SEARCH WITHIN CONTENT
Citation Information : Scandinavian Journal of Child and Adolescent Psychiatry and Psychology. Volume 8, Pages 38-47, DOI: https://doi.org/10.21307/sjcapp-2020-005
License : (CC-BY-NC-ND 4.0)
Published Online: 21-March-2020
The long-term negative consequences of mental health problems during adolescence highlight the need for effective treatments. Compassion-focused therapy (CFT) aims to help individuals to enhance their ability to support and care for themselves and to alleviate shame and self-stigmatization.
This non-randomized controlled trial examined the effectiveness of group-based CFT on perceived stress and the extent of self-compassion in a clinical sample of adolescents receiving psychiatric care for complex mental health difficulties and their parents.
The participants were 43 adolescents (ages 14–17; 83.7 % female) under treatment at a child and adolescent psychiatric outpatient clinic in Sweden and their parents (
The fathers scored highest on self-compassion and had less perceived stress at both times than the adolescents or mothers. There were no significant differences between the CFT and TAU groups in self-compassion or perceived stress at either time, and the effect sizes were small (
Some patients seem to benefit from group-based CFT in ways similar to the benefits of specialized therapeutic approaches. Further research into the utility of CFT for adolescents with MH problems and their parents, as well as the long-term clinical effects of CFT for this group is needed.
Having one or multiple diagnosable mental health or psychiatric conditions (hereinafter referred to as MH problems), such as externalizing disorders (e.g., attention-deficit hyperactivity disorder and conduct disorder) or internalizing disorders (e.g., depression and anxiety disorders) during adolescence is difficult for the adolescents and their families. The worldwide prevalence of these MH problems is about 10–20 % (1), and it continues to increase (2). Additionally, a substantial proportion of these problems persist into adulthood (3).
In addition to the social and educational problems they encounter, children experiencing MH problems are at higher risks of being stigmatized, such as being stereotyped and discriminated against, than their peers with other (or no) health problems (4). Feelings of shame and self-stigmatization (internalized stigma adopted from others) might interfere with their recovery from MH problems (5) and influence the extent of their MH impairments and chronicity (6, 7).
Furthermore, many adolescents diagnosed with MH problems do not disclose their conditions to others (6, 7), which might hinder them from obtaining social support from their families, peers, and others. Adolescents who believe their MH problems are lifelong conditions tend to have relatively high levels of self-stigma and perceive that they have relatively less control over their difficulties. High levels of shame and self-criticism have been found in samples of adults experiencing MH problems (5, 8-13). However, there is little research on self-criticism among adolescents experiencing MH problems.
Adolescents with one or more MH problems need more support than their healthy peers, and to support them and deal with their problems might be challenging and stressful for their parents (14). Parents of children who experience MH problems tend to perceive themselves as not good enough (15) or believe that they contributed to the causation, exacerbation, or continuation of their children’s MH problems (16). Parents tend to feel isolated in their parenting struggles and unsupported by the other parent, the child’s school, and/or healthcare professionals (15). Interventions are needed to help parents manage their stress and enhance their sense of agency and trust in their abilities to support their children’s development and success (17).
The impact and long-term negative consequences of MH problems during adolescence highlight the need for effective early treatments. Compassion-focused therapy (CFT) was originally developed to help people experiencing chronic and complex MH problems and those experiencing high levels of shame and self-criticism to cultivate a compassionate care orientation toward themselves and others (18, 19). CFT is an integrative multi-modal therapy based on evolutionary psychology, attachment theory, neuroscience, and cognitive behavior theory. An essential aspect of CFT is compassionate mind training to help individuals to experience warm, positive feelings and support within themselves.
