Introduction

A diagnosis of cancer is frequently experienced as traumatising by the individual and their family, with over a third of individuals diagnosed meeting criteria for clinically significant levels of distress (Carlson et al. 2004a; Mitchell et al. 2011). In addition, fear of recurrence has been indicated to be pervasive and may be experienced five or more years after the initial diagnosis (Koch et al. 2013). Clinical guidelines have proposed that psychological interventions should have an integral role in the services offered to individuals with cancer (National Institute for Health and Care Excellence 2004), with mindfulness-based interventions (MBI) representing a leading intervention (Shennan et al. 2011).

While mindfulness practices can be seen to have been incorporated into a number of third-wave behaviour therapies (e.g., Linehan et al. 1999; Hayes et al. 2006), MBI refers to the 8-week programmes of Mindfulness-Based Stress Reduction (MBSR; Kabat-Zinn 1982) and Mindfulness-Based Cognitive Therapy (MBCT; Segal et al. 2002). The benefits of MBI have gained empirical support across numerous studies, with meta-analyses demonstrating robust positive effects on a number of psychiatric, functional somatic, and stress-related symptoms (Baer 2003; Grossman et al. 2004; Khoury et al. 2013). Specifically in cancer populations, a number of significant benefits have been reliably associated with MBI participation; these include reduced symptoms of depression, anxiety, stress, fatigue, and fear of recurrence (Foley et al. 2010; Lengacher et al. 2009; Speca et al. 2000; van der Lee and Garssen 2012; Würtzen et al. 2013), and improved sleep quality, quality of life, energy and physical functioning (Carlson et al. 2004b; Lengacher et al. 2009). A recent meta-analysis of RCTs of MBI with cancer populations indicated significant effect sizes (Hedge’s g) of 0.44 and 0.37 for depression and anxiety respectively compared to wait-list controls (Piet et al. 2012). A relatively smaller number of qualitative studies have been published exploring the experience of persons who have undergone cancer treatment who participate in MBI. Participants in these studies have described benefitting from the cultivation of acceptance, sustaining mindful control, and taking responsibility for what they could (Dobkin 2008), in addition to the nonjudgmental and accepting approach of facilitators, the group process and sharing within the group, and increased awareness of the present moment (Kvillemo and Bränström 2011).

Despite the existence of a substantive literature which demonstrates associations between MBI and a range of positive outcomes, the processes by which MBI may affect these positive outcomes remains under debate. Several theoretical accounts have been published which have proposed various psychological processes as implicated in positive changes associated with MBI, including exposure, regulation of attention and emotion, awareness of body and action, cognitive change, and acceptance (Baer 2003; Baer et al. 2006; Brown et al. 2007; Holzel et al. 2011; Shapiro et al. 2006). One concept which has been consistently empirically associated with positive outcomes in MBI is self-compassion (e.g., Kuyken et al. 2010).

From a social psychological and Buddhist perspective, Neff articulated a tripartite definition of self-compassion which has been frequently cited. This definition is composed of self-kindness—being kind and understanding to oneself in instances of pain or failure, common humanity—perceiving one’s experience as part of the larger human experience, and mindfulness—holding painful thoughts and feelings in balanced awareness (Neff 2003a; p.225). These three components are understood to be interdependent and to mutually enhance and engender each other. The Self-Compassion Scale (SCS) was developed by Neff to measure these three components and their polar opposites, and has been employed extensively in the literature since its publication (Neff 2003a). Neff’s work has also drawn clear distinctions between self-compassion as compared to self-esteem, self-pity, self-indulgence, and other self constructs (Neff 2012).

From an evolutionary, developmental, and neurophysiological perspective, Gilbert articulated a theory of compassion which has become the foundation of a multimodal approach to psychotherapy: Compassion-Focused Therapy (CFT; Gilbert 2009). Fundamental to Gilbert’s theory is the distinguishing of three affect regulation systems—the threat system, drive system, and soothing system. Mental health difficulties are theorised to be underpinned by over-activity of the threat and drive systems and corresponding under-activity of the soothing system, with this pattern of activation linked to early experiences of relationships with significant others (Gilbert 2009). Gilbert has further outlined six attributes for engaging with compassion: motivation or care for wellbeing, sensitivity, sympathy, distress tolerance, empathy, and non-judgement; with self-compassion involving applying these attributes in relating to the self.

