Introduction

Couple violence suffered and perpetrated by sexual minority peopleFootnote 1 was largely understudied until a few decades ago (Kimmes et al., 2019). In contrast, research and public opinion focused primarily on violence within heterosexual couples, influencing and being influenced by a mainstream heteronormative discourse on couple violence mainly focused on violent men who abuse their female partner because of a patriarchal and sexist culture that justify these behaviors as expression of masculinity (Rollè et al., 2020, 2021).

Nevertheless, many studies demonstrated rates of IPV among sexual minority people that are comparable, if not higher, than those identified among heterosexual couples (e.g., Walters et al., 2013; West, 2012). Establishing firm conclusions regarding the prevalence of IPV among same-sex couples seems particularly complex because of methodological limitations (e.g., lack of generalizable data, differences in the operationalization of IPV) and differences between research (Rollè et al., 2018, 2019). In addition, studies with large or representative samples have been limited and mainly conducted in US states, while data from European countries are still lacking and other research in this direction are needed. However, a representative study by Walters et al. (2013) showed alarming results: nearly one-third of sexual minority men and one-half of sexual minority women in the USA reported having suffered psychological or physical IPV in their lifetime. In addition, no significant differences emerged in the prevalence of IPV between lesbian and heterosexual women, and gay and heterosexual men (Walters et al., 2013). A meta-analysis by Badenes-Ribera et al. (2015) confirmed these results among lesbian women, finding a mean lifetime prevalence of IPV victimization of 48%.

Despite the widespread prevalence of this phenomenon, few research has been conducted on IPV among sexual minority people. A study by Edwards et al. (2015) found that only 400 (approximately 3%) out of the 14,200 studies published between 1999 and 2013 that addressed couple violence examined participants with a non-heterosexual orientation.

Although attention on couple violence among sexual minority people has increased in the last decades, the data available are still scarce, and influenced by methodological limitations. For example, most researches have used convenience samples and are cross-sectional in nature. Differences in the operative definitions of violence and sexual orientation emerged as well and make it difficult to compare results and draw firm conclusions about characteristics, antecedents, and consequences of couple violence in sexual minority people (Mason et al., 2014; Murray & Mobley, 2009).

Many similarities have been found between IPV in sexual minorities and heterosexual people such as the cycle of violence (Messinger, 2011; Walker, 1979; Whitton et al., 2019), the forms of suffered abuse (i.e., physical, psychological, sexual, and controlling violence, and unwanted pursuit), and some of the associated factors — for example, relationship satisfaction (Balsam & Szymanski, 2005), mental health (Sharma et al., 2021), personality (Landolt & Dutton, 1997), adult attachment (Bartholomew et al., 2008a, b; Gabbay & Lafontaine, 2017b), family-of-origin violence (Fortunata & Kohn, 2003), and substance abuse (Wei et al., 2020a, b).

However, peculiarities of IPV among sexual minority people emerged as well. Specifically, as highlighted in the minority stress model proposed by Meyer (1995, 2003), sexual minority people suffer particular adverse conditions (i.e., experiences of discrimination, perceived stigma, internalized homonegativity, and sexual identity concealment) that affect their individual and relational wellbeing (e.g., Hughes et al., 2022; Pachankis et al., 2021; Pachankis et al., 2018), and which seem to increase the risk to suffer or perpetrate IPV (Edwards et al., 2015; Rollè et al., 2018).

In addition, sexual minority people are affected by some specific forms of abuse: threats of outing to significant others and homonegative attitudes expressed toward the partner emerged as specific abusive tactics acted out by sexual minority persons (Badenes-Ribera et al., 2016).

Furthermore, the help-seeking process within this population is influenced by unique complexities. According to several authors (Calton et al., 2016; Cannon & Buttell, 2015; Chong et al., 2013; Ollen et al., 2017; Rollè et al., 2021), the heteronormative and homonegative climate that still permeates our societies limits the opportunity of understanding, recognizing, and managing this phenomenon. The lack of services tailored to this population and the ineffectiveness of formal sources of support have been extensively documented (Freeland et al., 2018; Lorenzetti et al., 2017; Rollè et al., 2021; Santoniccolo et al., 2021). This negatively influences the possibilities of sexual minority people who are victims or perpetrators of IPV to find help and recover from this experience.

