Compassion-informed approaches for coping with hearing voices: literature review and narrative synthesis

ABSTRACT

Background

Compassion-informed talking therapies have gained increased attention with people who hear voices and practitioners alike. Developing inner kindness can reduce powerful critical voices and self-stigma, thus increasing people’s ability to cope with challenging voice hearing experiences, and potentially increase overall wellbeing.

Methods

The current review aimed to explore how compassion-informed approaches to coping with voice hearing are understood and reported across the existent voice-hearing literature. Academic Search Ultimate, MedLine and PsychINFO databases were searched for suitable papers, using the terms: [“hear* voice*” OR “voice hear*” OR “auditory hallucinat*”] AND compassion*.

Results

Fourteen papers were identified for inclusion: six quantitative studies, six qualitative reports, and two theoretical reviews. Included studies host a total sample size of 855 people, representative of clinical and community populations, across adolescence and adulthood. The reviewed research originated from the United Kingdom, Australia, France, Germany, Italy, Ireland, and the United States of America.

Discussion

Self-compassion and building a compassionate alliance with one’s voices can be challenging for many voice hearers to initially embrace. Over time, building a mutually compassionate relationship with one’s voices can help empower the voice hearer by resolving, rather than increasing, inner conflicts, and may increase opportunities to experience the presence of agreeable voices.

KEYWORDS:

• Hearing voices
• compassion-informed approaches
• coping
• review
• narrative

Voice content is commonly discussed in the context of people’s life experiences (Corstens et al., 2012; Maijer et al., 2019) and socio-cultural milieu (Larøi et al., 2019); with voices often providing insights into the hearer’s feelings about themselves (Gilbert, 2009) and their relationships (Heriot-Maitland et al., 2019). Although it can be distressing to hear voices that others do not due to voice tone, content, and culturally associated stigma, some find their voices to be helpful or comforting (e.g. Dudley et al., 2018; Parry & Varese, 2020).

Within mental health services, 67%–74% of voice-hearers report a trauma history (Maisey et al., 2021) and past trauma has been associated with an increased likelihood of experiencing “hallucinations” (Varese et al., 2012). In the context of trauma and severe stress, voices may represent a need for psychological protection (Corstens et al., 2012; Heriot-Maitland et al., 2019). There is a need to promote compassionate coping with the presence of voices that seeks to explore their potential function.

Compassion, self-kindness, and voice hearing

Compassion has been described as “a sensitivity to the suffering of self and others, with a deep commitment to try to relieve it” (Gilbert, 2010, p. 3). By this definition, compassion depends both on showing understanding or empathy for one’s own or someone else’s distress, and on committed action for change (Norman et al., 2020; see; Neff, 2003). Similar definitions of compassion appear to be broadly agreed upon in empirical studies of compassion across a wide range of disciplines (Mascaro et al., 2020).

Higher levels of self-compassion have been associated with helpful strategies for coping (Ewert et al., 2021). Further, self-compassion can protect against the impact and internalisation of stigma (Wong et al., 2019), making the concept and practice of self-compassion particularly relevant for people who hear voices.

Compassion-focused therapy (CFT) is a specific therapeutic approach, aiming to develop experiences of inner kindness and self-soothing, particularly for clients with high levels of shame and self-criticism (Gilbert, 2009). CFT and its briefer form, compassionate mind training (CMT), employ specific strategies that often connect to people’s senses, such as compassionate imagery or mindful attention to sense motivating stimuli. Parry et al. (2021) tentatively associated the multisensory coping strategies described by young people who hear voices with CMT techniques, offering an additional clue that compassion-informed approaches could be particularly beneficial for people who hear voices.

Compassion is an important feature of talking therapies that transcends modalities, however. Therefore this review aimed to bring together research in which these recognised elements of compassion could be identified; an awareness of own or other’s suffering, an attitude of kindness or sensitivity towards this, and a motivation to reduce suffering (Mascaro et al., 2020).