Empirical work over the last decade, employing the SCS, has indicated that individuals with higher levels of self-compassion tend to experience less mental health difficulties (MacBeth and Gumley 2012), greater subjective well-being (Allen et al. 2012), more social connectedness (Neff 2003a), increased acceptance and functioning in the face of pain (Costa and Pinto-Gouveia 2011; Wren et al. 2012), reduced rumination and worry (Raes 2010), more secure attachment (Neff and McGehee 2010; Raque-Bogdan et al. 2011), and use more adaptive emotional coping strategies (Neff et al. 2005). Research exploring the relevance of self-compassion in cancer is also beginning to emerge. A recent study with a cancer population revealed that self-compassion is associated with better psychological functioning, predicting reduced stress and depression symptoms and increased quality of life (Pinto-Gouveia et al. 2013). Another study also identified that self-compassion may act as a mediator between body image disturbance and psychological distress in breast cancer (Przezdziecki et al. 2013). With regard to intervention research, self-compassion has been studied most frequently in the context of MBI.

While the cultivation of self-compassion is not a direct or explicit focus of MBI, these interventions have nonetheless been indicated to be associated with significantly increased self-compassion (Dunn et al. 2012; Hollis-Walker and Colosimo 2011; Rimes and Wingrove 2011; Shapiro et al. 2005; Shapiro et al. 2007). In relation to cancer, the concept of self-compassion has also arisen unprompted as a finding in qualitative research with a cancer population following MBI (Kvillemo and Bränström 2011). A noteworthy study by Kuyken and colleagues demonstrated that the impact of MBCT on depressive symptoms was mediated by increases in self-compassion and trait mindfulness scores at 15-month follow-up (Kuyken et al. 2010). The results of this study are significant in going beyond demonstrating that MBI increases self-compassion to proposing that MBI’s may be effective in part because they cultivate self-compassion.

On the basis of research findings, the second edition of the MBCT manual now explicitly states that MBCT aims to cultivate mindfulness and self-compassion (Segal et al. 2013). However, despite a number of studies pointing to a relationship between self-compassion and MBI, the psychological mechanisms by which self-compassion may be increased in MBI remain unclear. Ingredients of MBI which have been suggested to facilitate the development of self-compassion include the continual emphasis on curiosity, kindness and befriending of experience (Feldman and Kuyken 2011), the group process (Allen et al. 2009), and the facilitator’s embodiment of the practice (Segal et al. 2013).

While the relationship between self-compassion and MBI has been repeatedly indicated in quantitative research, it remains unclear as to how this occurs or the processes through which self-compassion may be enhanced in MBI. This may be related in part to a virtually exclusive reliance on a single self-report measure, the SCS. The lived experience, meaning and sense-making of self-compassion have thus remained largely unstudied, with the exception of two identified published qualitative studies which specifically explored experiences of self-compassion outside of an intervention context (Patsiopoulos and Buchanan 2011; Pauley and McPherson 2010). The population of persons who have undergone treatment for cancer may be especially appropriate to considerations of self-compassion in light of the potential for multiple levels of suffering arising from a diagnosis of cancer, in addition to the well-established body of MBI research for this population.

The present study aims to qualitatively explore participant descriptions of self-relating and self-compassion in the context of an 8-week MBCT programme adapted for a cancer population. Three research questions are of interest. Firstly, how do participants describe changes in self-relating following MBI? Secondly, how do participants experience engaging with self-compassion? Finally, what processes from participating in MBI might facilitate changes in self-compassion?

Method

Participants

Ten individuals (9 females, 1 male; age range = 33–63 years) participated in the research interviews. Seven participants had previously received diagnoses of breast cancer, and the remaining three had been diagnosed with rarer forms of cancers. Time since finishing hospital-based treatment (e.g. surgery, chemotherapy, or radiation therapy) ranged from 4 months to 9 years; all participants had finished hospital-based treatment at time of commencing MBI. Inclusion criteria consisted of being a patient of the Oncology department of the hospital, being over 18 years of age, having participated in the most recently provided MBCT group, and provision of full written consent. Exclusion criteria consisted of diagnoses of brain injury, cognitive impairment or serious current psychiatric illness, and attending less than six of the eight sessions of the MBCT group. No participants met any of the exclusion criteria.