Self-compassion refers to an open and caring disposition toward oneself while experiencing an unpleasant event rather than a self-judging and self-critical attitude (20). Self-compassion includes an ability to balance one’s emotions during difficult times and an acknowledgment that struggling is part of the human experience. Adolescents who score high on self-compassion tend toward lower perceived stress (21), anxiety, and depression (22) in non-clinical studies. The relationships between self-compassion, perceived stress, and psychological wellbeing also have been found among adults (23). Parents with relatively high levels of self-compassion have been found to be less critical and to use fewer distressed reactions to cope with their children’s emotions than those with lower levels of self-compassion (24). CFT is considered a promising approach to easing various psychological disorders (25-28), but, since CFT is a relatively new psychotherapy, the evidence base regarding its efficacy is insufficient. Kirby et al.’s meta-analysis (26) of compassion-based interventions included papers on 21 randomized clinical trials (RCT), but only three were on CFT, and only one of those papers (29) reported on a clinical sample. Braehler and colleagues (29) compared a group receiving CFT plus treatment as usual (TAU) (n = 22) to a TAU-only group (n = 18) among individuals with schizophrenia-spectrum disorders and found a greater increase in compassion in the CFT plus TAU group. In support, a recent pilot RCT compared CFT plus TAU (n = 11) to TAU-only (n = 11) in a sample of outpatients with various types of eating disorders (30). They found more improvement in self-compassion, fear of receiving compassion, shame, and eating disorder pathology in the CFT plus TAU group than in the TAU-only group.
CFT studies not combined with TAU or cognitive behavioral therapy are scarce, and most of them did not use control groups (25, 28). One exception is a nonrandomized transdiagnostic group at a mental health service in Ireland that treated patients with either group-based CFT (n = 58) or TAU (n = 29) (31). The CFT group showed more improvements than the TAU group in psychopathology, fear of self-compassion, and social safety. In addition, group-based CFT has been applied in observational studies without control groups and found to benefit individuals struggling with personality disorders (32), dementia (33, 34), eating disorders (35), and complex MH problems (5, 36).
In sum, more research on stand-alone CFT implemented in clinical settings is needed. Studies conducted to date mostly included small samples without control groups, making generalization of findings problematic. It also is necessary to explore whether group-based CFT might help adolescents experiencing MH problems and their parents. CFT seeks to help people face their challenging feelings with courage, wisdom, warmth and care. This can help adolescents and their parents to be more compassionate toward themselves, rather that blaming themselves for the difficulties in life. The present study’s first aim was to explore whether group-based CFT would help improve self-compassion and reduce stress in a transdiagnostic group of adolescents experiencing complex MH problems and their parents. Since CFT was specifically developed to increase the capacities to experience compassion, self-compassion was chosen as primary outcome. The second aim was to compare the levels of self-compassion and stress participants receiving group-based CFT with adolescents and their parents receiving TAU.
The Regional Ethical Review Board in Gothenburg, Sweden approved this study (Number: 330-16). Participants were assured of data confidentiality, and permission to use data for reports that ensured the protection of the participants’ identities were obtained from all participants through informed written consent to participate, and their parents consented to their and their children’s participation.
The sample consisted of 43 adolescents (CFT, n = 19, TAU, n = 24, aged 14–17 years; 83.7 % female) under psychiatric care at a child and adolescent psychiatric outpatient clinic in the County of Södra Älvsborg, which includes both rural and urban areas, in Sweden, and their parents (n = 77; 61 % female). The clinic specializes in treating complex MH problems in children and adolescents aged 3–18 years, which is free of charge. Patients were included if they were: (1) aged 14–17 years, 2) receiving treatment at the clinic, (3) were experiencing one or multiple diagnosable mental health conditions according to the Diagnostic and Statistical Manual of Mental Disorders (DSM-5). Meanwhile, the exclusion criteria were: (1) acute suicidality or trauma, (2) acute psychotic symptoms, (3) acute serious self-harm, (4) substance abuse/misuse, (5) autism, or (6) inability to speak Swedish well enough to participate in the study. The reason for the exclusion criteria 1–5 were that in these cases, the adolescents need individualized treatment, and a group setting without individual support may have detrimental effects. The psychiatric clinic has special treatment units for patients experiencing these problems, such as a dialectical behavioral therapy unit for those engaging in serious self-harm.
This study was part of a larger research project that analyzed qualitative data on participants’ CFT experiences, and these results will be presented in two additional papers—one unpublished article reporting on qualitative data on adolescents’ experiences, and the other a published article about parents’ experiences with group-based CFT (37). The Transparent Reporting of Evaluations with Non-randomized Designs (TREND) statement was followed (38). The participants were recruited continuously from January 2017 until October 2018. Five CFT groups were observed, each of which had a minimum of three adolescents.