Given similarities and differences between IPV in sexual minorities and heterosexual people, and the negative consequences this phenomenon has on victims’ physical (e.g., injuries, risk of suicidality) and psychological (e.g., symptoms of depression, anxiety, and stress) wellbeing (Bartholomew et al., 2008a, b; Robinson, 2002; Strickler & Drew, 2015), understanding what variables are associated with the perpetration of IPV among sexual minority people can provide important information for clinical purposes.

Accordingly, the current paper aims to review and systematize the scientific literature focused on the exploration of factors associated to the perpetration of IPV among sexual minority people. Many studies have highlighted the lack of interventions tailored to sexual minority people who experience IPV as well as the ineffectiveness of mainstream formal sources of support, partly due by the lack of knowledge about LGBT+-related themes and specificities of IPV among sexual minorities people (see Santoniccolo et al., 2021 for a review on this topic). The implementation of policies and services capable of addressing the complexities and specificities experienced by sexual minority people involved in couple violence is still needed (Subirana-Malaret et al., 2019). Data obtained in the current review can provide empirical evidence in this direction, providing an exhaustive summary of the current knowledge on the phenomenon, which can guide the development of future prevention and intervention programs addressed to sexual minority people who perpetrate couple violence. Furthermore, the current paper aims to highlight limitations and gaps of the current literature and provide insights for future research.

Materials and Methods

Data Source and Search Strategy

The current systematic review followed the Preferred Reporting Items for Systematic Review and Meta-Analyses (PRISMA) statement (Moher et al., 2009; Page et al., 2021). Two independent reviewers (TT and LR) conducted a systematic search through EBSCO (Databases: APA Psycinfo; CINAHL Complete; Family Studies Abstracts; Gender Studies Database; Race Relations Abstracts; Social Sciences Abstracts [H.W. Wilson]; Sociology Source Ultimate; Violence & Abuse Abstracts) and PubMed. No temporal limits were imposed on the search. All the articles published from the beginning of the databases to July 2022 were screened.

The following keywords were applied: violence or abuse or aggression or batter* AND partner or couple* or domestic or intimate or dating AND “same-sex” or “same-gender” or gay or lesbian* or bisex* or lgb* or homosexual* or “m*n who ha* sex with m*n” or msm or “wom*n who ha* sex with wom*n” or wsw or “m*n who ha* sex with m*n and wom*n” or msmw or “wom*n who ha* sex with wom*n and m*n” or wswm or “sexual minorit*” or “m*n who love m*n” or “wom*n who love wom*n”.

Inclusion and Exclusion Criteria

The following inclusion criteria were applied to select the studies: (a) original research papers, (b) published in peer-review journals, (c) in the English language, (d) focused on the assessment of factors associated with the perpetration of IPV among sexual minority people (i.e., self-identified LGB+ people, people sexually or romantically attracted to people of the same-sex, people involved in same-sex relationship or people that reported non-heterosexual sexual behaviors); (e) only quantitative studies were eligible for the inclusion.

All the studies that did not match the inclusion criteria reported above were excluded. In addition, the following exclusion criteria were applied: (a) studies pertaining to IPV whose methods or results did not clearly differentiate between IPV among sexual minority people and heterosexual people; (b) validation studies, meta-analyses and literature reviews; (c) qualitative studies; (d) papers focused only on factors associated to IPV victimization among sexual minority people; (e) papers that assessed factors associated with any form of IPV (regardless of victim or perpetrator status) among sexual minority people which, however, did not differentiate between variables related to perpetration and those related to victimization. These studies were excluded because they do not provide clear information on factors associated with the perpetration of IPV, and thus do not provide data which can guide the development of interventions targeted to perpetrators. Finally, (f) articles mainly focused on transpeople or self-identified heterosexual people perpetrators of IPV were excluded. However, some of the studies included in the current systematic review involved small percentages of gender minorities or self-identified heterosexual people that based on their sexual behaviors or romantic attraction were classified as sexual minority people. These studies were retained because, from our perspective, they still provide data that can inform on factors related to IPV perpetration among cisgender sexual minority people, which was the population of our interest.

Study Selection and Data Extraction

The search through EBSCO returned 5956 articles, and 4028 papers were left after duplicates removal. Of these, 414 papers were selected for full-text review after the screening of title and abstract, and 73 papers were included. PubMed provided 2174 articles in total. After the screening of title and abstract, 216 papers were selected for full-text review. The removal of duplicates between databases left 143 articles and five were included. In total, 78 articles were included in the current systematic review after full-text reading and the application of inclusion and exclusion criteria.