Procedure

The format and content of the group followed closely the structure and resources outlined by Segal et al. (2002) in their 8-week MBCT group. A notable exception however was the replacing of the focus on depression with stress and the fight-or-flight response (Kabat-Zinn 1990); this was intended to more closely meet the needs of the cancer population. The group sessions were led by a clinical psychologist who had completed formal MBCT teacher training, with assistance from a counselling psychologist who had completed two personal 8 week mindfulness-based interventions.

Measures

Interview Schedule

The semi-structured interview schedule employed a funneling approach, with questions targeting the following domains: the self before the mindfulness group; motivation to attend the mindfulness group; experiences of the mindfulness group; perceived changes or impacts of the mindfulness group, with a progressive focus on changes in self-relating; the potential relevance of self-compassion, as well as barriers to experiencing self-compassion; and finally suggestions for the mindfulness group or research. Examples of questions included were “was there anything in particular that you noticed or learnt about yourself from practicing mindfulness?”, and “how would you say you treated yourself before the group started?”.

Due to the perceived risk of demand characteristics, the term self-compassion was not explicitly written in the information and consent form. Instead, the terminology used referred to a more general focus on self-relating, with self-compassion being explicitly referred to towards the end of the interview and named plainly in the debriefing sheet.

Interview Procedure

Interviews took place 4–6 weeks following the completion of the MBCT group, and were conducted by the primary researcher who did not facilitate the group. Interviews lasted on average 81 min duration (range 56 to 102 min).

Ethical Considerations

This research study was approved by both the hospital and the university ethics boards, and full informed written consent was obtained from every participant.

Data Analyses

A descriptive-interpretive approach, as outlined by Elliott and Timulak (2005), was adopted in analysing the data. A central feature of this approach is its recognition of and accounting for the broad organising structure often brought to qualitative analysis; this parallels techniques used in grounded theory and consensual qualitative research (Hill et al. 1997; Strauss and Corbin 1998). This structure is seen to be implicit in the formation of the research questions and the interview schedule, both of which are informed by an advance reading of the literature and the researcher’s pre-understandings of the topic. This domain structure remains flexible however and open to restructuring in order to best fit the data. In the present study, the initial domain structure had seven domains which were later collapsed into the three overarching domains in which the results are presented (see below).

Interview recordings were transcribed, then read and re-read to identify and extract meaning units: segments of transcribed text which were perceived to communicate something relevant to the research questions. These meaning units were then sorted under domain headings and collated across participants. Due to the large number of meaning units assigned to each domain, each meaning unit was reduced to a core idea (cf. Hill et al. 1997) capturing the key words contained in the text. These were then compared within each domain to observe and cluster together core ideas with similar meanings. These clusters then evolved to form categories and subcategories through a process of constant comparison (Glaser and Strauss 1967) and labeling. Meaning units for each domain were then re-sorted under the headings of categories and subcategories as formed from the abbreviated core ideas. Revisions in category structure were undertaken at this point through studying meaning units within categories, and category labels were modified to best represent the content of the meaning units.

Both descriptive and interpretive elements were brought to the formation of categories, where the researcher aimed to stay truthful to the common-sense meaning and original language of participants while also inevitably bringing previous understandings and theoretical ideas to bear in the analysis. The theoretical framework of the analyst (the first author) was informed by psychodynamic and third-wave CBT approaches, while the auditor’s (second author) was informed by humanistic theories, specifically emotion-focused. The third author provided critical feedback on an early configuration of themes from the analysis, and the fourth author assisted in the process of obtaining ethical approval.

A number of procedures were incorporated to ensure the trustworthiness of the data and analysis, as informed by the seminal work of Elliott et al. (1999). This included the keeping of a reflective diary, grounding the interpretation with frequent participant quotes, gaining and incorporating feedback from the analytic auditor and third author, and engaging in careful archiving and constant critical challenge of analytic steps.

