CES-D, SHI | Both experimental interventions resulted in significant increases in happiness at 6 months and significant decreases in depression at 3 months. Effects moderated by self-criticism and levels of dependence |
| Beaumont et al. 2012 (UK) | 32 | Non-RCT | Requiring treatment after trauma | CMT. Loving, caring, accepting imagery, compassionate letter-writing, grounding work + CBT (as control) | CBT techniques including cognitive restructuring, behavioural activation, graded exposure, relapse prevention and Socratic dialogue | Twelve weekly sessions (duration n.r.) | HADS Anxiety and Depression, IES Avoidance Hyperarousal Intrusion, SeCS Compassion | Significant reduction in depression, avoidance in both conditions. Improvement in scores greater for combined CMT + CBT. No difference in anxiety levels between groups. Greater improvement in self-compassion for CMT + CBT than CBT |
| Kelly et al. 2009 (Canada) | 75 | Non-RCT | Distressed chronic acne sufferers | CFT. Self-soothing condition. Slideshow about compassionate self-talk, visualization of compassionate image, compassionate letter-writing. Focus on warmth, acceptance, reassurance, desire to soothe distress | Attack-resisting intervention. Visualize confident, resistant, resilient image. Focus on strength, logic perseverance, self-confidence. Control condition: no exercises | Two weekly sessions. One × 1 h, one × n.r. + three × per day exercises for 2 weeks | DEQ, BDI, ESS, SKINDEX-16 | Depression reduced across conditions. Only attack-resisting condition lowered depression more than control. Moderating effect of self-criticism. Both experimental conditions showed greater reduction in shame than control |
| Gilbert & Irons, 2004 (UK) | 8 | Observational | Individuals from a self-help depression group who regard themselves as self-critical | Compassionate image use. Diaries recording critical thinking | None | Four 1-h sessions over 7 weeks (three consecutive weekly meetings + 4-week follow-up) | HADS, diary and quantitative ratings for self-criticism and self-soothing | No significant change in self-criticism. Significant increase in self-soothing/compassion, ease of generating images in a self-critical situation |
| Gilbert & Procter, 2006 (UK) | 6 | Observational | Patients currently in treatment for major/severe long-term complex mental health difficulties | CMT. Group therapy exploring self-criticism, compassion, self-attacking | None | Twelve weekly 2-h sessions | HADS, FSCS, FSCRS, Social rank variables, OAS, Social Comparison Scale, Submissive Behaviour Scale, weekly diary self-attacking/self-soothing | Significant reduction in HADS anxiety and depression, diary self-criticism, shame, inferiority and submissive behaviour.. Significant increases in self-compassion, reassure-self |
| Laithwaite et al. 2009 (UK) | 18 | Observational | Individuals with primary diagnosis of schizophrenia or bipolar affective disorder | Three modules of CFT-based programme. Understanding psychosis and recovery, understanding compassion and developing the ideal friend, developing plans for recovery after psychosis | None | Twenty sessions over 10 weeks | BDI-II, SeCS, SCS, RSE, SIP-AD, OAS | Significant improvement on BDI, SCS, OAS, RSE. No change on Self-Compassion Scale or SIP-AD |
| Lucre & Corten, 2013 | 10 | Observational | Individuals with chronic personality disorder and who regard themselves as self-critical | CFT groupwork programme. Exercises to develop capacity for self-soothing. Compassion-focused imagery exercises | None | Sixteen weeks with 1-year follow-up | SCS, SBS, OAS, FSCRS, DASS21, CORE | Significant decrease in depression and stress. Reduction in feelings of shame and social comparison. Reduction in self-hatred. Significant increase in self-reassurance |
| McEwan & Gilbert, unpublished data (UK) | 45 | Observational | Non-clinical volunteers | Four audio recordings including, soothing rhythm breathing, compassionate self- imagery, compassionate other imagery. Participants chose which exercise to practise | None | At least 5 min a day over 2 weeks. Amount of time varied by individual from several times a day to every other day | Social Safeness and Pleasure Scale, SeCS, Fears of Compassion Scale, Types of Positive Affect Scale, Experiences in Close Relationships, FSCSR, DASS | Significant increases in self-compassion, social safeness, active and relaxed positive affect and self-reassurance. Significant reductions in self-coldness, fear of comparison for others, avoidance of close relationships, inadequate self-criticism, depression, anxiety and stress |