For voice-hearers, self-to-self or self-to-voice compassion may involve understanding their voices in the context of lived experience. It may involve taking a non-judgemental and non-shaming attitude towards oneself as a voice-hearer. Self-to-voice compassion would involve mindfully acknowledging the voice, rather than seeking to avoid or eliminate it, and responding with kindness rather than fear (Maisey et al., 2021). For example, hearers who experience self-to-voice compassion would perceive their voices as serving a purpose, rather than as solely intimidating.

A “compassion-focused approach” was therefore defined in this review as any intervention or coping strategy that aims to increase voice-hearer’s compassionate feelings or behaviours towards themselves or their voices. The review aimed to develop a conceptual understanding of how such approaches could be helpful in sense making, developing coping strategies, and for managing one’s individual experiences alongside the sociocultural stigma (Degnan et al., 2021) associated with voice hearing across much of the Global North.

Method

Information sources

Sources searched were Academic Search Ultimate, MedLine and PsychInfo. Reference lists of the selected papers were also cross-checked to ensure the search had not missed potential papers for inclusion. Eligibility was assessed through title and abstract screening, followed by full-text screening of the shortlisted studies. Peer reviewed and published manuscripts were selected for review if they explicitly related to voice-hearing and/or auditory hallucinations, referenced “compassion” within the abstract or body of the paper, included adult and/or child populations, and if the full text was available in English. Due to this review’s specific focus on voice-hearing, studies were excluded if they related to “psychosis” or “hallucinations” without making it clear whether this included hearing voices.

Search strategy

The following search terms were employed: [“hear* voice*” OR “voice hear*” OR “auditory hallucinat*”] AND compassion*. The search returned 19 results, of which 14 met the inclusion criteria, relevant to both compassion and voice hearing (please see Figure 1).

Figure 1. Flowchart of inclusion/exclusion.

Screening

Inclusion and exclusion criteria were agreed by the first and third author. The first author conducted searches, screening and analysis, under supervision by the other two authors. Any emerging issues with the review were resolved collectively.

As this was a narrative review rather than a systematic literature review (Moons et al., 2021), a formal quality appraisal was not followed. Papers were selected based on inclusion criteria rather than on quality ratings.

Data items

Fourteen papers were selected for review; six quantitative (Bortolon & Raffard, 2019; Carden et al., 2020; Dudley et al., 2018; Maisey et al., 2021; Norman et al., 2020; Rosen et al., 2018b), six qualitative studies (Cheli et al., 2021; Hayward et al., 2018; Heriot-Maitland & Levey, 2021; Mayer et al., 2022; Mayhew & Gilbert, 2008), one case study review (McCarthy-Jones & Longden, 2015) and one theoretical paper on compassion focused approaches for voices (Heriot-Maitland et al., 2019; please see Table One). Three studies specifically focused on CFT (Heriot-Maitland & Levey, 2021; Heriot-Maitland et al., 2019; Mayhew & Gilbert, 2008). Three studies operationalised self-compassion as a variable for correlational analysis (Dudley et al., 2018; Maisey et al., 2021; Norman et al., 2020), and one gave a qualitative analysis of self-compassion amongst voice hearers (Cheli et al., 2021). A further four recommended compassion-focused therapies in the light of their research findings (Bortolon & Raffard, 2019; Carden et al., 2020; McCarthy-Jones & Longden, 2015; Rosen et al., 2018b), for example when the study focused on shame or trauma. One study (Hazell et al., 2017) mentioned compassion only in the context of a compassionate therapeutic relationship but contained interesting qualitative insights from participants, so was relevant for inclusion. Another discussed compassionate relational therapeutic approaches, such as role play and chair work, which are also techniques found in CFT (Hayward et al., 2018), so it was included. Finally, one qualitative paper related specifically to young people (Mayer et al., 2022).