Results

Shifts in Relating to Self and Others

This domain is presented as juxtapositions of categories of experiences in relating before and after the mindfulness group. All categories are presented in Table 1 below. This section will present each of the three results domains in turn, including a summary table of categories for each domain. Pseudonyms were used to protect participants’ confidentiality.

Table 1 Categories, subcategories and occurrences in the shifts in relating to self and others domain

Participants described how, before starting the mindfulness group, they experienced a pressured drive to be constantly engaged in activity. A variety of consequences related to this high-speed, incessant striving were observed, mainly centring on a lack of self-focused time, a loss of identity, and feelings of significant strain; for example Heather described feeling wrung out, I was just stretched…It was just too constant, there was no time off for me. When describing changes since completing the mindfulness group, participants described features of unhooking from this drive, for example having taken a step back (Irene), gave myself time (Debbie), and increased permission for the self to do nothing (Gina). Outcomes of this moving away from engaging with this pressured drive were portrayed as a reduced sense of speed and striving, and an increased sense of space and time for the self in the present moment.

Prior to the mindfulness group participants conveyed having not been aware that they were stressed, having actively rejected their felt bodily experience, or feeling disconnected from their sense of their self, with reflections of having lost myself (Beth). Participants attributed to their mindfulness group participation an enhanced ability to bring their own patterns of reacting into awareness, for example noting a greater ability to recognise the physical symptoms of what’s going on, and the effect it’s having (Cara). Outcomes linked to this enhanced awareness included reduced engagement with ruminative thought and an increased sense of choice and agency.

Frequent descriptions were made which indicated how, prior to the mindfulness group, participants’ self-relating appeared to be dominated by harsh or critical attitudes. Accounts also shared a sense of feeling trapped and powerless in relation to negative emotional cycles, such as feeling really low…I was in a hole (Ellen), and coping through avoidance. Reflecting on changes following group completion, many participants indicated that the valence of their self-relating had shifted towards a more positive and kinder focus. Underlying this shift appeared to be the ability to notice and release attachment to self-critical thought patterns, or let it go (Gina, Jackie). Relief from emotional tension, particularly low mood and anxiety, was also conveyed. Underlying this shift appeared to be the discovery that emotions can dissolve through approaching and observing their embodied experience; described as the more you get to know it, the more it dissipates (Alan).

When reflecting on their sense of self before they started the mindfulness group, participants described how they would have tended to overextend themselves in relationships with others, with some noticing how they were afraid to displease (Cara). Some further identified a double standard in offering consideration and sympathy to others but not to the self. Participants noticed a change following the group in how they were inclined to take more time to consider the self before committing to others, the outcome of which was described as increased assertive behaviour, or being able to stand up for myself more in taking on less (Cara). Some qualified, however, that this ability to assert boundaries with others was not at the expense of generosity and altruism.

Prior to the mindfulness group, participants described themselves as feeling very alone and disconnected from others; as Heather described I saw myself outside the herd, removed from it, not able to be part of it. Protective functions of this separateness were identified retrospectively, including the need to conceal one’s vulnerability from others or to put the brave face on (Cara). Since completing the mindfulness group, participants reflected on how they felt an increased sense of belonging and connection with others; noting how I don’t feel like I’m out of the herd now (Heather). This connection with others appeared to be facilitated by a shift from being more guarded to expressing oneself more openly and authentically.

Experiences of Engaging with Self-Compassion

In the latter part of the interview, participants were asked specifically about experiences of self-compassion and potential blocks to self-compassion. The findings from this enquiry are reported under the following domain, with sub-domains of Defining features of self-compassion and Influences impacting self-compassion. Categories are listed in Table 2 and elaborated below.

Table 2 Sub-domains, categories and occurrences in the experiences of engaging with self-compassion domain

The most frequent description of practicing self-compassion centered on taking time for the self. Participants described taking more opportunities (Beth) to do things that they enjoyed and that provided them with nourishment (Irene). This approach to practicing self-compassion was explained as being reinforced cognitively and emotionally by a shift in the way that {you} think about (Heather) giving the self time, and letting go of feeling guilty (Gina). It was also emphasised, however, that giving the self this time remained in balance with engaging in tasks of everyday living, described by Ellen as you build it into your day.