The treatment team determined eligibility during intake based on the inclusion and exclusion criteria. During intake, all individuals who met the inclusion criteria were asked if they wanted to receive group-based CFT. Individuals who did not want to participate in the group-based CFT were asked if they wanted to be in a control group receiving TAU. Figure 1 illustrates the enrollment and attrition process.
Although the study was intended as a randomized controlled clinical trial of CFT, only 30 adolescents volunteered for the group-based CFT, of which 19 (63 %) completed the questionnaires post treatment. Since blinded random assignment of volunteers would have created groups considered too small for this study, all of the volunteers were assigned to the group-based CFT. During the eight-week CFT treatment period, nine participants had no other MH support, four participants had one medical follow-up with a nurse or psychiatrist, and six participants had one meeting with a therapist or social worker for individual (e.g., at school) or family (with a parent) meetings.
The 24 volunteers for the control group who remained in the study until the end received TAU, which included cognitive behavioral therapy (CBT), systematic psychological treatments, clinical assessments, and psychosocial support. The participants in the control group were limited to TAU throughout the study period. Systematic psychological treatment refers to a mix of clinical assessment, network meetings with school and other support teams, and psychological treatments such as CBT, psychodynamic therapy, or family therapy.
A pretreatment questionnaire about self-compassion and perceived stress was distributed to all the participants (CFT = 83, TAU = 117). The questionnaire was distributed again to the 47 CFT participants during their last CFT session and to the 73 remaining TAU participants eight weeks after their TAU treatments began, presented in Table 2 as post treatment. A six-month follow-up questionnaire was distributed, but response rate was low (CFT group: adolescents, n = 12; mothers, n = 11; fathers, n = 7; TAU group: adolescents, n = 6; mothers, n = 9; fathers, n = 6). When no response was received, reminders or the opportunity to decline further participation was sent via a letter, for a maximum of three times, before it was reported as drop-out. Reasons for declining are shown in Figure 1.
The intervention was adapted and developed from the transdiagnostic CFT approach (18, 39). The feasibility and appropriateness to the participants of this group-based CFT intervention was assessed in a pilot study, which included six parallel CFT groups from August 2015 through December 2016, and the intervention was constantly adapted by considering feedback from the participating adolescents and parents. The most notable adjustments were lengthening the intervention from five to eight sessions and reducing the proportion of didactic teaching in favor of discussions, role play, and practical compassion work, such as making bracelets featuring a compassion statement (CFT adolescent group). Based on the results of the pilot study, we planned to include 34-36 participants in each CFT and TAU group, to detect an effect size of 0.5 with a power of 80 %.
The parents’ group leader was a clinical psychotherapist, and a clinical psychologist and psychotherapist (this paper’s first author) delivered the adolescents’ CFT, except for one group, which was led by a CFT-trained social worker. All of the group leaders completed CFT training at the Compassionate Mind Foundation in the United Kingdom. Two assistants not trained in CFT, but with significant experience in psychiatry and with academic backgrounds, provided support.
The group-based CFT sessions were eight parallel sessions lasting about two hours each. The adolescent and parent groups met at the same time but separately. The sessions followed the CFT principles, but they were adapted to meet the participants’ needs. The intervention had two phases. First, psycho-education about humans’ “tricky brains” and ways that we can understand negative threat emotions in an evolutionary perspective were covered (18, 40). The participants were encouraged to reflect on the balance in their threat, drive, and safeness system. The second phase was about attending to and perceiving the reasons for emotions and learning about the ways people are differently organized by various emotions. The subjects engaged in mindfulness and compassion-based exercises, such as soothing breathing, creating a safe place, visualizing a compassionate self-image, and exploring the role of one’s self-criticism. The structure and content of the sessions were the same for parents and the adolescents, and they were encouraged to complete homework assignments together at home.