Diagnoses disclosed by some participants included schizophrenia spectrum disorders (Bortolon & Raffard, 2019; Hazell et al., 2017; Maisey et al., 2021; Mayhew & Gilbert, 2008; Norman et al., 2020; Rosen et al., 2018b); bipolar disorder (Bortolon & Raffard, 2019; Rosen et al., 2018b); depression (Bortolon & Raffard, 2019; Hazell et al., 2017); anxiety, eating disorder, obsessive compulsive disorder (Bortolon & Raffard, 2019); brief psychotic disorder (Cheli et al., 2021); post-traumatic stress disorder (Maisey et al., 2021); borderline personality disorder and dissociative identity disorder (Hazell et al., 2017).

Analytic approach to qualitative synthesis

In addition to the overall synthesis, a narrative synthesis was undertaken with the qualitative studies reviewed to generate novel emancipatory narratives of compassion with voice hearing (Beaumont & Hollins Martin, 2015). Qualitative research allows for insight into under-represented first-hand perspectives (Marriott et al., 2014), highly relevant for the current review, which interpreted therapeutic engagements from a compassion-informed perspective. Six qualitative studies were included, exploring the experiences of 47 voice-hearers within and outside of mental health services. This review aimed to create a cohesive, nuanced narrative (Lisy & Porritt, 2016), reconsidering the diverse stories shared by participants and original interpretations of the studies’ authors. The concept of compassion was used as a lens for the synthesis and theoretical framework for understanding the emerging stories. This approach identified elements of compassion that were explicitly and implicitly narrated in the original papers, offering a novel interpretation of compassion for people who hear voices.

Results

Quantitative synthesis

Through correlational analysis, the quantitative studies explored who is likely to hear distressing voices and what psychological factors influence this distress. Most discussed shame (Bortolon & Raffard, 2019; Carden et al., 2020) and self-compassion (Dudley et al., 2018; Maisey et al., 2021; Norman et al., 2020) as potential factors for reducing both the likelihood of hearing voices and the distress associated with it. Half (Bortolon & Raffard, 2019; Carden et al., 2020; Rosen et al., 2018b) discussed the psychological benefits of improving relationships between hearers and voices, potentially whilst also addressing beliefs about voices (Norman et al., 2020).

Voice-related variables measured were: “hallucination-proneness”, “psychotic-like experiences”, mindfulness of voices, belief about voices, negative voice-content, and voice-related distress. These variables were studied in relation to broader psychological concepts: childhood adversity or childhood trauma, post-traumatic stress, attachment style, shame, self-compassion, self-esteem and perceived social rank.

Voice-content

Rosen et al. (2018b) operationalised negative voice content and voice-related distress as separate variables, along with adverse childhood experiences. Childhood adversity was associated with negative voice content; however, negative voice content was found to mediate the relationship between childhood adversity and distress. Therefore, more positive voice-content was associated with lower levels of voice-related distress, regardless of childhood adversity.

Rosen et al. argue that talking therapy interventions need to change the actual content of voices from negative communications to more positive ones, rather than encouraging the hearer to think differently about the voice. If the hearer communicates to their voice that they are not interested in or affected by what the voice says, this may increase negative voice content. The authors propose that compassion-informed approaches focused upon improving relationships between hearer and voice may contribute to important change.

Shame

Bortolon and Raffard’s (2019) study with 175 participants from the general population showed a significant relationship between shame and the likelihood of hearing voices. Shame mediated the relationship between childhood trauma and “hallucination-proneness”. Whether or not a survivor of childhood trauma developed voice-hearing was influenced by how much shame they felt, highlighting an urgent need to address shame for voice-hearers. Similarly, Carden et al. (2020) found higher levels of shame were strongly correlated with negative voice content, and negative relationships between hearer and voice. They argue that negative voices mirror the hearer’s perception of themselves as essentially “lesser than”. Interestingly, levels of shame were not correlated with the presence of positive voices, suggesting that positive voice qualities can be experienced even by participants with high levels of shame.

