Mental illness in a parent is a significant matter, affecting almost a quarter of children under 18 years of age (Maybery et al. 2009). A significant proportion of these children will experience mental health challenges at some point in their lives (Mowbray et al. 2006). Compared to their peers, these children are also more likely to experience developmental, social and emotional difficulties (Reupert and Maybery 2016), which can persist into adulthood (Dean et al. 2010; Rasic et al. 2014). The need to support these children throughout the lifespan has been well established (Goodyear et al. 2015; Hosman et al. 2009; Metz and Jungbauer 2019). However, to provide effective support, we need to understand the mechanisms driving these adverse outcomes.

There are several mechanisms involved in intergenerational patterns of mental illness, including genetics, parent- and child-related factors and socioeconomic issues. Beyond genetic contribution (Gatt et al. 2015), these children may be exposed to stress caused by social and economic disadvantages, family violence (Perera et al. 2014), stigma (Hinshaw 2005), lost education and unemployment (Yoshikawa et al. 2012). Moreover, some children may be exposed to unhelpful parenting behaviours (e.g. modelling of pathological coping strategies, less emotional availability, neglect or affective instability) (Reupert and Maybery 2007). Child-related factors may also play a part. Improving children’s mental health literacy, that is their knowledge and beliefs about mental health, maybe protective as it forms the basis of many preventative interventions for this population (Reupert et al. 2013). Nevertheless, other child-related factors, such as self-concept, have received less attention in the literature.

The psychological and social experiences of children in these families further inform current theories regarding outcome mechanisms for children. A systematic review and meta-synthesis of qualitative research conducted by Dam and Hall (2016) reported that children of parents with mental illness under 18 years commonly experience intense emotions, a sense of responsibility for their parents’ well-being, social isolation and shame in the context of parental mental illness. Murphy et al. (2011) found that adult children reflected on long-lasting strain in the parent-child bond, social isolation, persistent trust issues and the emotional “legacy” of parental mental illness. Other adult children highlighted the enduring impact of emotional needs gone unmet. Furthermore, these reviews indicate that children often assume caregiving roles. Caregiving may involve overt acts (watching over medication adherence, washing, cleaning and cooking) or covert acts of caregiving (monitoring the parent for signs of deterioration and adjusting behaviour accordingly). Current approaches to intervention may not always be sufficient to absolve children of a strong sense of responsibility for their parents’ well-being (Gladstone et al. 2014). Therefore, further inquiry is needed to examine additional approaches to support children to cope with difficult emotions, caregiving and reappraising responsibility.

Self-compassion is described as an adaptive way of responding to personal suffering (Neff et al. 2018). Operationally, self-compassion has been described interchangeably as a trait, a state, an attitude and a practice (Neff 2003a, b; Zessin et al. 2015). Neff (2003a, b) proposed that self-compassion involves non-judgemental openness to one’s own suffering, accompanied by a view of suffering as a shared element of the human experience and a motivation to alleviate one’s own suffering with kindness. Neff’s definition involves three interrelated components, each with two opposing parts, for example the presence of self-kindness versus self-judgement.

Self-compassion has gained attention due to its association with indicators of mental health and well-being (MacBeth and Gumley 2012; Zessin et al. 2015). Dispositional self-compassion has been associated with greater self-reported well-being (Zessin et al. 2015) and stress management (Galla 2016). The absence of self-compassion was also associated with psychopathology including depression, anxiety and stress symptoms (MacBeth and Gumley 2012). Self-compassion has been found to partially mediate the link between measures of maternal support, family functioning and well-being amongst adolescents (Neff and McGehee 2009). Moreover, self-compassion was found to be amenable to training. In their recent meta-analysis of 27 randomised controlled trials (RCTs), Ferrari et al. (2019) found that self-compassion interventions were moderately effective for increasing self-compassion and reducing self-criticism, depression, anxiety and stress compared to control conditions. The included RCTs involved a range of clinical and community samples. However, they only consisted of adults, predominantly females, living in Western countries. Nevertheless, the results indicate some promise for the effectiveness and clinical utility of interventions aimed at increasing self-compassion.

Self-compassion may be a clinically relevant concept for children and adult children of parents with mental illness. Firstly, self-compassion has been predicted to be important for people who experience high levels of shame and self-criticism (Gilbert and Procter 2006). Secondly, self-compassion has been proposed as an internal coping resource for adults who are caregivers (Iacono 2017; Robinson et al. 2017). It follows then that self-compassion may also be a resource for children who provide care to parents and who experience shame and self-criticism, such as those with parents who experience mental illness (Fjone et al. 2009).