Responses to the 12-item Self-Compassion Scale–Short Form (SCS-SF) were used to measure self-compassion (41). The response options are on a five-point scale ranging from 1 = almost never to 5 = almost always. Three dimensions of self-compassion are measured by four questions each: (1) self-kindness (e.g., “I try to be understanding and patient towards those aspects of my personality I don’t like”) versus self-judgment (e.g., “I’m disapproving and judgmental about my own flaws and inadequacies”), (2) common humanity (e.g., “I try to see my failings as part of the human condition”) versus isolation (e.g., “When I am feeling down I tend to feel like most other people are probably happier than I am”), and (3) mindfulness (e.g., “When something painful happens I try to take a balanced view of the situation”) versus over-identification (“When I fail at something important to me I become consumed by feelings of inadequacy”). The scores ranged from 12 to 60, and higher scores indicated more self-compassion. In the study by Raes et al., Cronbach’s alpha coefficient on the 12 items was .86 (n = 871, age 18–64). In a sample of Swedish adolescents, the coefficient was .87 (n = 226, ages 15–19) (42). Cronbach’s alpha coefficient at baseline in the present sample of adolescents and their parents (n = 200; CFT = 83, TAU = 117), was .89.
The Swedish version of the Perceived Stress Scale (PSS) (43, 44) was used regarding pretreatment and post-treatment perceived stress. The 14-item PSS concerns the perception of life as unpredictable, uncontrollable, and overwhelming during the previous month. Responses were on a five-point scale from 0 = never to 4 = very often. Seven items are negative and seven items are positive. The seven positive items were reverse-coded to calculate a total score, higher scores indicated more perceived stress, and the total scores ranged from 0 to 56.
The psychometric properties of the Swedish 14-item PSS had strong internal consistency in two Swedish adult samples. In a sample of women and men (n = 171, ages 20–45 years), the Cronbach’s alpha coefficient was 0.90, and the coefficient was 0.84 in a sample of women with stress-related disorders (n = 84, ages 35–55 years) (44). In a sample of Swedish adolescents, the alpha coefficient was 0.88 (n = 226, ages 15–19 years) (42). In the present sample of adolescents and their parents (n = 200; CFT = 83, TAU = 117), Cronbach’s alpha coefficient at baseline was .90.
The statistical analyses were performed using SPSS (ver. 25). Descriptive statistics were used to assess the normality of the distributions of scores and the means of the SCS-SF and PSS scores. Paired samples t-tests and independent samples t-tests assessed differences within and between groups. Bonferoni correction for number of statistical tests (45) was used, and adjusted p-values are presented in parentheses. To compare differences between the CFT and TAU groups, change scores were calculated by subtracting post-intervention scores from pre-intervention scores. Effect sizes were calculated using Hedges’ g, which is similar to Cohen’s d, because it is appropriate for sample sizes and/or standard deviations differ and it is frequently used to compare experimental to control groups.
Table 1 presents the descriptive statistics on the adolescents’ characteristics, MH diagnoses, the TAU adolescents’ types of treatment, parents’ ages, and mean number of sessions. There was no statistically significant difference between the CFT and TAU groups regarding age. The CFT parents received significantly more sessions than the TAU parents received, but the difference in the adolescents’ numbers of sessions was not significantly different between the groups.
Table 2 indicates there were no significant differences between the pretreatment and post-treatment mean scores on either the SCS or PSS for the CFT adolescent group or the TAU adolescent group. Among the four parent groups, only the TAU mother group was significantly different post-treatment, and these mothers experienced a significant mean increase in self-compassion (33.9 increased to 37.1).
There was no significant difference between the CFT and TAU adolescent groups in the extent to which self-compassion changed (p = .73). The effect size of 0.10 was considered very weak. Similarly, no significant difference was found regarding PSS change between the CFT adolescents (mean = -1.3) and the TAU adolescents (mean = -2.1), and the effect size of 0.14 was weak.
The results for the mothers and fathers were similar to those for the adolescents. There were no significant differences in SCS or PSS change scores between the CFT mothers and TAU mothers, with an effect size of 0.04. Moreover, there were no significant differences between the CFT fathers and the TAU fathers in SCS change (effect size = 0) or PSS change (effect size = 0.25).
This study’s results revealed no significant differences within or between group-based CFT and TAU treatments regarding self-compassion and perceived stress after eight weeks of CFT or TAU intervention (except for the control group mothers, whose mean self-compassion score significantly increased), and the effect sizes were small (g ≤ 0.25).
