A multi-sectoral evidence synthesis on religious beliefs, intimate partner violence and faith-based interventions


 In recent decades, more studies have emerged that examine the relationship between religion / spirituality, marital functioning and mental health. The scholarship on domestic violence (DV) / intimate partner violence (IPV) and religion does not appear to have integrated this evidence sufficiently and also lacks a multi-sectoral perspective. Ultimately, evidence from psychology and counselling needs to be bridged with evidence from public health and international development programming and studies in anthropology and sociology. A better integration could help to increase understanding of the multidimensional effects of religious parameters in the experience of DV / IPV and to identify how these effects could be leveraged resourcefully in faith-based interventions at community level and in psychosocial counselling involving victims, survivors and perpetrators in religious contexts. The current synthesis is the result of a systematic review that was conducted to start to bridge this multi-sectoral and multidisciplinary evidence to identify specifically: a) the influence of religious beliefs and experience on the rationalisations, behaviour and mental health of victims, survivors and perpetrators, and b) faith-based interventions in community and psychological counselling and their effectiveness. In pursuing these questions, we aimed to overcome the dominance of western definitions and paradigms of domestic violence or religion by combining evidence from sectors that are more international, such as public health and international development, and studies conducted in low-and middle-income societies that received less attention historically. We especially sought to distance ourselves from rigid understandings of ‘religion’ in an effort to make the review more relevant to diverse religious traditions and experiences.


Introduction
In recent decades, studies have paid considerable attention to the relationship between 'religion', individual mental health and intimate relationships (Bahnaru et  The scholarship on religion and domestic violence (DV) / intimate partner violence (IPV) does not appear to have integrated fully the evidence on religion, marital functioning and mental health. The literature on DV / IPV and religion also lacks a multi-sectoral perspective, with psychology and counselling insights still needing to be bridged with evidence from programmes in public health and international development and more anthropological and sociological study results. A better integration could help to identify with more speci city the multidimensional effects of religious parameters in the experience of DV / IPV and to identify how these effects could be leveraged resourcefully in faith-based approaches to DV / IPV at community level and in psychosocial support provision services involving victims, survivors and perpetrators of DV / IPV in religious contexts. It would also help to identify with more precision what works and what does not in faith-based interventions involving clergy, communities and counsellors.
While research on faith-based interventions has expanded in recent decades, studies that identify the causal mechanisms between religious parameters in DV / IPV interventions and their effectiveness are noticeably few.
The current synthesis is the result of a systematic review that was conducted to start to bridge this multisectoral and multidisciplinary evidence in order to increase understanding around the in uence of religious beliefs and religious experience on the rationalisations, behaviour and mental health of victims and perpetrators of domestic violence and to identify how victims, survivors and perpetrators in religious contexts can be supported better through effective and culturally adapted faith-based interventions. For the purposes of this review, DV / IPV was understood as violence affecting intimate romantic partners and could include all forms of abuses identi ed by study participants.
The review was led by the following two speci c questions: 1. How do religious beliefs and religious experience inform the individual conscience, mental state and behaviour of victims, survivors and perpetrators of intimate partner violence? 2. How effective has it been to integrate religious beliefs and religious experience in intimate partner violence victim, survivor and perpetrator support programmes, counselling and clergy-centred interventions and how can their effectiveness be increased?
We pursued these two questions with the recognition that the dominant evidence on DV / IPV, religion and mental health has historically emanated from western societies. We thus sought to diversify this literature by combining evidence from other sectors, such as public health and international development and studies from low-and middle-income societies (LMICs) in particular that received less attention historically. The main objective of the paper is to provide a synthesis of evidence, to the best of our ability, that can inform how DV/ IPV programmes and approaches can be better adapted to diverse religious communities, contexts and individuals.

Theoretical premises and methodological challenges
Historically, de nitions of 'religion', 'religiosity' or 'spirituality' have remained Eurocentric, re ecting conceptualisations that do not eschew the legacies of Enlightenment experience with theological dogmatism under western Christianity and 19th century secularisation processes. These legacies cannot be elaborated here, but it su ces to note that they have resulted in epistemologies of religion-ways of understanding and studying religious traditions-that have not always been shared by non-western religious societies (Istratii, 2020: chapter 2). Similar Eurocentric tendencies can be identi ed in the public health and international development sectors in LMICs and other non-western societies whereby DV / IPV has been approached through singlefactor reductionist theories, neglecting important psychological parameters and failing to ground the analysis of the problem in a substantive, ethnographic and people-centred engagement with local worldviews that can indicate relationships between gender, cultural, psychological and other parameters with religious beliefs (Istratii, 2020). While within industrialised societies, attention has been given to the psychological parameters of DV / IPV as much as to gender inequalities, drawing attention to intergenerational violence, psychological trauma and individual personality disorders, within the context of LMIC foreign-led research and practice, the assumption has been that DV / IPV is the outcome of gender asymmetries or culture-condoned pernicious norms and practices (Bowman, 2003;Jakobsen, 2014;Istratii, 2020). This has fostered tendencies to appraise cultural or institutional parameters as loci of female subordination contributing to women's abuse and to focus exclusively or disproportionately on gender-related societal approaches to address the problem, neglecting the in uence and potential resourcefulness of religio-cultural parameters in help-seeking attitudes and societal norms contributing to the problem and its mitigation (Narayan, 1977;Volpp, 2005;Istratii, 2020).
Owing to the differences in conceptual and methodological approaches, bridging different scholarships, disciplines and theoretical frameworks together can be challenging. For instance, numerous studies examining the relationship between religious worldviews and DV / IPV are qualitative (anthropological, phenomenological, etc.), while studies in psychology or public health tend to be quantitative (crosssectional, statistical, diagnostic, etc.). Anthropological or sociological studies often employ multidimensional, nuanced or exible conceptualisations of 'religion' compared to rigid de nitions encountered in psychology studies. On the other hand, international development and public health have shown tendencies to dissect 'religion' into beliefs and practices, to separate 'religion' from 'culture', place it at odds to reason, or bound it in the private sphere, which continue to re ect Eurocentric conceptualisations and experiences mentioned earlier. Such conceptualisations can be unhelpful in reallife contexts, especially in those LMIC contexts where religious traditions are highly salient in social life and are experienced as inextricable from social norms, in uencing collective understandings, as well as individual identities and life histories.
The review sought to distance itself from rigid understandings of 'religion' or theories of domestic violence in an effort to become more relevant to diverse religious societies and traditions, especially those found in LMIC contexts. To achieve this, the systematic review tried to capture the in uence of beliefs, values or practices that participants related to their religious tradition, religious upbringing or faith experience in their particular socio-cultural contexts.

Methods
To assess the available evidence in order to answer the two review questions an Arksey and O'Malley's (2005) scoping review methodology was employed. This approach is useful for determining the state of evidence on an issue which requires further clari cation before rigorous empirical studies are developed and conducted (Constand et al., 2014). It is a ve staged process involving de ning the research question, identifying relevant studies, study selection, charting the data and collating, summarising and reporting the results (Arksey and O'Malley, 2005).