Notwithstanding these potential links, there is a lack of research regarding self-compassion amongst children who provide care, including children of parents with mental illness. This scoping review aimed to identify research examining the experiences of children and adult children of parents with mental illness in relation to self-compassion.

Method

Procedure

Search Method

As this research sits at a point between two emerging fields of work (children whose parents have a mental illness and self-compassion), a scoping method, as described by Arksey and O’Malley (2005), was chosen and guided by the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR; Tricco et al. 2018). The scoping method allows for a review of a specific area of literature when a relative paucity exists, as is the case for this study. The research question and list of search terms followed a process of iterations as authors became familiar with the literature. Although “self-compassion” was included in the original search strategy, the search yielded zero results. Consequently, we chose to review available qualitative research regarding the children of parents with mental illness to determine whether themes related to self-compassion were present. Subsequently, self-compassion terms were removed, and qualitative research terms were added to the final search strategy (see Supplementary Materials for search terms).

A systematic search was last conducted on April 13, 2020, across the PsycINFO, SocIndex, CINAHL and MEDLINE databases. The search strategy included subject headings and MESH terms, along with truncation and adjacency commands to gather peer-reviewed literature of a qualitative nature relating to children of parents with mental illness. Search results were imported into an EndNote library, where duplicates were removed. The search strategy identified existing systematic reviews; their reference lists were hand-searched for papers not retrieved by the search strategy. The overall search strategy yielded 2048 unique papers.

Eligibility Criteria

The inclusion criteria meant that included papers (a) contained qualitative data; (b) reported on the experiences of children (children defined as individuals of any age based on a relationship to a parent a mental illness, inclusive of adult children) of parents with mental illness (mental illness defined as a diagnosed mental illness, screened as part of research or inferred by recruitment source; or an undiagnosed mental illness based on participant self-referral, i.e. a participant considered themself to be the child of a parent with a mental health issue or mental illness) and (c) included children’s experiences relevant to self-compassion (as defined by Neff (2003a, b) and Gilbert (2009), including self-kindness, self-criticism, self-blame, shame, self-stigma, sense of isolation, sense of common humanity or shared experience, subjugating own needs, mindfulness, overidentification, hypervigilance, self-care and self-soothing). The review was restricted to studies published in English after 1990. This date limit was imposed to retain feasibility and identify a breadth of key research in this field, whilst also acknowledging that the treatment of families where parents have mental illness has shifted significantly in the past 30 years (Reupert et al. 2015). Furthermore, papers were excluded if (a) children’s perspectives were not provided; (b) they focused exclusively on children of parents with a somatic illness, physical disability, cognitive disability or substance use disorder; (c) the primary purpose of the study was to evaluate an intervention; or (d) the study focused exclusively on cases of neglect and abuse.

Selection Process

ADS screened all article titles and abstracts to eliminate papers based on the inclusion and exclusion criteria. Subsequently, blind abstract screening was completed by ADS and AR using the Rayyan QCRI review software (Ouzzani et al. 2016). Any conflicts passed to full-text review. Fifty-five papers underwent blind full-text screening by A.D.S. and either J.S., M.L. or A.R. Five disagreements occurred at full-text review and consensus was reached by referring to the individual paper and the inclusion criteria. Results of the screening process are presented in the PRISMA flow diagram (see Fig. 1; Moher et al. 2009). Data were extracted from the published articles by ADS into a table tested and agreed upon by the research team (see Table 1 for a list of the included studies and their characteristics).

Fig. 1
figure 1

PRISMA flow diagram

Full size image
Table 1 List of included studies with quality appraisal, methodological characteristics and sample characteristics
Full size table

Quality appraisal was completed for each article using the Critical Appraisal Skills Program checklist for qualitative research (Critical Appraisal Skills Program 2018). The Critical Appraisal Skills Program (2018) was chosen as it is a simple, evidence-based tool for quality appraisal which can be used to generate a score out of ten for each study.