Self-compassion

Norman et al. (2020) explored the role of self-compassion and other constructs used in cognitive behavioural models of voice-hearing, namely self-esteem, perceived social rank, and beliefs about voices, in predicting the level of voice-related distress. They found levels of self-compassion, rather than self-esteem ratings, strongly predicted negative voices and voice-distress. The two are distinct, as self-compassion is based in recognition of universal humanity, while self-esteem is conditionally based on favourable social comparisons and social rank (Neff, 2003), highlighting the important influence of one’s self-perception in and of their social milieu.

Voice malevolence mediated the relationships between self-compassion and levels of voice-related distress. This may support Rosen et al.’s (2018a) emphasis on the content of voices and their perceived relationship with the hearer.

Scheunemann et al. (2019) investigated correlations between self-compassion, the frequency of “psychotic-like experiences” (PLE), and PLE-related distress in a community sample of 234. Higher self-compassion was associated with lower frequency of, and lower distress related to, PLEs, including voices. The aspect of compassion most strongly correlated with lower distress was “common humanity”. Participants with a stronger sense of common humanity, and shared recognition of others’ successes and struggles, are less likely to perceive themselves as abnormal or alone.

Similarly, Dudley et al. (2018) found that voice hearers reported significantly lower distress if they reported greater self-compassion and mindful awareness of voices. A synergistic effect was seen, whereby the benefits of self-compassion were greater for participants who also scored highly for mindfulness. Therefore, mindful awareness of voices could enhance the flow of compassion. Dudley et al also explored the role of people’s attachment style. The authors hypothesised that people from “low affection or abusive backgrounds” are more likely to have fears of self-compassion, and fearful attachment styles. A significant correlation was found between fearful attachment and low self-compassion. However, no correlation was found between insecure attachment and voice-related distress. Perhaps this echoes Bortolon and Raffard’s (2019) finding that voice-related distress was more linked to shame than to adverse childhood experiences. Additionally, Dudley’s research suggests an important role for the therapeutic relationship in terms of nurturing safety in the relationship to reduce a fear of compassion.

Maisey et al. (2021) examined whether self-compassion moderated the relationship between post-traumatic stress (PTS) and voice-related distress, finding self-compassion was negatively correlated with both PTS symptoms and voice-related distress. However, greater self-compassion was associated with lower distress, regardless of PTS threshold, highlighting the benefit of nurturing self-compassion for someone’s wellbeing.

The synthesis of the aforementioned studies suggest that it is self-compassion and shame, rather than trauma or insecure attachments, which specifically determine voice-related distress. Therefore, it could be hypothesised that shame-reducing and compassion-increasing interventions could benefit voice-hearers and hold therapeutic promise, even for people who have experienced significant trauma.

These studies’ authors called for further quantitative research to be carried out with longitudinal (Carden, Dudley, Maisey, Norman) or experimental (Bortolon, Maisey, Norman) designs, as none of these correlational studies could establish the direction of causality between distressing voice-hearing and the other variables (shame, self-compassion, mindfulness and attachment) studied. Authors noted the importance of controlling for potential confounds such as anxiety and depression (Bortolon, Dudley), or wider social factors including poverty or ongoing victimization (Rosen). It was recommended that RCTs be carried out to establish if CFT or CMT interventions do in fact reduce distressing voice-hearing (Maisey, Norman, Scheunemann). Furthermore, some authors (Maisey, Norman) called for their study designs to be replicated with non-clinical populations, to explore whether higher self-compassion is a protective factor against requiring services.

Narrative synthesis

The qualitative and theoretical papers were synthesised through a process of familiarisation, writing narrative summaries of each paper, and constructing emancipatory narratives from across the publications. Analysis was conducted of the entire results and discussion sections, including quotes from participants and interpretations from original authors. This textual data was synthesised to find common themes and differences across papers, to begin telling a coherent “story” based on the literature to date. Interpretations and synthesis were developed in discussions between the three authors. This process resulted in two narrative layers that tell an overall story of voice hearing through a lens of compassion-informed therapeutic approaches.