Half of participants’ accounts of practicing self-compassion captured a sense of relating to the self with reduced pressure and increased kindness. This was described as a shift in attitudes towards the self, being more accepting of myself (Jackie), and associated with outcomes of reduced subjective stress, or not getting het up (Ellen). This shift in self-relating was experienced for some as an internal soothing voice, as Cara conveyed: when I think I’m failing I then have to say sure you poor thing, I mean you were trying, it’ll be okay.

A third feature of practicing self-compassion was concerned with not automatically yielding to others’ agendas, but instead giving greater consideration to the self. This appeared to involve re-dressing an imbalance for those who recognised that they were inclined to prioritise the needs of others, as described by Gina: if this is the right thing for me to be doing for myself right now, that that’s as much as I can do, whereas I would have usually been the reverse. Moreover, rather than reserving acceptance, forgiveness and care only for others, these qualities were described as being directed towards the self when engaging with self-compassion.

When reflecting on potential barriers to adopting self-compassion, some participants noted having grown up in very busy and task-focused households. Irene, for example, indicated having undergone a revision of this family-based schema of the need to be busy, observing just because you’re sitting down doesn’t mean you’re lazy. Some participants further saw Irish Catholic culture as incompatible with self-compassion in its perceived discouragement of showing affection or acknowledging suffering, but rather that you didn’t talk about your emotions (Fran).

When reflecting on how self-compassion may have been enhanced through completing the mindfulness-based intervention, participants observed that the message of treating the self with kindness was the core of everything (Cara). They noticed how messages of kindness, allowing, and congratulating appeared frequently in both the facilitator’s language and the guided meditations CDs, and suggested that self-compassion may have filtered in (Ellen) through this process. When asked if they had concerns around relating to themselves with kindness and compassion, participants expressed fears that they might lose self-control and not engage with everyday functional activity. A further threat perceived to be associated with self-compassion was the risk of being judged as selfish or wasteful by others if you do kind of run away with it (Gina). Finally, there were three participants who felt that they did not have an experience of a compassionate other growing up; as reflected poignantly by Cara: There wasn’t any single person I ever felt totally safe with.

Processes Which May Facilitate Changes in Self-Compassion in MBI

This domain reflects processes that participants described experiencing which they related to their engagement with the mindfulness group, as well as the broader context of their cancer experience in the context of participating in MBI. These processes were divided into sub-domains reflecting different levels of focus, as depicted in Table 3 and elaborated below.

Table 3 Sub-domains, categories and occurrences in the processes which may facilitate changes in self-compassion in MBI domain

Dedicating time to the self was perceived by many participants as an inherent quality of engaging in mindfulness practice; regarded as an investment in me (Ellen). This practice time was mostly described as a gesture of care and nourishment for the self, with some likening mindfulness practice to receiving a gift. Occasionally, however, practice was portrayed as a place where discomfort and strain were witnessed in being with the self.

Half of participants described a process of gaining greater awareness of areas of vulnerability in their self which they had previously avoided or turned away from. These areas included patterns of relating to others, self-protective defences, fear of cancer progression, and rejection of the embodied self post-cancer treatment. A common thread across participant accounts was a strengthened ability to turn towards and confront these difficult areas, as described by Cara It brought up issues but made me face them as well, and deal with them.

Strong emotions were described by participants as surfacing and being released during or following mindfulness practice sessions. This contributed to intensified emotional vulnerability for some, with Cara feeling it stripped me to the core. Others, however, described a sense of revelation and wonder in relating to this felt emotional experience, as Jackie reflected I never allowed myself to feel it maybe…it was just incredible.

A desire to build up psychological resilience was identified by half of participants when reflecting on their reasons for commencing the mindfulness group. Some described having come from a context of gruelling medical treatment, where they observed that there was no room for the emotional aspects of the self during this time. There was a recognition in accounts that it isn’t enough to get better physically (Cara), with this acknowledgement providing an impetus to be mentally stronger (Ellen).