Data Analyses

To synthesise the qualitative content related to self-compassion, we adapted the directed qualitative content analysis (DQCA) approach proposed by Assarroudi et al. (2018). In the first stage of analysis, we defined main categories and a coding scheme consistent with self-compassion as outlined above in eligibility criteria. We decided to focus the analysis on the content of participant quotes and authors’ interpretations (if they were supported by participant quotes). The included articles were each repeatedly read to identify codes based on the coding scheme; codes were then extracted into the categorisation matrix. These codes were compared and contrasted to identify themes relevant to self-compassion. If a theme was observed once or more in a study, it was considered to be present. Themes were reported if they were observed repeatedly across the reviewed studies and were specifically linked to self-compassion. A.D.S., J.S. and A.R. have had significant exposure to children of parents with mental illness through their life experiences and/or work. Prior experience brings with it potential biases and preconceptions, along with understanding and knowledge which can facilitate abstraction (Smith and Osborn, 2015). Therefore, we engaged in self-reflection, journal keeping, robust discussion of themes, comparison to existing research and knowledge from clinical experience working with children of a parent with mental illness. Additionally, we sought to establish the trustworthiness of these findings through a 12-month period of engagement and immersion in the data, providing thick descriptions of themes and detailing the analytic process (Korstjens and Moser 2018).

Results

Study Characteristics

Thirty papers met the inclusion criteria. Of the 30 papers, three represented one data set from a study by Murphy et al. (2015, 2016, 2017), two were based on a single sample (Mordoch and Hall 2008; Mordoch 2010) and one involved analysis of data which was initially gathered in an earlier study which did not meet inclusion criteria for this review (Riebschleger et al. 1993 as cited in Riebschleger 2004). Ultimately, the review represents 27 unique sets of data, involving 374 participants aged from 6 to 78 years of age. Extracted data from each study are presented in Table 1. The total sample was approximately 32% male and 68% female; this is skewed by studies of adult participants (approximately 77% female), where recruitment was almost exclusively self-identification, compared to child studies (approximately 58% female), where recruitment was based on parental consent and engagement with mental health services.

Studies took place across ten countries including six from Australia, six from Canada, five from the USA, eight from eastern Europe, two from the UK and one from Iran. The overall sample was predominantly Caucasian, though exact data on sample ethnicity were not presented for all the studies.

There were some recurring aims throughout the research, including children’s experiences of everyday life and how they coped with or understood their parents’ mental illness. Some studies sought to examine experiences of caregiving, stigma, “burden” and loss. Others asked for children’s perspectives on the support they and their families received or should have received. Studies of adult children have also examined participants’ experiences of parenting their own children.

The studies included employed a range of methodologies to collect data including semi-structured interviews (15/27), focus groups (3/27), narrative interviews (3/27) and others (see Table 1). Researchers employed a range of analytic approaches including phenomenological analysis (6/27), grounded theory (5/27), thematic analysis (5/27), content analysis (5/27) and others (see Table 1).

The average Critical Appraisal Skills Program (2018) score was 9.2 out of 10. Almost half of the included papers did not provide detail regarding reflexivity or the relationship of the researcher to the participants (14/30 papers). Other flaws included not clearly articulating ethical considerations (4/30 papers), not describing the analysis procedure (4/30 papers), not clearly articulating the aims of the study (1/30 papers) and possible sampling bias (1/30 papers). Fourteen articles had no issues with quality based on the CASP.

Directed Qualitative Content Analysis

Three themes were found to be relevant to self-compassion; Isolation; Concealing emotions and Self-judgement (see Table 2 for the presence of themes). Themes are described below, alongside an aspect of Neff’s self-compassion to which they relate, with accompanying representative quotes and corresponding participant demographic details, where available.

Table 2 The studies which themes were present
Full size table

Common Humanity vs Isolation: Isolation

A key element of self-compassion is a sense of common humanity, a sense of connection with humanity on the basis that all people are vulnerable to experiencing suffering (Neff 2003a, b). Alternatively, one might feel isolated or separated from others by experiences of suffering (Neff 2003a, b). Participants in 25 of the 27 studies discussed these concepts. Isolation was observed across the studies, whilst a sense of common humanity was almost exclusively described in relation to peer support groups.

Children of parents with mental illness experience isolation in that they sense they and their families are wrong or different compared to normal families, “Our family was wrong…we were just a dysfunctional, broken family, we just weren’t normal like other families.” (Adult; McCormack et al. 2016, p. 335) Differences were noted when children compared their family with families they perceived as “normal”. Perceived “wrongness” stemmed from awareness and, often, internalisation of social stigmas regarding people with mental illness. Therefore, these children felt not just different, but ashamed and embarrassed.