Two independent reviewers analyzed the full-text and proceeded with the data extraction. Any disagreement was discussed between the reviewers in order to obtain a unanimous consensus. See Fig. 1 for a summary of the study selection procedure.

Fig. 1
figure 1

Flowchart of the selection procedure

Full size image

Results

Seventy-eight studies were included in the current systematic review and most of them (54 studies) were conducted in the USA. Two studies were conducted in Canada, and two in the USA and Canada. Nine studies were conducted in Europe: three in Italy, two in England, one in Germany, one in Turkey and one in Spain; one study was in Turkey and Denmark. Five studies were conducted in China, one in Spanish-speaking countries (Spain, Mexico, Chile, and Venezuela), one in Latin American countries (mainly in Mexico), one in Puerto Rico, one in South Africa, and one in Hong Kong. Finally, one study assessed factors associated with IPV perpetration in the USA, Canada, Australia, UK, South Africa, and Brazil (see Table 2).

The last two decades have seen an increase in the number of articles published on the topic of our interest. Specifically, while three studies were published between 1988 and 1999, and 15 between 2000 and 2010, most were published between 2011 and 2021 (60 studies; see Table 2).

Methodological Issues

Several methodological differences between the included studies emerged, which must be accounted for when considering the obtained results.

First, differences in the operationalization of violence and in the used instruments were found. Thirty-nine studies assessed IPV perpetration with some version of the conflict tactics scale (CTS; Ayhan Balik & Bilgin, 2021; Bartholomew et al., 2008a, b; Causby et al., 1995; Balsam & Szymanski, 2005; Chong et al., 2013; Craft & Serovich, 2005; Craft et al., 2008; Edwards & Sylaska, 2013; Edwards et al., 2021; Leone et al., 2022; Lewis et al., 2017; Li et al., 2019, 2022; Li & Zheng, 2021; Gabbay & Lafontaine, 2017a; Gabbay & Lafontaine, 2017b; Jones & Raghavan, 2012; Kelley et al., 2014; Landolt & Dutton, 1997; Lewis et al., 2018; Mason et al., 2016; McKenry et al., 2006; Milletich et al., 2014; Oringher & Samuelson, 2011; Pepper & Sand, 2015; Pistella et al., 2022; Stephenson et al., 2011b; Stults et al., 2021b; Swan et al., 2021; Swann et al., 2021; Taylor & Neppl, 2020; Tognasso et al., 2022; Whitton et al., 2021; Wu et al., 2015; Stults et al., 2015b; Stults et al., 2016; Waterman et al., 1989; Whitton et al., 2019; Zavala, 2017). Other used assessment tools were the Intimate Partner Violence Among Gay and Bisexual Men (IPV-GBM) Scale (10 studies; Davis et al., 2016; Jones & Raghavan, 2012; Sharma et al., 2021; Stephenson & Finneran, 2016; Stephenson & Finneran, 2017; Suarez et al., 2018; Wei et al., 2020a, b, 2021; Zhu et al., 2021); the Psychological Maltreatment of Women Inventory (three studies; Bartholomew et al., 2008a, b; Leone et al., 2022; Lewis et al., 2018) or its short form (one study; Lewis et al., 2017); the 30-item Abusive Behavior Inventory (ABI; two studies; Telesco, 2003; Turell et al., 2018); the Multidimensional Measure of Emotional Abuse (two studies; Fontanesi et al., 2020; Ummak et al., 2021); the Cold Violence Scale (one study; Li & Zheng, 2021); the SGM-Specific IPV Tactics Scale (one study; Swann et al., 2021); the Conflict in Adolescent Dating and Relationships Inventory (one study; Reuter et al., 2015); the Relational Pursuit-Pursuer Short Form (one study; Derlega et al., 2011); the Sexual Coercion in Intimate Relationships Scale (one study; Fontanesi et al., 2020); the Psychological Maltreatment Inventory (one study; Landolt & Dutton, 1997); the Sexual Experiences Survey (one study; Krahé et al., 2000); the Perpetration in Dating Relationships (PDR; one study; Jacobson et al., 2015); The Safe Dates—Psychological Abuse Perpetration scale (one study; Jacobson et al., 2015); the 19 items of the Psychological Abuse in Intimate Partner Violence Scale adapted to be used with same-sex couples (EAPA-P; Longares et al., 2018a; one study; Longares et al., 2018a); the Fundamental Interpersonal Relations Orientation-Behavior (FIRO-B; one study; Poorman & Seelau, 2001); The Coercive Control Scale (one study; Whitton et al., 2019); SGM-Specific IPV Tactics Scale (one study; Whitton et al., 2019); the Cyber Abuse Scale (one study; Whitton et al., 2019); an adaptation of the scale developed by Smith et al., (1995; one study; Wong et al., 2010). In contrast, 19 studies (Bacchus et al., 2017; Bartholomew et al., 2008a, b; Bogart et al., 2005; Carvalho et al., 2011; Finneran & Stephenson, 2014; Finneran et al., 2012; Kelly et al., 2011; Li & Zheng, 2021; Longares et al., 2018b; Messinger et al., 2021; Miltz et al., 2019; Reuter et al., 2015; Schilit et al., 1991; Stephenson et al., 2011a, 2013; Stults et al., 2015a, 2021a; Toro-Alfonso & Rodríguez-Madera, 2004; Waterman et al., 1989) used items developed by authors to assess at least some forms of IPV perpetration (see Table 2 for more details).