Identifying relevant studies
A literature search using the search engines MEDLINE, PubMed, Web of Science, APA PsycINFO, Scopus, CINAHL, International Bibliography of the Social Sciences (IBSS) and Atla Religion Database was performed. Keywords used in the search included domestic violence, intimate partner violence, spousal violence, wife abuse, religion, domestic violence AND religion, OR religion AND intimate relationships AND psychology, religion and intimate partner violence. Use of Boolean operators enabled exploration of a combination of terms. A search was also conducted using Google and Google Scholar to identify studies not published in indexed journals. In addition, the reference list of each article was scrutinised to identify studies that may not have been listed in the searched databases.
In this review we looked rst for studies that explored the relationship between religion and DV / IPV from the victim's or perpetrator's perspective and secondly for studies that presented on faith-based interventions with DV / IPV victims and perpetrators involving counsellors, clergy and communities. The review was open to studies from any context in the world, with a focus on non-western societies, religious minority communities in secular contexts or communities where religious traditions or faiths have been known to be prevalent in society, as encountered in many LMIC contexts. The studies read as a result of this review were combined with other relevant studies from anthropology, international development, gender and religious studies, psychology and other disciplines that the authors were aware of to achieve a more comprehensive synthesis of the evidence. Included studies had to be based on empirical data, written in English, and published in a peer reviewed journal between 2000-2020. Papers such as reports, case series, editorials, commentaries were excluded. PhD theses were included only if they presented an empirical study. We also excluded studies on spirituality that did not make explicit reference to participants' religious beliefs and experiences. Within psychological and counselling studies, 'religion' and 'spirituality' are sometimes used interchangeably or together, with spirituality being typically conceptualised in more existential ways, and religiosity or religious belief being conceptualised as a subset of spirituality. While spiritual-existential concerns and understandings can in uence the conscience, mental health and behaviour of individuals in intimate relationships with important implications for victims, survivors and perpetrators of DV / IPV, such experiences extended beyond the scope of this systematic review and were excluded.

Selection of Studies
The initial search resulted in the identi cation of 2,381 potentially relevant studies. Removal of duplicates resulted in reducing the number to 1,722. Initial screening of the articles resulted in the further removal of 1,495. A total of 227 articles were retrieved (130 under question 1 and 97 under question 2), and review of title and abstracts of each study resulted in removal of another 158 irrelevant studies, with 28 studies under question 1 and 41 studies under question 2 remaining. The full text for these studies was retrieved and both authors read these articles to determine inclusion as shown in PRISMA chart in Figure 1. An additional 11 studies were included on the themes of religion, domestic violence from other disciplines that the authors were aware of but did not show on the search. Titles and abstracts of articles were independently reviewed by the authors. If articles were representative of the inclusion criteria, the articles went through two full-text independent reviews by the authors.

Charting data
A data extraction template was constructed and used to record relevant information such as purpose, research design, sampling method, sample characteristics, data collection method, method of data analysis, the results of the study, limitations and comments. Data were systematically charted in Microsoft Excel.

Collating, summarising and reporting results
The ndings of the review are presented under appropriate headings in the following section. The data was extracted differently depending on whether it was extracted from qualitative or quantitative studies. The studies were grouped thematically and summarised with emphasis being placed on study objectives, context, key ndings and limitations.

Results
The reviewed studies are presented in two sections, which re ect the two questions driving the review. The rst section presents the evidence on the in uence of religious beliefs on victim, survivor and perpetrator psychology and behaviour, placing attention to speci c mechanisms that relate faith to (un)helpful help-seeking and coping responses by victims, survivors and un(helpful) rationalisations by perpetrators.
The second section presents the evidence on faith-based interventions that engage congregations and clergy in religious spaces and community settings or involve victims, survivors and perpetrators directly, such as in individual counselling or perpetrator treatment groups. The studies presented speci c faithbased programmes or interventions with clergy, faith-sensitive approaches involving therapists, psychologists, social workers and other domestic violence practitioners, and international development and public health programmes involving faith actors and communities. In the presentation of the evidence, emphasis is placed on the ways in which religious parameters were integrated in these interventions, the effectiveness of this integration and strategies for improving effectiveness.
3.1. The in uence of religious beliefs on victims, survivors and perpetrators of Domestic Violence / Intimate Partner Violence 3.1.1 Types and focus of studies Studies in this thematic category came from the USA, Brazil, India, Indonesia, Iran, Israel, Pakistan, South Africa and Zimbabwe. Most studies were conducted in a community setting, with the majority being quantitative cross-sectional studies, although the sample also included a large number of qualitative studies. A large majority explored the relationship between spirituality, resilience and IPV (de la Rosa et al., 2015) the relationship between IPV and suicidal ideation (Fischer et al., 2016), help-seeking behaviours (Tonsing and Barn, 2020), and coping strategies used by women to manage and cope with IPV (Bhandari,