A number of participants remarked on being struck by the level of empathy that they felt when relating to others in the group. This prompted some participants to re-appraise their own circumstances and health, with some noting that they were lucky or more fortunate compared to others’ suffering. Empathising with others also caused some participants to re-appraise their own self-relating through an increasing realisation of the double standard with which they related to others compared to their own self, as explained by Irene: why am I pushing? And it’s the seeing of the other people I think, that’s what does it. It’s the very much they’re human in front of me. I don’t see myself.

Participants related to the group as an environment where they experienced being on an equal plane and sharing a common intention with others. While a couple of participants identified with others’ body language as mirroring their own struggles, the most common medium of identification with others was through listening and sharing during guided group discussion. This led to insights for some around their shared patterns of reacting, as Heather described it holds a mirror up to you. This process of identifying with others’ experiences led participants to fundamentally recognise that they were not alone, and that they could draw from the common humanity inherent in their struggles, as noted by Irene it’s hard for everybody, and it was lovely to just hear ah God I’m human.

The approach of the facilitators was consistently framed as helpful, with descriptions of understanding, sensitivity and fairness. A couple of reflections pointed specifically to the value of guided, Socratic-style questioning in eliciting insight about the self.

Finally, some participants identified shifts in perspective which they attributed to their experience of cancer, particularly relating to mortality and the value of life. Two participants noted that it was hard to distinguish the different influences of mindfulness group participation and cancer.

Discussion

This study aimed to explore shifts in self-relating, in particular self-compassion, in the context of MBI. Participants described shifting from patterns of driving, disconnection, and preoccupation towards slowing down, giving the self time, enhanced awareness of self-experience, increased self-kindness, and relief from emotional tension. Participants further noted interpersonal changes following the group, including a shift towards more assertive communication and an increased sense of belonging and connection. Defining features of the lived experience of self-compassion were making time for and being less pressurising towards oneself, and accounting for one’s own needs in relationships. Factors identified as influencing the development of self-compassion included cultural and family backgrounds and fear of losing control. Finally, proposed facilitators of changes in self-compassion spanned increases in self-awareness and emotional engagement, identification with other group members, facilitator guidance, and the broader cancer experience.

Descriptions of a pressured drive involving high speed and relentless activity are similar to Gilbert’s portrayal of the drive system, in particular its focus on doing and achieving and associated feelings of being energised or hyped up (Gilbert 2009). Parallels are also apparent with mindfulness theory’s depiction of mental modes, specifically the driven-doing mode and its focus on activity and achieving goals (Segal et al. 2013). Participant descriptions of slowing down following the group, on the other hand, appear similar to Gilbert’s depiction of the soothing system as involving feelings of peacefulness, wellbeing and a state of non-seeking (Gilbert 2009), and mindfulness theory’s being mode (Segal et al. 2013). A shift from reliance on the drive system towards occupying the soothing system is supportive of increased compassionate self-relating in CFT (Gilbert 2009).

Accounts of becoming kinder and less hard towards the self suggest increased self-kindness and reduced self-judgement, consistent with Neff’s conceptualisation of self-compassion (Neff 2003b). Participant descriptions of less engagement and identification with thoughts are further consistent with a reduction in rumination. Decreased rumination has been associated with higher levels of self-compassion (Neff et al. 2007; Raes 2010; Samaie and Farahani 2011), as well as being an observed outcome of MBI (Hawley et al. 2013). Finally, participants noted how negative emotions dissipated, which appears to be consistent with proposals of improved emotion regulation skills in both the self-compassion (Neff et al. 2005, Vettese et al. 2011) and the MBI literatures (Chambers et al. 2009; Holzel et al. 2011; Robins et al. 2012).

Identified shifts towards more assertive behaviour are consistent with studies where higher levels of self-compassion were related to a reduced likelihood to self-subordinate needs, less desire to please, greater authenticity, and a greater ability to say no (Barnard and Curry 2011; Yarnell and Neff 2013). In the MBI literature, themes of improved communication, saying no, and balancing needs and responsibility in relationships have also been identified (Bihari and Mullan 2014; Hoffman et al. 2012). The increased sense of belonging, connection and greater ease with others portrayed by participants also echoes findings in both the self-compassion and MBI literatures related to greater social connectedness, emotional connection to partners, and a sense of wellbeing in relationships (Allen et al. 2009; Bihari and Mullan 2014; Neff 2003b; Neff and Beretvas 2013; Neff et al. 2007; Yarnell and Neff 2013).