Second, only 15 studies (Balsam & Szymanski, 2005; Davis et al., 2016; Edwards et al., 2021; Fortunata & Kohn, 2003; Jones & Raghavan, 2012; Sharma et al., 2021; Stephenson & Finneran, 2016; Stephenson & Finneran, 2017; Suarez et al., 2018; Swann et al., 2021; Wei et al., 2020a, b; Whitton et al., 2019; Whitton et al., 2021; Zhu et al., 2021) included instruments or items that addressed abusive tactics specific to sexual minority people (e.g., threats of outing, homonegativity, and negative HIV-related attitudes). Furthermore, only slightly more than half of the studies were focused on the variables associated to the violence perpetrated within a same-sex relationship (40 studies; Ayhan Balik & Bilgin, 2021; Balsam & Szymanski, 2005; Bartholomew et al., 2008a, b; Bartholomew et al., 2008a, b; Carvalho et al., 2011; Chong et al., 2013; Craft & Serovich, 2005; Davis et al., 2016; Edwards & Sylaska, 2013; Fontanesi et al., 2020; Gabbay & Lafontaine, 2017a; Gabbay & Lafontaine, b; Jacobson et al., 2015; Jones & Raghavan, 2012; Kahle et al., 2020; Kelley et al., 2014; Landolt & Dutton, 1997; Leone et al., 2022; Lewis et al., 2017; Lewis et al., 2018; Li et al., 2019, 2022; Li & Zheng, 2021; Longares et al., 2018b; Mason et al., 2016; Milletich et al., 2014; Pepper & Sand, 2015; Pistella et al., 2022; Poorman & Seelau, 2001; Schilit et al., 1990; Sharma et al., 2021; Stephenson & Finneran, 2016; Stephenson et al., 2011a; Stephenson et al., 2011a, b; Stephenson et al., 2013; Suarez et al., 2018; Telesco, 2003; Tognasso et al., 2022; Wu et al., 2015), while the remaining articles did not specify in what kind of relationship the violence occurred (i.e., if within a relationship with a same-sex or opposite-sex partner).