Cooperation between hearer and voice

From participants’ and researchers’ narratives, it was clear that developing a mutually co-operative relationship with voices is challenging, and an unfamiliar concept to many voice-hearers. Cheli et al. (2021) interviewed two participant groups: six participants diagnosed with Brief Psychotic Disorder and six without diagnosis. Differences emerged in how these two groups described the flow of compassion between themselves and their voices. Participants without a diagnosis welcomed their voices and perceived them as nurturing. By contrast, participants with a diagnosis described being spoken to, or at, by the voices, with little or no reciprocal dialogue. The content of voices was aversive or abusive. The authors comment that “the voices seemed to want to underline a lack or deficit in the hearers, and a consequent need to change … their intention was never caring or compassionate” (p. 8).

Yet if compassion consists of a recognition of suffering, plus the commitment to change, the voice’s intentions may indeed be positive. Within the CFT model, it has been observed that many people attempt to help themselves by telling themselves to “get a move on” or “pull yourself together”, for example, yet inadvertently do this in a way that feels critical and judging. Clients are trained to use supportive tones as well as content, when speaking to themselves (Gilbert, 2010). In Hayward et al.’s (2018) Relating Therapy, role plays were used to change the relationship between hearer and voice. Participants felt they became more assertive with their voices and described benefits of being able to challenge them. Participants felt less controlled and overwhelmed as a result. However, these accounts did not suggest that participants saw compassion for their voices as a priority. Interestingly, although some voices responded to the role plays by becoming kinder, others developed more unpleasant content. Hearers felt this was because the voices felt threatened by the therapy. The same difficulty was highlighted by Hazell et al. (2017). Hearers found that their voices often obstructed their attempts to get help. This was apparently due to the voices’ perception that they were being talked about, rather than to, and that the hearer and therapist wanted to eliminate them. This potential threat to the voice(s) increased the likelihood of disengagement from therapy. Participants spoke about the value of a compassionate therapist, but to develop a compassionate self was more challenging, and compassion for voices was absent from their narratives.

By contrast, Heriot-Maitland and Levey (2021) described a process of the client beginning to offer gratitude and reassurance to a voice she had previously perceived as unhelpful. Another voice responded to this by offering soothing words to the client, suggesting a multi-directional “flow” of compassion.

This review uncovered one study with children and adolescents, with a small sample of six (Mayer et al., 2022). Some of these young voice-hearers identified voices as having positive roles, such as reducing loneliness or helping keep the young person safe in threatening social situations. However, all hearers had felt controlled, or “bullied”, by their voices. Their narratives showed ambivalence towards the idea of voices being part of themselves, especially when the voices were threatening, or encouraged actions that misaligned with the client’s own values. Young people spoke of not needing to understand the purpose or causative factors behind their voices in order to lead fulfilling lives; this contrasts with adult literature, which suggests that voice-hearers benefit from “sense-making” (Corstens et al., 2012; Steel et al., 2019). For them, an indication that they were coping better would be that they could live their lives despite the voices. From these accounts, it is unclear what sense young people may make of an intervention that directly encouraged them to form an allegiance or engage compassionately with their voices. However, living well with the voices seemed more important than only understanding the presence of the voices.

In summary, it appears that the notion of mutual co-operation between hearer and voice is largely absent from traditional therapies or mental health interventions, and from many voice-hearers own responses to their voices. Nurturing this sense of co-operation may therefore be a valuable step forward.

Negative voices as social coping strategies

Many voice hearers and clinicians struggle to see a positive intention behind voice hearing. This is understandable given how often voices are found to be critical, bullying, shaming or threatening. However, compassion-informed approaches interpret shame and self-criticism as survival strategies, arising out of necessity in stressful conditions (Heriot-Maitland et al., 2019). In our social world, rejection by others is experienced as a genuine threat to the safety and psychological wellbeing of the individual (Gilbert, 2009).