Key ndings
3.1.2.1 Effects of faith or religiosity on how individuals respond to or engage in victimisation, violence or con ict As early as 2001, Ellison and Anderson published a study that involved adults in the US, which found that regular religious attendance was inversely associated with the perpetration of IPV. Among men, this protective effect was evident only for weekly attenders, whereas among women, the protective effect also surfaced for monthly attenders. Another study from the US by Wang et al. (2009) examined the religious beliefs and practices in relation to IPV experiences of 1,476 randomly selected Christian women. Findings suggest that women as well as their intimate partners who reported attending church more regularly were less likely to perpetrate or experience IPV. Women who had abusive partners at the time of the study or previously reported that while their church communities were less accepting of divorce in cases of DV, they had found that strength in God to nd the courage to leave.
In an attempt to understand better what aspects of religious experience can serve as a protective or deterrent factor in DV, scholars have attempted to capture the effects of speci c religious practices, such as prayer or both partners being religious. For example, Sharp (2010) conducted a study that involved a sample of 62 current and former victims of IPV from a wide range of religious, socioeconomic, ethnic, racial, and geographical backgrounds in the US and found that prayer served as an imaginary social support interaction that provided individuals with resources to perform individual emotion management strategies. This could overall act as a positive resource for preventing IPV. Austin and Falconier (2013), in turn, examined whether spirituality and dyadic coping protected partners from becoming psychologically aggressive toward each other. The study included 114 adult Latino partners, living together for at least 1 year, the majority of whom identi ed as Catholic. The researchers found that each partner's spirituality had a direct negative effect on their own psychological aggression and a direct positive effect on their own supportive dyadic coping and the couple's common dyadic coping. Moreover, each partner's spirituality also had an indirect effect on both partners' psychological aggression through increases in the couple's common dyadic coping. However, supportive dyadic coping was not found to mediate the relation between spirituality and psychological aggression. In another study, religious beliefs and practices among American Catholic and Orthodox Christian families (such as prayer and attending worship services) were related to seeking forgiveness and to resolving con ict in order to build and maintain uni ed and harmonious family relationships (White et al., 2018).
More recently, Zavala and Muniz (2020) conducted a study to determine whether religious involvement could reduce IPV victimization indirectly through the three elements of routine activities theory: motivated offenders, target suitability, and capable guardianship. The study involved 4,126 college students, who comprised the American subsample of the International Dating Violence Study. The authors found that participants with higher levels of religious involvement were less likely to report IPV victimisation but overall concluded that IPV victimisation could be reduced using faith-based strategies, such as religious counselling or preventive and intervention programmes that increase prosocial bonds between partners.
The evidence emerging outside the US and Europe moves along similar lines, suggesting a potentially positive effect of religious experience and faith on IPV, whilst evidencing that religious couples do not eschew violence. For example, Ghodrati et al. (2019) investigated the relationship between religious attitudes and DV against women in Iran. The study involved 210 Muslim participants from Shiraz selected from health centres. The researchers found that in cases with high levels of religious attitudes, the violence was reported to be lower, whilst the most severe violence was observed against participants with moderate religious attitudes. In Brazil, De Bernardin Gonçalves et al. (2020) explored whether religiosity was associated or not with diminishing violence in a Brazilian population-based representative sample. The study seemed to concur an inverse relationship between religiosity and violent outcomes, which was mediated by alcohol consumption. More speci cally, individuals with a religious a liation were found to be less involved in ghts and detention and those who considered religion important in their lives were less involved in ghts.
Some of the available evidence is more psychologically grounded, relating religious imagery to attachment theory and IPV. Certain studies have suggested that individual reliance on and appraisal of God as an attachment gure is not unrelated to attachment models that individuals develop through relationships with other humans in their life. Scholarly opinions vary on whether the relationship is one of correspondence (whereby attachment patterns with humans correspond to attachment patterns in individuals' experiences of God) or compensation (whereby attachment patterns with humans and God do not correspond because God functions as a substitute attachment gure). For example, Birgegard and Granqvis (2004) found that individuals who developed secure parent-internal working models tended to display more secure relationships to God. Hall et al. (2009) argued that correspondence operated at implicit levels of spiritual experience, nding that human attachment patterns were not associated with explicit spiritual functioning. Pollard, Riggs and Hook (2014) found that attachment avoidance was most strongly linked to the lack of a secure, positive relationship to God, but attachment avoidance could be less detrimental to marital functioning when the individual used more positive religious coping strategies.

Studies with victims and survivors
Studies exploring women's responses to DV / IPV from the perspective of their faith are abundant and generally agree on a dual in uence of religious beliefs on victims' and survivors' help-seeking attitudes and coping. While many victims may hesitate to leave an abusive situation as a result of what they understand as religious interpretations or standards, they generally report also nding support in their faith in order to cope and to become more resilient.
The study by Popescu et al. (2009) speci cally explored religious belief-based barriers that deterred women IPV survivors in a conservative Christian community in the US from changing their circumstances. The sample consisted of 42 adult women who self-identi ed as a current or former Seventh-day Adventist and had experienced an abusive intimate partner relationship. Nearly all of the women in this sample (93%) held at least one belief-based barrier, which included beliefs about marriage and divorce, stereotypes about Christians, and beliefs about Christian gender roles. The study observed that such beliefs were externally reinforced by other people in their lives, such as clergy, church members, family members, and partners resulting in holding them in abusive relationships. Khng et al. (2009) also explored factors that in uenced women's decision making in their process of leaving their abusive relationships, this time in an Indonesian sample. Study participants included eight abused wives from three different ethnic groups in Singapore, among whom two were Muslim, two Roman Catholic and one Buddhist. The study found that at the level of the exosystem / external factors, personal religious faith had a great in uence on the women's process of leaving. A realisation that the relationship cannot improve and will always stay abusive and that their abusive experiences are affecting their children were reported to in uence women's decision to leave abusive relationships. Zakar et al. (2012), in turn, explored coping strategies adopted by women victims of IPV in Pakistan. The study involved 21 married women of reproductive age (15-49 years) with experience of IPV all of whom were Muslim from the lower economic class. It was found that women used a combination of strategies to cope with violence, the majority of which were positive and helpful. One emotion-focused strategy was described as increased engagement in individual religious activities. Along similar lines, Drumm et al.
(2014) explored the spiritual coping processes for surviving and healing used by 41 adult Christian women victims of IPV who self-identi ed as a current or former Seventh-day Adventist. The analysis suggested that women moved from coping to building resilience and healing, which suggested a potentially empowering role for faithfulness and religious belief in psychosocial support provision.
In a different study context, Katerndahl et al. (2015) assessed the impact of spiritual symptoms and religious coping on attitudinal, behavioural and clinical outcomes among women in violent relationships. Adult women were recruited from six care clinics, who came primarily from the low economic classes and the Hispanic population. This study found that religious and spiritual factors had a consistently negative effect on clinical outcomes while the use of spiritual resources had variable effects. In this study, religious coping was only associated with staying in the relationship.
De la Rosa et al. (2016) examined the correlation between spirituality, resilience, and IPV among Mexican American women born in Mexico. The study included 54 women with experience of some type of violence from an intimate partner or ex-partner in the past 6 months residing in three different domestic violence shelters. Their religious a liation was Roman Catholic (47%) and Evangelical Christian (30%). The study found that higher levels of spirituality were correlated with increased resilience with 71% of the variance in resilience scores being explained by spirituality and the other controlled variables. Increased time spent in a shelter was associated with an increase in reported spirituality after controlling for education (β = −.12, SE = .05, p ≤ .05). Higher levels of perception of spirituality were associated with fewer types of abuse experienced (Exp[β] = 1.02, p ≤ .02) after controlling for resilience and education.
A study among abused Fijian women by Tonsing and Barn (2020) also explored help-seeking behaviours and practices and the role of religiosity as a coping strategy. The study sample included 8 Fijian women in abusive relationships aged 18 and above with children. In the interviews conducted the women did not justify IPV by cultural or religious justi cations and reported often seeking help rst from religious personnel, as opposed to secular domestic violence services. The authors explained that the women utilised their agency to cope with their situation through their faith and personal relationship with God, reading the Bible, praying, and having faith in God as main coping strategies.
Adjei and Mpiani (2020) explored the in uence of religious beliefs and practices on the perpetration of husband-to-wife abuse and the entrapment of victims in Ghana. The study involved 40 participants, including 16 women with experiences of physical and / or sexual abuse from current or past marital partner and 16 men who had in icted physical and / or sexual abuse on a current or past marital partner and 8 key informants (police, traditional elders, religious leaders, social and local government workers) from rural and urban Ghana, with the majority being identi ed as Christians. As a more negative outcome, the authors found that perpetrators constructed husbands' conjugal authority over their wives in terms of prescriptive religious norms, while victims constructed their entrapment in abusive relationships in terms of proscriptive theology of divorce in the bible.
Chadambuka (2020) explored the strategies that women in rural areas used to manage and cope with IPV in Zimbabwe. The study included 25 women who had experienced IPV and had left abusive relationships and 7 key informants (nurse, police o cer, two village heads, one gender o cer and one social worker). The study identi ed two types of coping mechanisms: culture-based coping mechanisms (e.g., acceptance and placating; self-blame; de ance) and religious-based coping mechanisms (e.g., pastoral mediation and prayers).
A few studies took a more psychologically-grounded approach in their assessment. For example, Fischer et al. (2016) examined existential and religious well-being as two potential protective factors in the IPVsuicidal ideation link in a sample of African American women in the US. The study included 111 African American women of low socioeconomic status between the ages of 18 and 56, who presented to a large public sector hospital for medical or psychiatric reasons. The study con rmed a direct and positive effect of IPV on suicidal ideation and found that existential well-being was a mediator of this relationship. However, religious well-being was not found to serve a similar mediating purpose, pointing to the signi cant importance of holistic well-being.
Bhandari (2019), in turn, focussed on the coping strategies of women in India using a problem-focused and emotion-focused conceptual framework. The study involved 21 abused women over 18 who had been abused by an intimate partner and then sought help from agencies serving DV victims and had the following religious a liations: Hindu (n = 13), Buddhist (n = 5), Muslim (n = 2), and Christian (n = 1). Five themes emerged into problem-focused and emotion-focused coping strategies. Pertinent to this review, emotion-focussed coping strategies included: (c) spirituality (d) hope that the abuser will change, keeping quiet, crying a lot, and keeping oneself busy.
Other studies examined IPV attitudes and responses engaging religious communities. For example, Ml et al. (2010) explored patterns of victimisation and coping in a conservative Christian denomination in the US. This was a cross-sectional study involving 1,823 female and male participants in a community setting, 68% of whom identi ed as conservative in terms of their religious practice. The study found that negative coping was signi cantly predicted by childhood victimisation and current victimisation and that religiosity had a small protective effect against negative coping-which disappeared completely when childhood victimisation was taken into account. Witnessing parental violence was also a strong predictor for negative coping, with its effect being slightly diminished by current victimisation. Band-Winterstein and Tuito (2018) explored how ultra-Orthodox Jewish women in Israel affected by IPV experienced the spouse selection process. Most of them felt that it was the matchmaker's fault for putting them into violent relationships. They did not refer to any personal and interpersonal relationships or intergenerational transmission of violence related factors as a reason for their experiences of being in a violent relationship but felt that it was the matchmaker and system's fault that put them in the speci c situation.
Another study relied on interviews with 24 social workers in daily contact with Haredi women experiencing IPV (Band-Winterstein and Freund, 2016). Findings suggested that religion was used as a double edge sword and that women experienced emotional as well as spiritual violence where the husband used religion to not only justify the violence but also to control the woman. Evidence from more sociological studies of domestic violence suggest that religious women who experience husband abuse might tend to endure and to forgive it, often as a direct result of how they understand and embody religious teachings or vernacular religious traditions (Shaikh, 2007;Merry, 2009, 68;Mardsen, 2014; Nason-Clark et al., chapter 2). At the same time, female victims may resort to religious beliefs to condemn the abuse and through their ordeals may acquire a more justice-oriented understanding of their faith, helping them to address the harmful situation (Shaikh, 2007;Johnson, 2015).