Third, differences in the characteristics of the involved population emerged. Only one study used probabilistic methods of sampling (Bogart et al., 2005). Most (39 studies) included self-identified LGB+ people (Ayhan Balik & Bilgin, 2021; Bacchus et al., 2017; Bartholomew et al., 2008a, b; Bartholomew et al., 2008a, b; Bogart et al., 2005; Chong et al., 2013; Derlega et al., 2011; Edwards et al., 2021; Finneran & Stephenson, 2014; Finneran et al., 2012; Fontanesi et al., 2020; Fortunata & Kohn, 2003; Gabbay & Lafontaine, 2017a; Gabbay & Lafontaine, b; Jacobson et al., 2015; Kelly et al., 2011; Lewis et al., 2017; Lewis et al., 2018; Longares et al., 2018a; Longares et al., b; Mason et al., 2016; Oringher & Samuelson, 2011; Pistella et al., 2022; Poorman & Seelau, 2001; Reuter et al., 2015; Schilit et al., 1990; Sharma et al., 2021; Stephenson & Finneran, 2016; Stephenson & Finneran, 2017; Stephenson et al., 2011a; Stephenson et al., 2013; Swan et al., 2021; Taylor & Neppl, 2020; Telesco, 2003; Tognasso et al., 2022; Toro-Alfonso & Rodríguez-Madera, 2004; Turell et al., 2018; Ummak et al., 2021; Zavala, 2017). Regardless of their self-identified sexual orientation, 17 studies recruited participants on the basis of their involvement in a same-sex relationship (Balsam & Szymanski, 2005; Craft & Serovich, 2005; Craft et al., 2008; Edwards & Sylaska, 2013; Jones & Raghavan, 2012; Kahle et al., 2020; Kelley et al., 2014; Leone et al., 2022; Li et al., 2019, 2022; Li & Zheng, 2021; Milletich et al., 2014; Pepper & Sand, 2015; Stephenson et al., 2011a, b; Suarez et al., 2018; Waterman et al., 1989; Wu et al., 2015), while twelve articles were focused on people who reported non-heterosexual sexual behaviors (Davis et al., 2016; Krahé et al., 2000; Miltz et al., 2019; Stults et al., 2015a, b, 2016, 2021a, b; Wei et al., 2020a, b, 2021; Zhu et al., 2021). In five studies, participants were recruited based on one of the criteria reported above (i.e., sexual orientation, same-sex relationship, sexual behaviors; Messinger et al., 2021; Swann et al., 2021; Whitton et al., 2019, 2021; Wong et al., 2010). Five articles did not specify based on what criteria they defined their participants as sexual minority people (Carvalho et al., 2011; Causby et al., 1995; McKenry et al., 2006; Landolt & Dutton, 1997; Schilit et al., 1991). In terms of age, education and ethnicity of the enrolled participants, most of the studies (54 studies) included in the current systematic review were focused on adult participants, mainly White and well-educated (Ayhan Balik & Bilgin, 2021; Bacchus et al., 2017; Balsam & Szymanski, 2005; Bartholomew et al., 2008a, b; Bartholomew et al., 2008a, b; Bogart et al., 2005; Carvalho et al., 2011; Chong et al., 2013; Derlega et al., 2011; Edwards & Sylaska, 2013; Edwards et al., 2021; Finneran et al., 2012; Fontanesi et al., 2020; Fortunata & Kohn, 2003; Gabbay & Lafontaine, 2017a; Gabbay & Lafontaine, b; Jacobson et al., 2015; Kahle et al., 2020; Kelley et al., 2014; Kelly et al., 2011; Krahé et al., 2000; Landolt & Dutton, 1997; Leone et al., 2022; Lewis et al., 2017; Lewis et al., 2018; Li & Zheng, 2021; Li et al., 2022; Longares et al., 2018a; Longares et al., b; Mason et al., 2016; McKenry et al., 2006; Milletich et al., 2014; Miltz et al., 2019; Oringher & Samuelson, 2011; Pepper & Sand, 2015; Pistella et al., 2022; Schilit et al., 1990, 1991; Sharma et al., 2021; Stephenson & Finneran, 2017; Stephenson et al., 2011a; Stephenson et al., 2011a, b; Suarez et al., 2018; Swan et al., 2021; Taylor & Neppl, 2020; Tognasso et al., 2022; Toro-Alfonso & Rodríguez-Madera, 2004; Turell et al., 2018; Ummak et al., 2021; Waterman et al., 1989; Wei et al, 2020a, b, 2021; Zhu et al., 2021). Only five studies involved adolescents (Poorman & Seelau, 2001; Reuter et al., 2015; Stults et al., 2015b, 2016; Whitton et al., 2019), three were mainly focused on HIV-positive participants (Bogart et al., 2005; Craft & Serovich, 2005; Wu et al., 2015), and in 20 articles, ethnic minorities or ethnically diverse people were the population of interest (Craft & Serovich, 2005; Craft et al., 2008; Davis et al., 2016; Finneran & Stephenson, 2014; Jones & Raghavan, 2012; Li et al., 2019, Messinger et al., 2021; Reuter et al., 2015; Stephenson & Finneran, 2016; Stephenson et al., 2013; Stults et al., 2015a, b, 2016, 2021a, b; Swann et al., 2021; Whitton et al., 2021; Wong et al., 2010; Wu et al., 2015; Zavala, 2017). The study by Causby et al. (1995) did not specify the characteristics of the included participants.