Theoretical literature has explored the evolutionary function of self-shaming and submission, and its relevance to voice-hearers. McCarthy-Jones (2017) argues that shame is an adaptive strategy used by people who find themselves in a subordinate position, to protect themselves from harm from a more powerful other. This can be likened to animals adopting submissive behaviours to avoid another’s aggression (Heriot-Maitland et al., 2019). Acts of submission have been described as internalised shame (Gilbert, 2009, p. 84). If shame serves to protect us, this adaptation may have been necessary in the lives of many voice-hearers. McCarthy-Jones (2017) interpretation of historical case documents suggests that shame plays a causal role in voice-hearing and may be a confounding factor in the relationship between trauma and voices.

Mayhew and Gilbert’s (2008) participants described how their distressing voices served the purpose of keeping the hearer “in check”, protecting them against external judgement or persecution. However, after their experiences of CMT, participants appeared to make less use of submission as a safety strategy. Similar to Hayward et al. (2018) Relating Therapy clients, they felt less subordinate to their voices. Mayhew and Gilbert’s participants also described a direct link between their self-thoughts and the quality of their voices; a change in their self-talk brought about a corresponding change in their voices’ content, following which voices appeared to show more positive intentions as supportive and reassuring, rather than critical and intimidating.

An important aspect of compassion-informed therapies then, may be to help the voice-hearer make sense of their voices’ intention to keep the hearer safe in the social world by attempting to avoid rejection by others. Voice-hearers can be encouraged to develop more reassuring self-talk in response to these fears of rejection or social shaming.

Discussion

This review has shown that developing self-compassion is a struggle for many voice hearers, particularly in clinical populations. Many voice-hearers have described either an antagonistic relationship with their voices (Cheli et al., 2021) or finding a compromise, whereby they are able to resist, stand up to, or go on with life despite their voices (Hayward et al., 2018; Mayer et al., 2022).

Some hearers reported that going to therapy, or attempting to ignore their voices, caused the voices to become more unpleasant or obstructive (Hazell et al., 2017). Relatively few participants described positive allegiances with their voices, or a mutual flow of compassion (self-to-other (voice) and (voice) other-to-self). Compassion-informed approaches may therefore play an important role in empowering the voice hearer, whilst ensuring that the voice is not antagonised by the process (Heriot-Maitland & Levey, 2021). Voices can be conceptualised as part of the self, and in CFT it is considered vital that the different parts do not judge or seek to suppress one another (Heriot-Maitland et al., 2019). CFT places great emphasis on reducing shame, an emotion that has been identified as playing a causative role in voice-hearing (Bortolon & Raffard, 2019; McCarthy-Jones, 2017). Related to this is the experience of subordinate patterns of relating to others, which can be reflected in the hearer’s relationship with their voices if the voice becomes seen as an “other” (Hayward et al., 2018).

Shame was discussed across all papers reviewed. Mayhew and Gilbert’s (2008) participants described the voices keeping tabs on some shameful aspect of the voice hearer. Heriot-Maitland and Levey (2021) describe overcoming the shame of being thought “mad”. Mayer et al. (2022) young participants also spoke about the shame of voice hearing. Whilst McCarthy-Jones (2017) suggests that shame causes voices, it is important to note that for many people, being a voice-hearer also causes shame. As Carden et al. (2020) suggest, shame can be caused by the pathologising effect of using services or being defined by a diagnostic label, which may go some way towards explaining the differences in how Cheli et al.’s (2021) diagnosed and non-diagnosed participants perceived their voices.

Compassion-informed approaches emphasise that painful experiences are part of universal human experience (Gilbert, 2009; Neff, 2003) and that many of our unwanted feelings, behaviours and experiences started out as necessary and effective safety strategies (Heriot-Maitland et al., 2019). As such, these approaches are de-pathologising and may contribute well to reducing the stigma and isolation to which voice hearers can be vulnerable (Iudici et al., 2019).