Studies with perpetrators and the general male population
Studies involving perpetrators or the male population more generally were fewer in the sample. In one of the studies reviewed, Hayati et al. (2014) explored men's views on masculinity and the use of violence within marriage in order to gain knowledge on how to engage men in prevention of DV in rural Indonesia.
The study involved holding focused group discussions with 44 male community leaders. The authors identi ed three different groups of men with different positions of masculinity and thoughts about gender and attitudes towards violence in marriage: the traditionalist (with higher acceptance of DVA as a tool to uphold men's position in marriage), the pragmatist (who believed DVA was undesirable but sometime necessary to correct the wife's behaviour), and the egalitarian (who saw men and women as equal and complementary to each other and felt DVA was unacceptable).
Another study investigated the interactive effects of alcohol use, hostile sexism, and religious selfregulation on male-perpetrated IPV (Lynch and Renzetti, 2020). The study involved 255 male, aged 18 or older, living in the US. With regard to religious self-regulation, the authors found that introjected religious self-regulation (when religious behaviour is performed because of pressure of others and external in uences rather than personal positive motive) was positively associated with hostile sexism and positively associated with perpetrating physical IPV. Identi ed religious self-regulation was negatively associated with physical violence perpetration. The implication of the study was that alcohol abuse elevated the risk for physical violence perpetration among men high in introjected religious self-regulation and low in hostile sexism, and reduced the risk for perpetration in men high in identi ed religious selfregulation and low in hostile sexism.
As it was mentioned earlier, Adjei and Mpiani (2020) found that in Ghana perpetrators employed a prescriptive theology of male headship and authority in marriage, which was related to IPV. Additional insights emerge from within religious, gender, anthropological and sociological studies from the US, Canada, South Africa and other contexts. These suggest that some abusive men may have distorted understandings of religious teachings, not unrelated to the family environment they grew up in, which they could use to justify their abusiveness (Shaikh 2007