Ten of the 19 CFT adolescents and 15 of the 24 TAU adolescents who completed the sessions had higher self-compassion scores and lower perceived stress scores after the treatment; however, those results were obscured by the effects on the means of the scores that did not change or decreased. Therefore, the effect sizes were weak on both measures. The small effect sizes in the present study were anticipated because of the short treatment period for adolescents experiencing complex MH problems. The results are in line with a previous finding, which showed that efficacy of psychotherapy when compared to TAU is limited with an effect size of 0.30 or below (46). However, according to a systematic review of adolescent depression (47), most studies used clinicians’ measures of symptom change, while patients’ and parents’ perspectives, as well as their personal growth, interpersonal relationships, and school performance were not assessed. A qualitative study of parents’ experiences showed that the CFT intervention enabled parents to find their agency and inner confidence to support their children (37). The parents found it easier to care for their children, when they were able to care for themselves. This qualitative study suggested that CFT could be a promising approach to support parents of young people experiencing MH problems. Future studies need to explore the factors affecting lack of participation or withdrawal of participation in both CFT and TAU conditions, in order to develop interventions tailored to the needs of the patients.
CFT was developed to reduce self-criticism and shame among individuals with multiple MH problems (31), which could help adolescents who experience complex MH problems. In separate analyses, we found that some of the adolescents in the CFT group found the intervention difficult, particularly regarding mentalizing with themselves and/or others and trying to identify various emotions. Some adolescents obtained lower scores on self-compassion after the treatment, which might be related to their lack of awareness of their degree of self-criticism before CFT. Furthermore, some patients had difficulties with the abstract content of the CFT program. That finding supports McFarlane’s results on compassionate mind training in a school setting, in which some adolescents were helped, whereas others were challenged (48). More research is needed to help us understand the characteristics of those who benefit from CFT and those who might be overwhelmed by it. A single-subject design might be particularly appropriate in this case.
Regarding the parents, the fathers scored higher on SCS and lower on PSS than the adolescents or the mothers on the pretreatment and post-treatment tests. This finding supports previous studies’ findings that men are more self-compassionate than women (49).
Conducting research in applied contexts is challenging, and the present study has a number of limitations such as low sample sizes, insufficient description of the content in the TAU condition, and risk of self-selection bias. The nature of sample selection made it impossible to randomly assign the adolescent participants to the treatment and control groups because only few adolescents volunteered for the group-based CFT intervention. Furthermore, the response rate for the post-treatment questionnaire was low (CFT = 57 % (total n = 47), and TAU = 62 % (total n = 73), which, although similar to previous studies of psychological interventions with adolescents (48), limits our ability to draw conclusions. Although the group-based format was not mentioned as a reason for declining CFT, this may also have contributed to non-response. The majority of the adolescents in both groups were female, suggesting that males were less likely to want to participate. Although the reasons for non-participation are not known, there might have been more girls than boys who met the criteria at the enrollment stage (n = 720), or there might be a gender-based stigma related to self-disclosure. Last, the SCS and PSS measures might not have captured the actual change resulting from the intervention. Given that the adolescents and their parents had the same change results, it is plausible that perceived stress and self-compassion changed similarly for all the participants because of the help they received from the mental health service.
The present study conducted a nonrandomized clinical trial to examine the effects of an eight-week, group-based CFT on a sample of vulnerable adolescents. No previous studies have examined whether CFT is a helpful psychotherapy method in adolescents having mental health problems, and therefore this study contribute to the field. Adolescents with mental health diagnoses have complex problems related to their developmental stage. Further studies are need to investigate whether a compassion-focused approach can help these children and their parents to cope with their challenging situations. There were no significant changes in self-compassion or perceived stress after treatment in the experimental or control group of adolescents or their parents, and there were no differences between the experimental and control groups (except for the control group mothers, whose mean self-compassion score significantly increased). This suggests that some patients seem to benefit from group-based CFT in ways similar to the benefits of specialized therapeutic approaches. However, some patients might need a more individualized approach. Further research into the utility of CFT for adolescents with MH problems and their parents, as well as the long-term clinical effects of CFT for this group is needed.
Group-based CFT for adolescents experiencing complex MH problems and their parents may be a complement to current psychiatric treatments. For the adolescents, individual sessions between the group sessions, may be necessary to capture individual needs, and also to detect any adverse effects of the treatment.
Some participants found the abstract content of the CFT program including visualizing and mentalizing hard. On the other hand, the group-based format made it possible for the participants to share their thoughts, and listening to others can be valuable for those who had mentalizing difficulties.