As identified as a defining feature of self-compassion, themes of making time and creating space for the self, including enhanced self-care practices, have been previously identified in MBI research (Allen et al. 2009; Weitz et al. 2012). In relation to the contextual influences noted to impact the development of self-compassion, correlations have been identified between self-compassion and family functioning (Neff and McGehee 2010), and cross-cultural differences have also been observed (Neff et al. 2008). Furthermore, fears of compassion and perceived difficulties engaging with self-compassion have been outlined and scales developed to measure these fears (Gilbert et al. 2011; Gilbert and Procter 2006; Pauley and McPherson 2010).

In terms of group processes, experiences of empathy for other group members are consistent with MBI studies (Bihari and Mullan 2014; Arch et al. 2014), and may also reflect Gilbert’s empathy attribute of compassion (Gilbert 2009). The process of identifying with other group members appears to further hold parallels to Neff’s common humanity component of self-compassion, which captures feelings of connection and reduced isolation in recognising the universality of human suffering (Neff and Tirch 2013). Only one other study was identified which explicitly named the role of facilitators as instrumental in MBI, specifically referring to their non-judgemental and accepting approach (Kvillemo and Bränström 2011). The developers of MBCT have recently theorised that participants may develop self-compassion through the facilitators’ embodiment of kindness, particularly their personal warmth, attentiveness, and gentle responding (Segal et al. 2013). In compassion-focused approaches, the role of the therapist has also been emphasised in helping clients to experience ‘safeness’, acceptance and care in the relationship which may in turn activate the soothing system (Gilbert 2007).

The more open-ended approach afforded by the qualitative design is regarded as a strength of this study, allowing consistencies and variations in experiences to emerge without imposing a pre-defined framework of what constitutes experiences of self-compassion; this is in contrast to the virtually exclusive reliance in previous research on the Self-Compassion Scale (SCS; Neff 2003a). This study is seen to add further considerations to the scope and experience of self-compassion which extend beyond what may be captured by items on the SCS; in particular descriptions of shifts in pressured driving and interpersonal relating and the proposed potential processes underlying these shifts.

Due to the qualitative nature of this study, it is not possible to draw more definitive statements as to the nature, sequencing or directionality of associations between aspects of self-compassion and mindfulness. Other limitations relate to the female gender bias, which is consistent with previous research (e.g., Ledesma and Kumano 2009), and the 4–6-week latency between group completion and research interview, which precluded an exploration of longer-term effects. Finally, while these qualitative findings may offer transferability to other persons and contexts, they cannot be generalised indiscriminately (Flick 2006).

Future research would be valuable to further examine the processes and mechanisms which may contribute to increased self-compassion, described here as shifts from pressured driving, disconnection and not asserting one’s needs towards making time for oneself, connecting with oneself and others, and accounting for one’s needs in relationships. In particular, greater attention to the interpersonal processes involved in self-compassion and MBI would be valuable in future research in recognition of the predominantly intra-individual focus of previous research (e.g., Bihari and Mullan 2014; Neff and Beretvas 2013; Neff and McGehee 2010; Yarnell and Neff 2013). The role of facilitators was also raised in this study and linked to findings in previous research. This would be a valuable area for further research, which may wish to seek to elaborate on the qualities effective facilitators may embody, and the interaction styles of effective facilitators.

With regard to clinical implications, a number of participants identified perceived barriers to engaging with self-compassion, including familial, cultural and cancer contexts, and fears of loss of control. When working therapeutically to promote self-compassion, whether individually or in a group context, it may be valuable for clinicians to explore these fears and resistances with clients. This may include validating developmental and relational experiences which neglected the person’s needs for soothing and safeness. Barriers may also be compassionately reframed, for example as safety strategies as in Gilbert’s approach to compassion-focused formulation (Gilbert 2009). Finally, clients who have underdone cancer treatment may represent a group who are particularly vulnerable to dominant activity of the threat or drive systems, for example in facing their mortality or striving to recover or to distract the self.