Interventions involving clergy
A common reference for the scholarship that looks at faith-based interventions involving clergy is the understanding that religious personnel, the discourses they use and their responses to communities can both contribute to the continuation of the problem of domestic violence and serve as a positive in uence in efforts to address the problem. While the level of awareness and preparedness among clergy to respond to domestic violence has improved in the past decades, problems and challenges remain. The existing evidence suggests the need for more customised, theology-speci c training for clergy and for building more religious-secular equitable collaborations.
A literature review conducted by Shannon-Lewya and Dull published in 2005 that looked at how Christian clergy in the US can be a resource to victims of domestic violence found that clergy have different levels of understanding and preparedness, and that often their understanding of theology and exegesis could hinder them from responding to victims in ways that minimise the risk of being abused. The authors cited studies with American Protestant clergy, which found that the involved clergy still upheld hierarchical understandings of gender relations by bestowing husbands' authority over wives, which could contribute to a tolerance of women's abuse in marriage. Other beliefs, such as that suffering is virtuous or that marriage is sacred and indissoluble were also identi ed as exacerbating or contributing to an ambiance of tolerance and unresponsiveness to situations of DV / IPV. Moreover, the article reported that while the majority of the interviewed clergy held the men responsible for most of the abuse, they still placed some responsibility on women either for inciting the situation or for staying in the relationship, which could overall contribute to an insu cient level of responsiveness to victims.
Studies that examine clergy perceptions and responses to domestic violence in their congregations generally recognise shortfalls in their preparedness and the need for more training. For example, the study by Brade and Bent-Goodley (2009), which involved Protestant religious leaders, found that clergy perceived the problem of domestic violence to be salient in their communities and held that additional training and resources on domestic violence were needed. The study seemed to imply that partnerships between social workers and clergy were needed to achieve a more adequate response. Along similar lines, a study by Moon and Shim (2010) that examined nine protestant pastors' perceptions of and responses to IPV found that pastors tried to support victims and even couples together and would not hesitate to support divorce if necessary. The analysis of the interviews evidenced that many participants' hesitation to speak about IPV was related to their feeling unprepared to respond to the problem with con dence, for example, due to lacking training or resources.
A study by Sisselman-Borgia and Bonanno (2017), in turn, examined Rabbis' opinions about and responses to domestic violence. As in an aforementioned study, the participants were very willing to help, but felt unequipped to provide appropriate counselling support. A 2019 study by Gezinski et al. on the lived experiences of IPV victims who approached their religious leaders for guidance in a community of Latter-Day Saints, found again that generally pastors were unprepared to respond to IPV in an informed manner, placing faith rst and limiting their support to prayer and to encouraging church attendance. The study highlighted the need for comprehensive training for religious leaders, education that recognises the impacts of trauma and that teaches religious leaders practical, trauma-informed approaches for addressing IPV. A more recent study by Shaw et al. (2020) that explored Black female clergy's role as responders to IPV among Black women in their congregation in the USA found that female clergy considered it their responsibility to care for domestic violence victims and sought to provide holistic, culturally-sensitive support. However, they felt that they had limited support from the wider Church they a liated with and limited external resources to rely on.
Similar results emerged in studies with migrant populations. A study by Behnke, Ames and Hancock (2012) examined 28 US Latino church lay clergy and pastors' beliefs about domestic violence and their reactions to three domestic violence vignettes to explore how they might react. Their responses included actions such as providing counselling to both victims and perpetrators, offering spiritual guidance, suggesting legal action, referring victims to other community services to protect them from risks and providing job-related advice. While many responses were constructive and genuinely concerned, other suggestions for action evidenced the clergy's limited familiarity with safeguarding risks, as seen wanting to advise a perpetrator while still living with the abused party.
A survey conducted by Choi (2015) with 152 Korean American Protestant clergy in their congregations found that the majority of respondents reported providing counselling support to IPV victims, although one third reported they had never referred IPV victims to additional / external resources. Only 16% felt prepared to respond to domestic violence in their communities, which evidenced the need for more training for clergy to understand how domestic violence manifests in their speci c cultural context and how it might be addressed. Another study by Choi and Cramer (2016)  Some of the key ndings included: communities' own recognition for the role that faith and religious leaders should play in addressing gender-related issues, the di culty for faith leaders to separate religious teachings from cultural standards, and the tendency of religious leaders to interpret sacred texts in a rigid manner that could be conducive to the problem and its continuation. A cited study from DRC, for example, found that members of faith communities felt compelled to comply with religious leaders and their positions, such as regarding hitting one's wife or sleeping with her when she did not want to. This motivation was also strong with those who did not belong to these faith communities, evidencing the in uence that religious leaders have in shaping public attitudes more generally. The overall evidence suggested that people did not consider faith leaders to be well-disposed or equipped to respond effectively to victims of sexual and gender-based violence (SGBV). The paper noted the need for multilevel engagement and integration and equipping clergy with theological knowledge to move towards interpretations that promote gender justice. However, any such engagement and support for clergy would need to be sustained over time to ensure personal change and move beyond approaches based on a mere assimilation of information.
In another study published in 2017, Le Roux and Bowers-Du Toit drew on data collected during a scoping study on the role of faith communities and organisations in the prevention and response to sexual and gender based violence. The review included academic and gray literature, as well as organisational reports, research reports, internal evaluations, manuals, and pamphlets and was accompanied by 20 key stakeholder interviews and an electronic survey involving faith and gender-based violence (GBV) specialists around the world. The ndings from the consultations emphasised the faith leaders' complicity within the continuation of patriarchal traditions, leading to the misinterpretation of sacred texts, the double role of religion in serving as a support and healing system for victims but also making them vulnerable as a result of certain misplaced beliefs about how a 'good' religious person should respond to abuse, and the clergy's disbelief or lack of awareness of the extent of GBV in their communities. However, the overall evidence con rmed that clergy are often willing to address the problem, but they may feel unequipped to do so and may need continuous training and support to be effective.

Examples and effectiveness
The literature on clergy-centred interventions that address DV is substantial. The public health and international development sectors also include numerous examples of faith-based and clergy-centred approaches implemented at community level. Most assessments of such programmes are qualitative, with a visible dearth of studies that sought to evaluate programme effectiveness using more quantitative methodologies, such as randomised controlled trials.
In a 2009 paper, Danielson et al. presented the Set Free Ministries programme, a comprehensive faithcentred domestic violence programme initiated by four women attending Moody Church in Chicago, Illinois, who found the need to address domestic violence in their congregation and to support female victims of childhood abuse. The programme focused on providing mothers with support in the context of domestic violence, although it recognised the need to work also with perpetrators in order to reduce the likelihood of intergenerational violence. In its implementation, the programme incorporated faith-based healing through prayer, the Bible, support from the wider community and referral to other community resources. In terms of prevention, the programme included the provision of training to families and pastors to help develop healthy individual identities and relationships. Direct services provided by the programme included operating a hotline crisis response line, conducting risk assessments for the women seeking support, planning for the women's physical, spiritual and emotional safety, operating a women's recovery support group, holding individual counselling sessions and offering social support at court hearings.
The Religion and Violence e-Learning (RAVE) programme could be considered another such intervention.