| Judge et al. 2012 (UK) | 42 | Observational | Range of diagnoses including depression, anxiety, OCD, bipolar, personality disorders, social anxiety, deliberate self-harm. Those referred on to CFT were high in self-criticism and shame | Group CFT | None | Between 12 and 14 weekly sessions of 2 h with a 15-min break | Depression, anxiety, inadequate self, hated self, reassured self, self-correction, self-persecution, internal shame, external shame, social comparison, submissive behaviour, self-critical thought, self-soothing thoughts (BDI, BAI, FSCRS, FSCS, ISS, OAS, SCS, SBS, weekly diary for self-attacking/self-soothing) | Significant improvement for all measures apart from self-correction |
| Gale et al. 2014 (UK) | 139 | Observational | Individuals diagnosed and being treated for an eating disorder | Group-based CBT, integrating CFT. Two-step treatment programme consisting of didactic teaching with in-session written activities and homework, designed to increase understanding of their eating disorder and be actively involved in deciding if they are ready to engage in treatment | None | Step 1: patients are offered a 4-week 2-h per week group-based psycho-education programme, followed by a 20-session group-based recovery programme, taking place over 16 weeks, with two sessions a week for the first 4 weeks and then 12 weekly sessions. Groups last 2–2.5 h, with 2 h of homework each week | EDE-Q; SEDS; CORE-OM | EDE-Q: significant improvement in all EDE-Q subscales after intervention. People with bulimia nervosa improve most between times 1 and 5. SEDS: significant improvement in all subscales except low assertiveness. People with bulimia nervosa improved more than the other two groups between times 1 and 5. CORE-OM: significant improvement for all subscales after intervention. People with bulimia nervosa improved the most between time 1 and 5 |
| Mayhew & Gilbert, 2008 (UK) | 3 | Case series | Individuals diagnosed with schizophrenia experiencing auditory hallucinations | CMT on a one-to-one basis. Focus on understanding and compassionate to safety behaviours. Discussion about self-compassion, tasks that can be used to develop self-compassion | None | Twelve weekly 1-h sessions | BAVQ, FSCS, FSCRS, SCL-90, Voice Rank Scale, SeCS, weekly diary | All participants showed decrease in SCL-90, inadequate-self scores. All showed improved BAVQ (total scores reduced, all voices became less malevolent and less persecuting). Two of three heard more reassuring voices |
| Ashworth et al. 2011 (UK) | 1 | Case study | A 23-year-old female with acquired brain injury after road traffic accident | CFT and CBT sessions targeting low self-esteem, self-criticism, CMT to self-soothe using and brainstorming self-nurturer imagery | None | Twenty-four weekly 1-h and 50-min sessions | Robson SCQ, BAI, BDI, State Trait Anger Expression Inventory, eating disorder symptoms | Reliable decreases in BAI, BDI and SCQ. Anger Expression Inwards Scale fell within normal range. Beliefs relating to key cognitions indicated improvement |
CFT, Compassion-focused therapy; CMT, compassionate mind training; CBT, cognitive behavioural therapy; ISS, Injury Severity Scale; SCS, Social Comparison Scale; FSCS, Functions of the Self-Criticizing/Attacking Scale; FSCRS, Forms of the Self-Criticizing/Attacking and Self-Reassuring Scale; DASS, Depression Anxiety and Stress Scale; SBS, Submissive Behaviour Scale; OAS, Other As Shamer Scale; EDE-Q, Eating Disorder Examination Questionnaire; SEDS, Stirling Eating Disorders Questionnaire; CORE-OM, Clinical Outcomes in Routine Evaluation – Outcome Measure; SIP-AD, Self-Image Profile for Adults; RSE, Rosenburg Self-Esteem measure; SeCS, Self-Compassion Scale; HADS, Hospital Anxiety and Depression Scale; DEQ, Depressive Experiences Questionnaire; BDI, Beck Depression Inventory; BAI, Beck Anxiety Inventory; SCQ, Self-Concept Questionnaire; ESS, Experiences of Shame Scale; IES, Impact of Events Scale; CES-D, Centre for Epidemiological Studies Depression Scale; SHI, Steen Happiness Index; SSC-SF, Smoking Stage of Change – Short Form; PANAS, Positive and Negative Affect Scale; FORSE, Fear of Recurrence Scale; PBIQ-R, Personal Beliefs about Illness Questionnaire – Revised; CGI-I, Clinical Global Impression – Improvement Scale; NRSS, Narrative Recovery Style Scale; TAU, treatment as usual; BAVQ, Belief About Voices Questionnaire; n.r., not recorded; RCT, randomized controlled trial; SCL-90, Symptom Checklist-90; OCD, obsessive–compulsive disorder.