For young people, normalisation, the support of family, and being seen as one’s own “self”, rather than being defined by one’s voices, seemed to help combat this perceived stigma (Mayer et al., 2022). There is still relatively little research into the voice-hearing experiences of children and adolescents, despite voice-hearing being more common before adulthood (Maijer et al., 2018). Daalman et al. (2011) found that participants with an earlier age of voice-onset were significantly less likely to need clinical care than those whose voices began at a later age, suggesting that very young voice-hearers may possess a novel perspective that is as yet unexplored but likely to be less influenced by the stigma held within many western cultures in relation to hearing voices. A particularly interesting point made by Maijer et al. (2019) is that although voice-hearing often occurs at times of stress or trauma, the voices predictably discontinue when the traumatic circumstances change. Consequently, the authors suggest that appropriate interventions for voices would include universal interventions, such as anti-bullying programmes in schools. A profound shift in the conceptualisation of voice-hearing would be needed if similar principles and systemic social interventions were to be applied in the case of adult voice-hearers.

Strengths, limitations and future research

A strength of this review is its inclusion of a range of quantitative and qualitative methodologies; facilitating a novel synthesis of the varied ways in which compassion appears in the literature on voice-hearing. Theoretical literature has proposed that developing self-compassion and reducing shame are helpful in interventions for voice-hearing. Quantitative studies into the relationships between shame, self-compassion and voice-related distress appear to support this. This review therefore demonstrates that compassion is a robust and worthwhile concept to explore in relation to voice-hearing and voice-related distress.

A limitation of the quantitative studies is that most used a cross-sectional design, which identified correlations between factors such as shame, self-compassion and voice related distress, but could not infer causation. One way to improve this would be by conducting longitudinal research, exploring how these different factors develop chronologically. Another approach might be to focus more specifically on childhood and adolescence, when voice-hearing has been shown to be particularly dynamic and often transient. Through qualitative accounts, young people may be able to elucidate how internal and external events in their lives have influenced the development and the quality of voices. Interventions to address triggers and maintaining factors could then be more targeted and effective.

This review purposefully included studies pertaining to both clinical and non-clinical populations. This allowed for a broad view of the “spectrum” of voice-hearing, ensuring the review was not limited by assumptions that voice-hearers necessarily require mental health services. This could also be a limitation, as it is not certain whether the findings from a clinical study would apply to a non-clinical population, or vice versa. As the research in the current review was conducted in the Global North where voice hearing is largely understood as indicative of mental health difficulties, it would be helpful to learn from conceptualisations of voice hearing across the world to develop a better understanding of cross-cultural coping strategies from diverse sociocultural and holistic perspectives, further extrapolating voice hearing as an experience from western constructs of mental health or illness.

Finally, this review including both adult and youth literature has allowed for breadth but not depth of understanding of hearing voices across lifespan. The review has begun to explore how voices are experienced differently by age groups and what may be helpful at different developmental stages, indicating children and adolescents have valuable perspectives to share on factors that increase vulnerability to, or protect against, distressing voices. However, further research is certainly needed to gain more comprehensive insights into the relationships that young people have with voices and how support options may need to differ across developmental ages and stages.

Conclusion

In summary, this review has demonstrated that compassion-informed approaches for coping with hearing voices include relational work, appreciating the purpose or role of voices, and developing a flow of compassion to facilitate new options for relating to and communicating between the self and voices. These approaches are particularly helpful for individuals with high experiences of shame. Based on the findings of this review, it could be hypothesised that increasing the flow of compassion between voice hearers and their voices could improve the relationship, reduce shame, voice-related distress and self-stigma, improving people’s wellbeing.

Disclosure statement

No potential conflict of interest was reported by the author(s).