…some people know that my Mom was mentally ill and some people don’t . . . It was a secret I was hiding. There was a lot of shame with it. It was like . . . my Mom is crazy, so they are going to think I’m crazy. (39-year-old-female; Carroll and Tuason 2015, p. 1068)

Parental mental illness was often a “taboo subject, an unspoken family secret, which was neither talked about in the family nor outside.” (34-year-old-male; Metz and Jungbauer 2019, p. 3) Keeping this family secret meant children were isolated with few opportunities to establish a sense of common humanity.

Accordingly, many children did not talk about their experiences of parental mental illness with friends. Moreover, most children believed that other people could not help anyway because, without lived experience, they would never understand.

Some of my best friends say that I can talk to them when I need to, and that is certainly a good thing, but regardless of how much I tell them, they will never really understand. I feel very alone in the middle of all this. (Adolescent; Trondsen 2012, p. 180)

In practice, this meant many children were alone to deal with troubling circumstances and feelings. This feeling of isolation remained for many adult children, described by this woman:

I think probably the overwhelming thing was how lonely and isolated and how little I talked and reached out to people and that it continues to this day. I have to make a conscious effort to reach out for help. I tend to try and deal with things by myself…then when I finally do ask for help it’s usually when I’m at crisis point. (Foster 2010, p. 3147)

However, participants described gaining a sense of common humanity when they shared their experiences with other children who had parents with mental illness. Such rare opportunities provided relief from the isolation and left children feeling validated, connected and understood.

People were going through like, “My dad is sick.” And I’m like, “Mine too.” It was really weird how it can be such a relief just to hear somebody else say, “My dad’s nuts!” and actually mean it and understand. (Adolescent; Harrison et al. 2014, p. 100)

Even for adult children, opportunities to share their experience of having a parent with a mental illness are rare yet valuable. As told by the 34-year-old son of a mother living with schizophrenia:

Where silence and speechlessness previously prevailed, something becomes ‘talkable’ and thus appears less threatening. It took me – like many other children of mentally ill parents – a very long time until I dared to tell my story. (Metz and Jungbauer 2019, p. 8)

Mindfulness vs Concealing Emotions: Concealing Emotions

In self-compassion, mindfulness is a non-judgemental state of mind which involves observing negative thoughts and feelings with openness, acceptance and balanced awareness, without suppression, denial or getting caught up in storylines of suffering (Neff 2003a, b). These concepts were discussed in 23 of the 27 studies.

Several children described engaging in mindful awareness of their difficult emotions through self-talk, journaling, listening to emotive music, talking to peers, family members, toys, pets or imaginary friends. Some spoke about acceptance as a way to cope.

You have to accept it before you can feel better about it. That’s what I had to learn to do. (Child/Adolescent; Mordoch and Hall 2008, p. 1138)

Conversely, some adult children’s parenting narratives reflected their unmet desire for mindfulness in their own childhood. Participants who were parents themselves, described endeavouring to allow their children to express their emotions and their individuality, an experience they did not have as children (Patrick et al. 2019). At the same time, others described how they were unable to parent in the way they hoped as their childhood left them emotionally numb.

A tendency to conceal emotions was pronounced. To illustrate, participants described strong emotions, involving dread, grief, anger and shame, combined with efforts to control and suppress their emotions.

Suddenly a huge amount of emotions, deep emotions. There are no words to describe the amount of fear and guilt. We only tried to cope somehow. In practice, it meant that everybody kept the emotions inside. (Pölkki et al. 2004, p. 158)

In a world that could be consistently unpredictable, children developed a “self-induced lithium” (76-year-old female; Carroll and Tuason 2015, p. 1069) so as to not ride the “roller coaster” of their parents’ tumultuous emotions (Adult female; Foster 2010, p. 3145). Some children assumed the role of the strong one for their family, looking for external guides for how to behave.

I try to be strong, but am often very sad, even though I don’t show it to the rest of the family… I notice that I keep a lot inside me. (Adolescent; Trondsen 2012, p. 180)

Emotional concealment was used as a means of coping with their experiences. Mordoch and Hall (2008) found that children reported hiding their distress to “better manage their situation”. However, authors noted that “children who appeared calm during participant observation often described feeling emotionally drained and distressed” (p. 1130).