Finally, while the characteristic of both partners can influence and being influenced by IPV perpetration, only eleven studies used dyadic techniques of data analysis (Landolt & Dutton, 1997; Leone et al., 2022; Lewis et al., 2018; Li et al., 2019, 2022; Schilit et al., 1990; Sharma et al., 2021; Stephenson et al., 2011a, b; Stephenson et al., 2013; Suarez et al., 2018; Turell et al., 2018), and only two adopted a longitudinal design (Lewis et al., 2018; Stults et al., 2021a), while the remaining 76 articles were cross-sectional in nature.

Main Findings

The following classes of variables (see Table 1 for a definition of each class of variables) were identified as factors associated to the perpetration of IPV among sexual minority people: sociodemographic factors; psychological factors; relational factors; social and community-level factors; feminine and masculine gender expression; intimate partner violence; family of origin-related factors; substance use; medical conditions; sexual behaviors; sexual minorities-specific factors (see Table 2 for a summary of the results found within the studies included and their methodological characteristics).

Table 1 Definition of the IPV-associated factors analyzed in the results section
Full size table
Table 2 Characteristics of the included studies
Full size table

Sociodemographic Factors

Gender Differences

No gender differences in perpetration of IPV were found in eight studies (Chong et al., 2013; Craft et al., 2008; Fontanesi et al., 2020; Gabbay & Lafontaine, 2017b; Longares et al., 2018b; Pistella et al., 2022; Reuter et al., 2015; Waterman et al., 1989). In contrast, in the study by Derlega et al., (2011), the results showed that men who were rejected after the breakup of a relationship engaged in more pursuit behaviors (i.e., invasive and annoying, but not necessarily threatening behaviors) than women did, though gender differences were not found considering aggressive behaviors (i.e., invasive and threatening behaviors).

Age Suarez et al. (2018) identified a negative association between participants’ age and IPV perpetration. Similarly, Turell et al. (2018) identified a negative relation between partner’s age and IPV perpetration. In contrast, two studies showed a positive association between age and couple violence perpetration (Messinger et al., 2021; Zavala, 2017). In Finneran et al. (2012), participants aged between 25 and 40 years old were at increased risk of perpetration of sexual IPV only in the USA, while no significant association between age and IPV perpetration was found in Canada, Australia, UK, Brazil, and South Africa. The remaining articles that assessed the association between IPV perpetration and participants’ age (Balsam & Szymanski, 2005; Bartholomew et al., 2008a, b; Chong et al., 2013; Finneran & Stephenson, 2014; Pistella et al., 2022; Stephenson et al., 2011a; Stephenson et al., 2011a, b; Turell et al., 2018; Whitton et al., 2019; Wei et al., 2021) or age differences between the partners (Stephenson et al., 2011a, b; Turell et al., 2018) did not find significant results.

Gender Identity No differences between participants’ gender identities (i.e., cisgender vs. gender minorities) in the perpetration of IPV were found in two studies (Turell et al., 2018; Whitton et al., 2019). In contrast, in Stults et al., (2021a, b), transgender participants reported a higher injury perpetration prevalence compared to cisgender participants, while this relation was not significant when considering physical, psychological, and sexual IPV prevalence. Minor sexual IPV chronicity was instead more common among cisgender participants than transgender ones in this study (Stults et al., 2021a, b). No differences emerged in relation to the other assessed forms of IPV. In the study by Whitton et al. (2019), compared to participants with cisgender female partners, those with gender minority partners were at increased risk for perpetrating coercive control and abusive tactics specific of sexual and gender minorities.

Sexual Orientation Bisexual people were at increased risk of IPV perpetration compared to homosexual people in four studies (Reuter et al., 2015; Stults et al., 2021a, b; Ummak et al., 2021; Wei et al., 2020a, b). In the study by Reuter et al. (2015), the linear regression model further showed that sexual orientation was only related to severe teen dating violence (TDV; i.e., a combination of physical and sexual IPV) perpetration, while it was not related to TDV when any TDV score was considered as a dependent variable. Finally, Turell et al. (2018) found an association between having a bisexual partner and IPV perpetration. In contrast, sexual orientation was not associated with IPV perpetration in five studies (Chong et al., 2013; Finneran & Stephenson, 2014; Miltz et al., 2019; Wei et al., 2021; Whitton et al., 2019).