RAVE was established as an innovative online platform for the delivery of DV training and resources for clergy and congregations developed on the basis of many years of research by the authors and their
colleagues at the University of New Brunswick in Canada. According to its creators, the RAVE website has served as an online training venue for religious leaders, and others working in domestic violence responses publishing training materials that are adapted to the needs of clergy, social workers, affected individuals and other interest groups. Another aim of the RAVE programme has been to promote more partnerships between secular and religious stakeholders for a more effective response to domestic violence.
In a 2014 article, Hancock, Ames and Behnke described a project designed to stimulate and support appropriate responses to family violence in rural immigrant churches in the USA. The project team implemented a community-based participatory research approach by involving relevant stakeholders and by proceeding in culturally-relevant ways. In order to achieve this cultural sensitivity the programme sought to promote women's protection without undermining the traditional position of the father and husband in the community, which required developing a biblically-informed approach that held men accountable of the abuse and encouraged behavioural change on men's part. As a way of assessing this intervention, the authors conducted a post-test data collection that enquired about the pastors' learning before and after the intervention. The survey indicated that most of the pastors had been receptive to the content and planned to incorporate it into their church work, but some felt that they needed more training on how to speak to and help abused women. This implied that following up the intervention with continuing support for pastors would be needed to contribute to more effective outcomes.
In another paper, Kim and Menzie (2015) presented and reviewed Shimtuh, a Korean domestic violence and sexual assault programme located in the San Francisco Bay area, and its collaborative models with secular organisations. The programme reached primarily Christian and Buddhist communities and comprised direct services and advocacy for women and children affected by domestic violence, and community organising and engagement to change norms that contributed to the problem and its continuation. Among other activities, the programme delivered a two-day annual domestic violence training to faith leaders, which was bilingual and culturally adapted. Leveraging on their role as social change agents, Shimtuh organised the clergy to participate in campaigns to end domestic violence. Over the 12 years that Shimtuh had been operating, the models of collaborating with faith leaders and institutions shifted from Outreach, Training and technical assistance to Community organising campaign and Shared leadership. These models re ected Shimtuh's commitment to engage religious stakeholders in an active manner and see them take on the leadership in their communities to prevent and to address domestic violence. What appears to have worked particularly well in this programme is the exibility of Shimtuh to adapt to the evolving relationship with faith leaders and faith-based institutions and to learn from each phase of the collaboration. with communities to raise awareness and with municipalities to inform policy-making. SADD, in turn, implemented the 'Church Preventing Gender-based Violence' project to provide biblical, theological and advocacy resources to address violence affecting women offering social, pastoral and material support to women eeing domestic violence. The authors reported that the regular work of SADD with clergy and non-ordained religious leaders gradually led the latter to engage more directly with the issue, incorporating it in sermons and responding better to victims and survivors. Given an increasing salience of religious identity in policy-making, especially around gender issues, the paper stressed the need for faith-inclusive approaches and employing theology to denounce violence affecting women.
In another 2016 paper, Petersen presented a study from South Africa that re ected on the work of the South African Faith and Family Institute (SAFFI), a multi-faith non-pro t organisation established in 2008 to address the faith dimensions of violence against women (VAW) in the country. The paper discussed feminist approaches to VAW, which the author argued had been historically in uenced by white feminists' interests and had not paid su cient attention to the context-speci c needs of South Africa and many women's faith-oriented lives. SAFFI was established in order to advance in the country a coordinated, multi-sectoral and culturally competent approach to addressing domestic violence. The means included providing resources to clergy to support victims/survivors and to hold perpetrators accountable, employing theology-informed discourse to counter 'patriarchal traditions' conducive to women's abuse and using scripture to encourage mutually supportive and ful lling models of marriage. In its approach, SAFFI employed a theoretical framework drawing from the concept of Ubuntu and the Human Development theory, ultimately aiming to advance the dignity of the human being. One of the programme's main achievements was its inter-faith character, bringing together clergy from faith traditions that would not otherwise communicate with each other. Moreover, the programme's respectful engagement with clergy had reportedly brought many to re ect on their own practices and discourses and even to open up about challenges in their own family lives.
In a more recent study (2020), Le Roux et al. presented an evaluation of a three-year intervention that addressed VAWG, and especially sexual violence, by engaging with communities of faith and their leaders in DRC. Two community surveys were conducted, one before and one after the intervention, in three health areas in Ituri Province in the DRC. At both baseline and endline, data was collected from male and female members of randomly selected households in 15 villages ( ve per health area) in which the intervention was being implemented. At baseline the sample comprised 751 respondents (387 women, 364 men) and at the endline 1,198 respondents (601 women, 597 men). The study showed signi cantly more equitable gender attitudes and less tolerance for IPV at the endline. Positive attitudinal change was not limited to those actively engaged within faith communities, with a positive shift being noted across the entire community around gender attitudes, rape myths and rape stigma scores. This again seemed to point to the in uential role of religious discourse in shaping social attitudes and norms, affecting the wider society.

Challenges to faith-secular collaborations
One recurrent suggestion in the literature reviewed is the need for more collaborative approaches between secular and religious stakeholders and providers to achieve more effective responses. However, numerous factors were identi ed to stand in the way of fruitful partnerships.
In a 2010 paper, DeHart examined current and envisioned models of faith-secular collaboration, and the perceived role that such collaboration could play in the lives of crime victims. The paper presented ndings from a descriptive needs assessment that preceded a national project to link faith-based organisations and victim service programmes in ve high-crime neighbourhoods. Approximately 90 participants were interviewed, including faith leaders of Christian, Muslim, Jewish, and Hindu background, victim service providers, and other professionals. The ndings showed that collaboration was hindered by numerous factors, including differences in philosophies and misperceptions or stereotypes about each other that hindered constructive engagement with each other. The authors proposed that building mutual understanding of the sources of differences could help to overcome those.
In addition, trainings addressing ethical standards, such as con dentiality and appropriate approaches, could help to address concerns about skills and approaches to bring the different stakeholders in alignment in terms of standards of good practice.
In a paper published in 2017, Le Roux and Loots explored how the faith versus secular binary had in uenced how GBV interventions had been implemented within public health and international development. The authors drew insights from a scoping review conducted in 2015 funded by DFID that included 20 key informant interviews that were done with faith and GBV experts (eight men and 12 women) from 15 different international organisations, 13 identifying overtly as faith-based organisations (FBOs). The analysis established the existence of what the authors referred to as secularist biases in the post-secular era, which were identi ed as two of the main obstacles to a productive engagement with faith in GBV interventions. Interviewees, for example, spoke about their marginalisation in programmes involving secular actors or their general and consistent portrayal as negative actors. Others spoke about a widespread underlying view in the sector that faith is backwards, or the frequent instrumentalisation of religion just to tick a box with secular agencies with tools and resources used by faith organisations being appropriated or used by secular agencies in a non-equitable model of collaboration. According to the study, many admitted that some of this critique was not unwarranted, explaining that the clergy often taught through a cultural patriarchal lens that contributed to women's abuse and that faith communities themselves could contribute to stigma and shame silencing the victims. It was reported that faith leaders could also hesitate to work with secular actors to address the problem together. In addition, clergy often faced di culty in questioning fundamental beliefs upheld within the faith and often perceived suspiciously other religious traditions, which challenged collaborative approaches. In parallel, interviewees spoke about a perpetuated myth that secular approaches and discourses were neutral in contrast to what were often presented as biased or value-laden approaches employed by FBOs. In discussing the ndings, the authors stressed the importance of overcoming the religion-secular binary as this not only alienated faith stakeholders, but also mistakenly represented faith actors as engaging only in 'faith-related' activities as de ned by a very rigid, Eurocentric mainstream secular discourse.