Some children described considerable difficulty with observing and recognising their own emotions and their own needs, a key aspect of mindfulness. Moreover, children reported difficulty trusting their own emotions and wondering how they “should” be feeling. This was an experience which persisted into adulthood. In the words of this woman whose parent experienced schizophrenia:

I mean it’s probably more that I don’t know what I should be feeling...It’s not the feeling I’m feeling but I just need to be angry because I didn’t know how to express other feelings. (Adult female; Murphy et al. 2016, p. 671)

Akin to overidentification, children in 23 out of the 27 studies described concerns about inheriting their parents’ illness. Contrary to mindfulness, strong emotions were perceived by children as a warning sign or reminder of their genetic risk, further compounding their original emotions with worry and dread.

If I felt down or something or the other, then I was worrying, well am I going to be like that too, so...It sort of makes me wonder like sometimes if I haven’t got some of his [Pause], you know, um the affective disorder or something. If it’s not, cause I’ve heard like it’s heredity or something. (17-year-old female; Meadus and Johnson 2000, p. 387)

Self-kindness vs Self-judgement: Self-judgement

The third element of self-compassion is a desire to alleviate one’s own suffering with kindness. Self-kindness involves offering oneself warmth and understanding during times of suffering. Alternatively, one might respond to suffering from harsh self-judgement, criticism or blame. This theme of self-judgement was observed in 25 of the 27 studies in the review. Once again, whilst some children described instances of self-kindness, examples of self-judgement were more common.

Self-judgement was evident in children’s descriptions of feeling “worthless”. In some cases, self-worth was associated with unmet attachment needs or instances of abuse during childhood. For some young adults, rare moments of affection were viewed as a “validation of existence” (28-year-old female; McCormack and Sly 2013, p. 306). Worthlessness and self-hate were most often described in cases when parents were diagnosed with trauma-related or comorbid substance use disorders, such as this young man whose father suffered from post-traumatic stress disorder after serving in the Vietnam war.

It was always “you’re hopeless, you’re no good, you’re useless” … so I guess I’ve carried that all through my life. (Adult male; McCormack and Devine 2016, p. 282)

A sense of self-blame and guilt was reported by children, often associated with having caring responsibilities in relation to their parents’ mental health, an experience described in 23 of the 27 studies. Children reported feeling responsible to keep their parent well, lift their low mood, monitor medication adherence and prevent relapse or suicide, and many felt guilty if they were unable to do so.

I always had that fear that—um if anything happened to him while I was away that I hadn’t done enough to help him (crying) . . . I tried every way I could from every angle to stop him. (34-year-old female; McCormack and Sly 2013, p. 308)

Guilt and blame were felt by many, often younger children, who interpreted their parents’ symptoms as signs that they had done something wrong or misbehaved. This was reinforced when some were explicitly blamed by their parent or overheard discussions of post-natal illness onset. Nonetheless, adult children described how self-criticism, hypervigilance and determination drove them to achieve, though at a cost. This woman spoke with ambivalence about how anxiety, abuse and her guilt over perceived failure as a carer drove self-abuse, but also perseverance and achievement.

I was always mature, I was switched on, I can see that has been negative in a lot of ways, because I am so anxious and, that I can’t relax. But I do see it as a big positive too, that I’ve just never stopped so I have achieved lots. (Adult female; McCormack et al. 2016, p. 339)

Some children provided clear examples of how they employed self-kindness and mindfulness of emotions in efforts to independently manage difficult emotions and hard times.

I’m saying sorry to myself about when I was really sad. Like, “Sorry you’re sad”. It helps solves my problems. I just do it myself. (Child/Adolescent; Mordoch and Hall 2008, p. 1135)

For some adult children, redistributing responsibility for parental health, creating parent-child boundaries, engaging in self-inquiry and developing positive relationships had allowed them to repair their sense of self. In doing so, adult children “gave themselves permission to embrace…self-compassion” and “self-forgiveness” (McCormack and Devine 2016, p. 284). The therapeutic value of confronting self-blame and the role of social support in this effort are described by this teenager when reflecting on her school’s peer support group for children of parents with post-traumatic stress disorder.

…you get so low on yourself, sometimes thinking, “What did I do wrong? I did something wrong.” But then you talk to this other person, and they’re like, “We did nothing wrong. It’s them. They’re sick” and you have to realise that. (Adolescent female; Harrison et al. 2014, p. 100)

Discussion

This study sought to examine the experiences of children and adult children of parents with mental illness, in relation to self-compassion. Our review suggested there were no papers directly examining self-compassion amongst children of parents with mental illness. Nonetheless, aspects of self-compassion were observed in the representations of children’s experiences in qualitative research. Within the literature, children described examples of self-compassion, though instances denoting a lack of self-compassion (isolation or a lack of common humanity, concealing emotions or a lack of mindfulness and self-judgement as a lack of self-kindness) were common.