Faith-based approaches in counselling and psychological support
Faith and spirituality in psychological approaches The integration of spiritual and religious parameters in psychological approaches has been gradual and heavily skewed to re ect the realities of industrialised western societies. According to Hodge (2005), this integration started to become more visible starting in 1985, with more counsellors becoming interested in knowing how to engage with spiritual parameters in their practice and the eld of psychological counselling granting increasing attention to cultural competency in an effort to cater to clients from diverse religio-cultural traditions. Other motivations have included the clients' own interest in having their spiritual or religious experiences incorporated in counselling sessions and new research evidence on the potential positive effects of religious beliefs on family and married life. Writing in 2005, Hodge noted that spiritual and religious parameters had not yet been well integrated in psychological and counselling education, leading him to produce a paper that presented six recently developed tools to integrate spirituality / religious beliefs in counselling. In presenting these approaches, the author followed the guidelines of the Joint Commission on Accreditation of Healthcare Organizations (JCAHO), recommending an initial assessment on the importance of spirituality and religious beliefs for the client. The assessment approaches were described as: spiritual histories, spiritual lifemaps, spiritual genograms, spiritual ecomaps and spiritual ecograms.
In a 2008 paper, Marterella and Brock spoke about the need for therapists to become more educated and con dent to engage with spiritual and religious parameters in their counselling practice. The authors discussed the importance of therapists being able to engage productively with the religious beliefs of their clients, but also being able to challenge those beliefs that can be contrary to the objectives of the therapy, without disrespecting or questioning the importance of religious convictions in their clients' lives. The authors referred to the particularly challenging task for therapists to not allow their own beliefs to interfere with a productive engagement with their clients' religious beliefs and stressed the need for selfre exivity. In the authors' understanding cultural sensitivity should not be dissociated from religious beliefs since in many communities cultural and religious parameters intertwine and in uence the identity of the individual in complex ways. The paper listed what were known techniques for incorporating religious parameters in therapy, including by use of spiritual genograms and spiritual ecomaps.
It is also notable that de nitions of spirituality / religion have started to be diversi ed and have themselves increasingly become the subject of research. Most of the studies reviewed adopted different de nitions for spirituality and religious belief, which were however linked. Generally speaking, spirituality has been conceptualised as a broader existential meaning and relationship to a higher power, while religious beliefs more directly with one's faith and faith-driven practices. Hodge, for example, de ned spirituality as an individual's existential relationship to God / Ultimate Transcendence, seeing religion as owing from spirituality and expressing this relationship in beliefs, forms and practices developed as a result of speci c religious traditions. Marterella and Brock (2008) de ned spirituality as "the human experience of discovering meaning, purpose, and values, which may or may not include the concept of a God or transcendent being" and religion as "the formal institutional contexts for spiritual beliefs and practices" (p. 64).
More recent studies have actually explored how those immediately involved have themselves de ned and engaged with spirituality in their practice, recognising that there has been a diversity in understandings, especially across different geographies and cultural contexts. A study that looked at domestic violence practitioners' own de nitions of spirituality involving 12 countries found that the large majority (45 percent) spoke of spirituality in conjunction with the values of unity, peace, equity, unconditional love, forgiveness and letting go (Pandya, 2017). About a third spoke of spirituality in reference to mysticism and transcendence. Overall, the study found that practitioners from European countries, United Kingdom, United States, Canada, and Australia, Christians and those with higher spiritual openness tended to speak of spirituality in terms of mindfulness, peace and letting go in supporting domestic violence victims and survivors. On the other hand, practitioners from Asian and African countries placed emphasis on the role of spirituality in helping victims and survivors to let go and to achieve empowered reconciliation, where women negotiate space within a rights framework.
Faith-based approaches in counselling: exemplary programmes and effectiveness In a 2003 paper, Ronel and Tim presented the Grace Therapy programme, a model for male batterers' group therapy based on a 12-Step programme, which directly responded to historical barriers in treating male perpetrators of domestic violence, including a systematic emphasis being placed on the criminality of the perpetrator hindering a true commitment to treatment. In uenced by the model of Alcoholics Anonymous, the programme used spirituality language to achieve the men's self-transformation by replacing self-centeredness with "God-centeredness" and by instilling within the male batterers a win-win principle of partnership that is based on unconditional giving. Spirituality was de ned in this case "as the quest for God, as one understands God" (p. 65). The programme reportedly differs from cognitive, behavioural or other approaches in its assumptions, such that abuse is both a mental and spiritual imbalance, with the latter being understood as the self-centeredness or sel shness and the result of feeling powerless as a result of attempting to control what one cannot control. The authors explained, Faith in God was not a requirement to be part of the therapy, but a possible outcome as a sign of spiritual development. Like most available programmes, the Grace Therapy Programme used a group therapy approach informed by evidence that providing therapy in a group format fosters identi cation with other participants and the development of a cohesiveness in the group that can lead to productive results.
In another article, Ronel and Claridge (2003) presented how the Grace Therapy programme could be effective in treating substance abuse in male batterers of domestic violence as a means of reducing the severity and occurrence of the problem. The authors presented the programme that was implemented at the Tel Aviv Center for the Treatment of Domestic Violence involving over 300 men who had participated in on-going groups between 1994 and 1999. The men's participation could last from weeks to years and many of the men suffered from substance abuse as well, allowing the current study. The authors mentioned that male batterers spoke of feeling powerless to control their circumstances and to manage their life, leading them to frustration and fostering their abusiveness. Recognising that they were never in full control and that others were not responsible for their powerlessness was identi ed as a vital step for participants to work towards recovery.
In a 2014 paper, Hook et al. presented a naturalistic study that examined the role and effects of religion in couple therapy. The study included 44 therapists and 68 couples entering couple therapy with religious couple therapists. Clients assessed six religious and nine non-religious techniques used in therapy, which were related to religious commitment and relationship satisfaction. The results showed that religious techniques were used in fewer than half of the sessions, which were related to the clients' religious commitment. This was taken to mean that therapists were probably adapting to religious clients, taking into consideration the importance of religious beliefs in their clients' lives. The study also reported that clients expressed improvements in relationship satisfaction overtime, which suggested that incorporating religious parameters in couple therapy was effective with this group of clients. One limitation of the study was that the therapists consulted identi ed as Christian, which meant that the study was less able to indicate what happens when there is a discrepancy between the therapist's religious commitment (e.g. when a therapist identi es as secular) and a client's religious commitment. Alliance. The study used a quasi-experimental design to assess how religious values and high Christian beliefs relate to ratings of marital therapy situations. Participants were asked to rate four marital therapy situations: Christian therapy using Christian practices (e.g. prayer), Christian therapist using psychological techniques only, non-Christian therapist willing to use Christian techniques and non-Christian therapist using psychological techniques only. The study found that highly religious Christians received therapy different than Christians categorised as low to moderate in religiosity. Those with high Christian beliefs expressed a signi cant difference between the non-Christian therapist using psychological practices only (mean 1.88) and all other types of therapy (mean range: 4.33-5.71), as well as between the non-Christian therapist (mean 3.12) and the Christian therapist (mean 5.09). The main implication of the study was that the use of religious techniques and the religiosity of the therapist may in uence highly religious couples in their decision to engage in therapy.

Receptiveness of faith-based counselling approaches and religious clients
In another study, Fowler et al. (2011) examined differences between shelter and faith-based service utilisation and satisfaction in a sample of 73 women residing at a domestic violence shelter in central Texas, US. In the case of this study, focus was placed on spirituality, with religiosity being subsumed within the former. Along the lines discussed earlier, spirituality was de ned as pertaining to a way of being, an awareness of the transcendent, deeper meanings about life and an interaction with a higher power. Participants were asked to complete questionnaires in order to collect information regarding psychosocial characteristics, intimate partner abuse, spirituality, service utilization, and satisfaction.
Overall, the study found that spirituality explained roughly 16% of the variance in domestic violence shelter service utilisation. Moreover, women survivors with higher levels of spirituality were less likely to utilize shelter services despite the IPA levels experienced. However, women survivors who reported more IPA experiences were less likely to report satisfaction with faith-based resources.

Conclusions
The current paper made an attempt to integrate the evidence on religion and domestic violence from numerous disciplines and sectors, including psychological counselling, international development and public health, and anthropological and sociological research in order to improve understanding of how religious beliefs in uence victim, survivor and perpetrator thinking and behaviour and how they can best be leveraged in domestic violence interventions and counselling. The review was particularly interested in diversifying the current scholarship by collecting studies from across the world and to depart from rigid, Eurocentric conceptualisations of religion or spirituality to account for diversity in religious experience as encountered cross-culturally.
A rst important insight that emerged from the synthesis of the evidence is the double effect that religious beliefs and discourses can have in DV / IPV experiences. On the one hand, religious beliefs can feed hesitation among victims to leave their abusive situations and can be employed by perpetrators to justify the abuse. On the other hand, religious beliefs and faith were consistently reported to serve as coping mechanisms for victims, survivors that could help them to regulate their emotional states and to build resilience to exit pernicious situations. Spirituality and a God-centred conscience were also found to help perpetrators minimise of overcome their abuse in some treatment programmes.
The effects of religion on domestic violence victims and perpetrators could not be properly assessed without recognising the different de nitions of religiosity or spirituality adopted in the diverse studies that were reviewed. Indicative of this variability was the international study that found that domestic violence practitioners from western societies and those with higher spiritual openness tended to understand spirituality as mindfulness, peace and letting go in their work with domestic violence victims and survivors, while practitioners from Asian and African countries placed emphasis on the role of spirituality in helping victims and survivors to let go and to achieve empowered reconciliation, where women negotiate space within a rights framework. This could re ect the different gender realities and social structures in these countries, with gender inequalities being more prominent and an important hurdle for female survivors to overcome or exit abusive situations in more traditional patriarchally structured societies.
Moreover, different levels of religiosity or spirituality seemed to be associated with different effects. As some studies showed, in some cases a very high degree of religiosity or spirituality could have pernicious effects, as compared to a low or medium level of religiosity. Very spiritual women who were victims of IPV, for example, were found to be less likely to utilise domestic violence services. The effects on men, and the more speci c group of male perpetrators, was equally nuanced. Some studies found that perpetrators could use religious language or theology to justify their authority over their intimate partners or spouses and some could even abuse their partners spiritually. Studies from other parts of the world found that men identifying as religious had a lower likelihood of engaging in ghts and were more likely to condemn abuse or consider it inappropriate. These differences could re ect different relationships to one's faith, but also the variable theological teachings and the diverse religio-cultural contexts in which research took place.
In regards to faith-based approaches, the literature evidenced well-established practices within psychological counselling to integrate religious and cultural parameters. The studies that were reviewed pointed to a diversity of clients who related differently to their faith, in uencing their receptiveness to and satisfaction with counsellors employing secular psychological methods versus counsellors who integrated religio-cultural parameters in their counselling and counsellors who themselves identi ed as religious. The faith-based approaches that seemed to be particularly effective within counselling psychology were those that understood that religious beliefs had to be addressed within the speci c cultural context of the client and that were based on the psychologist's re exivity about their own beliefs and how these could interfere with counselling their religious clients.
What seemed to be missing from the literature on faith-sensitive or spiritual psychological approaches was a discussion of how evidence on attachment models and spirituality or dyadic regulation in religious couples applied or were leveraged in supporting religious clients. Future studies should aim to identify more closely and systematically how these psychological dynamics can improve the effectiveness of counselling when therapists are aware and leverage on this knowledge.
The evidence on faith-based interventions in congregations and communities revealed numerous examples of effective programmes and approaches led by a variety of actors, including churches, congregation members, priests or organisations working in community development. The common thread of these very diverse approaches seemed to be the importance of cultural sensitivity in these programmes to ensure that these were relevant and impactful with those they aimed to support. This cultural sensitivity was equally relevant to those working in LMICs and those supporting migrant populations in western industrialised societies, with the important insight that religious understandings among migrant communities could re ect newly acquired theological in uences encountered in the host society and not solely or necessarily beliefs upheld in the home country.
Interventions involving clergy evidenced numerous commonalities across geographies, but also some differential needs given the context in which they operated. While the evidence from the US and other western contexts pointed to an increased awareness amongst clergy of domestic violence in their communities, a lack of trauma-centred training and the need for continuous support from religious institutions to address the problem adequately, studies from non-western, studies in tradition-oriented or LMIC contexts focussed on the central role of the clergy in addressing marital problems and in mediating domestic violence situations and how best to leverage on this role. In such societies, the pervasive nature of religious idiom in social life has meant that religious tenets have often been indistinguishable from cultural practice, reinforcing rigid interpretations amongst the clergy and the community. On the other hand, the in uence of the clergy in collective and individual life has placed the clergy in a key position to in uence collective attitudes and human behaviour.
Regardless of geography, it seemed to be agreed that clergy-centred interventions could be more effective if clergy were better equipped with theological / exegetical knowledge to respond to distorted deployments of religious language in the community, were willing to address the topic of domestic violence in public sermons, understood how to respond to domestic violence victims and perpetrators with consideration of their psychological states and likely safeguarding risks facing them, and integrated more substantively with secular systems responding to domestic violence. In fact, the divide between secular and faith actors and organisations, as seen within international development and public health, seemed to be a key hurdle to achieving integrated approaches and constructive partnerships. The literature review suggests that more attention should be given to address biased preconceptions that hinder religious and secular stakeholders to learn from each other and to collaborate constructively. This could be achieved by fostering the understanding that the divide between religious and secular actors and organisations is itself more mythical than realistic since neither are secular organisations without religious in uences, legacies or motivations, nor are faith-based actors necessarily different in a radical way from their secular counterparts.
Relating the two group of studies and their ndings as outlined above, would lead to the conclusion that faith-based interventions can become more effective if they recognise and understand that two-way effects that religious beliefs can have on victims / survivors and perpetrators, that these effects will differ depending on how one relates to their faith or level of spirituality, and that religious in uences need to be understood in reference to the religio-cultural context of the target group and their wider sociological realities. The evidence suggests the need to move towards a more context-speci c approach in designing faith-based interventions, that is theologically-grounded and considers carefully the sociological and psychological conditions of the individuals and groups it aims to support.

Limitations
In considering these insights, it is important to recognise the limitations of this review. While a rigorous search strategy was used to collect evidence from different disciplines and sectors, it is possible that other relevant and important studies were missed. Moreover, given the extensive evidence reviewed under the two questions driving the review, it was not possible to incorporate evidence from the psychological studies that examined the effects of religious beliefs on mental health and personal psychology more widely and beyond the context of domestic violence. Such an integration will be attempted in a subsequent paper building on the results of this one. Lastly, it is signi cant that the review was limited to studies in English. It is likely that many more relevant studies exist in LMIC contexts in national, regional or local languages not assessed in this review.
In terms of the studies reviewed, it is important to recognise that di culty of comparing across study results due to differences in study design, context and methodologies to assess interventions. Therefore, in presenting the result emphasis was placed on understanding the individual studies, and relating their ndings to each other only where and as feasible. A more robust comparison of studies that have assessed interventions could be undertaken in the future, exploring also the prospect of a meta-analysis where possible.
Declarations Figure 1 PRISMA ow diagram for systematic review