Included studies were drawn from a range of countries and adopted a range of data collection and analysis approaches with children of various ages, including adult children. In terms of the quality of the reviewed studies, researchers had frequently failed to discuss the potential influence of personal bias or relationship with participants on their findings. This may have resulted in subsequent analyses being represented as objective, rather than inherently subjective—which is the nature of good, transparent qualitative research (Newton et al. 2012).

Participants’ descriptions of stigma, silence, secrecy and shame provide the context in which children experienced isolation associated with experiences of parental mental illness. Participants tended not to report a sense of common humanity unless they were connected with siblings or a peer group with similar experiences. Isolation in the context of stigma aligns with research by Corrigan and Miller (2004), who described how children can be perceived as “contaminated” by a parent’s mental illness. We observed that if children internalise societal stigma, they may struggle to offer themselves kindness and compassion when they suffer in relation to a parent’s mental illness.

In this review, children reported a tendency toward concealing emotions. This finding may suggest that the difficult emotions felt by children in relation to a parent’s mental illness may not be acknowledged or addressed. Participants made sense of their tendency to conceal difficult emotions in various ways, with some suggesting that this persisted into adulthood. Within Neff’s (2003a, b) model of self-compassion, mindfulness is distinctly contrasted with overidentification, but not concealment or subjugation of emotions. Thus, Neff’s model may not adequately represent the experience of self-compassion for this population.

Descriptions of self-judgement, particularly self-blame, were frequently cited in the included studies and may indicate barriers to self-compassion. This finding extends the notion of Corrigan and Miller (2004), as it suggests that, like spouses and siblings, children experience blame for the mental illness of their loved one. Children who assumed responsibility for their parents’ mental health, reported experiencing a sense of self-blame and self-judgement when parent’s symptoms worsened or were unresolved. These children also described barriers to engaging in self-care and self-compassion, including when they described themselves as selfish for considering their own needs.

Limitations and Future Research

There are some notable limitations to this review. An a priori framework was adopted to identify content associated with self-compassion, and we did not focus on content that did not relate to self-compassion. This approach introduced the risk of confirmation bias. Therefore, we note that 13 of the articles which underwent full-text screening were excluded from the final review as they did not include content relevant to self-compassion. Consequently, these findings are drawn from a subset of the extant qualitative literature, in which self-compassion was presented, and this may not represent the experiences of all children of parents with mental illness. Additionally, the direct quotes included in qualitative papers are typically exemplar quotes, and only represent a small fraction of the collected data. Furthermore, as examining self-compassion has not been the primary aim of any qualitative literature to date, some self-compassion-related content may have been collected but not presented in the results of those studies. Additionally, it is important to note that we did not seek to determine whether self-compassion explains children’s outcomes. Even so, the findings of this review suggest that self-compassion may be relevant for some children of parents with mental illness.

Furthermore, our prior exposures to children of parents with mental illness—through life and/or work experiences—bring with them potential biases and preconceptions. However, as discussed in methodology, steps were taken to reduce the risk of bias directing the findings, whilst also drawing on existing knowledge to facilitate the abstraction process.

Notwithstanding the noted limitations, the findings of this review indicate directions for future research. Researchers might consider recruitment of more ethnically and culturally diverse populations and strengthen the consideration and reporting of concepts of trustworthiness in qualitative research, a noted limitation in some of the papers identified. Qualitative research is needed to explicitly examine the experiences and perspectives of children of parents with mental illness in relation to self-compassion, including the presence of ambivalence. Future quantitative research might examine the personal, relational and cultural antecedents of self-compassion in children of parents with mental illness, including whether self-compassion is implicated in relationships between childhood experiences and their outcomes. Furthermore, as children of parents with substance use disorders may be particularly at risk of childhood experiences associated with the development of self-criticism (Hinrichs et al. 2011), future research might also examine self-compassion amongst their population. Finally, the themes identified in this study may provide relevant targets for interventions to address the mental health and well-being outcomes for children of parents with mental illness. The emotional experiences reported by participants, including isolation, self-criticism, shame and emotional concealment, may highlight specific challenges faced by this group which need to be addressed. Thus, future research is needed to determine the clinical utility of self-compassion embedded into interventions for children and adult children of parents with mental illness and evaluate whether self-compassion interventions designed for these children and their families might mitigate their significant hardships.