Ethnicity Non-White participants were found to be at increased risk of IPV perpetration in three studies (Messinger et al., 2021; Stephenson et al., 2011a; Zavala, 2017), and Black/African American and indigenous participants reported higher rates of IPV perpetration in the study by Turell et al. (2018). Similarly, Black and Latin participants were at increased risk of IPV perpetration compared to White participants in the study by Whitton et al. (2019), while in a more recent study by Whitton et al. (2021), only Black, and not Latin participants, were at increased risk of IPV perpetration compared to White participants. In contrast, Latin participants were at increased risk of IPV perpetration compared with Black and White participants in the study by Stults et al. (2021a). In a more recent study by Stults et al., (2021a, b), Asian participants had higher levels of minor sexual IPV perpetration prevalence compared to White participants (and not compared to Latin, Black, or multi-ethnic participants), while this relation was not significant in relation to physical, psychological, and severe sexual IPV perpetration prevalence. There were no differences in IPV perpetration prevalence between White, Latin, and Black participants. On the other hand, IPV perpetration chronicity was more common among White and Black participants than among Asian participants in this study (Stults et al., b). Caucasian participants were more likely to report physical and emotional, but not sexual, IPV perpetration than African American participants in one study (Wong et al., 2010). Turkish participants were at increased risk of IPV perpetration compared with Danish participants in Ummak et al. (2021). Six studies did not identify a significant association between ethnicity and perpetration of IPV (Finneran & Stephenson, 2014; Finneran et al., 2012; Fortunata & Kohn, 2003; Miltz et al., 2019; Stephenson et al., 2011a, b; Wei et al., 2021).

Education Less educated people were at increased risk of IPV perpetration in 9 studies (Balsam & Szymanski, 2005; Bartholomew et al., 2008a, b; Chong et al., 2013; Finneran & Stephenson, 2014; Milletich et al., 2014; Miltz et al., 2019; Stephenson et al., 2011a; Stephenson et al., 2011a, b; Stults et al., 2021a, b). However, in the study by Bartholomew et al., (2008a, b), the association between physical and psychological IPV and education was not significant when controlling for the effect of IPV victimization (i.e., bidirectionality of abuse). In addition, the negative association between education and lifetime physical and sexual IPV perpetration that emerged in the study by Balsam and Szymanski (2005) was not significant when recent IPV was considered as the dependent variable. Finneran and Stephenson (2014) found that only physical IPV perpetration was negatively associated with education, while sexual IPV was not. In contrast, less educated people were at increased risk of perpetrating sexual and emotional, but not physical, abuse in the study by Stephenson et al., (2011a, b). In the study by Finneran et al. (2012), participants with more than 12 years of education were at increased risk of IPV perpetration only in Canada, while this relation was not significant in the USA, Australia, UK, Brazil, or South Africa. Education was not related to IPV perpetration in one study (Wei et al., 2021).

Income Income was negatively associated with physical, but not sexual, IPV perpetration in one study (Chong et al., 2013). Similarly, in the study by Bartholomew et al., (2008a, b), income was negatively associated with physical, but not emotional, IPV perpetration. However, this relation was no longer significant when controlling for IPV victimization. Fortunata and Kohn (2003) found that batterers’ partners had lower income than non-batterers’ partners, while income was not associated with IPV perpetration in Balsam and Szymanski (2005). Finally, income was negatively related to severe injury and severe sexual IPV perpetration prevalence, and positively related to minor psychological IPV prevalence and minor sexual IPV perpetration chronicity in Stults et al., (2021a, b). Income was not associated to minor injury, minor sexual, severe psychological, and physical IPV perpetration prevalence, and to chronicity of injury, physical, psychological, and severe sexual IPV perpetration in this study (Stults et al., 2021a, b).

Employment None of the studies that assessed the association between employment and IPV perpetration highlighted significant results (Finneran & Stephenson, 2014; Fortunata & Kohn, 2003; Miltz et al., 2019; Wei et al., 2021).

Psychological Factors

Mental Health (General Mental Health; Emotion Regulation; Depression, Anxiety, and PTSD; Loneliness; Stress; Suicidality; COVID-19) In Wei et al., (2020a, b), general mental health was associated to any, emotional, controlling, and monitoring IPV, while these results were not supported in relation to physical and sexual IPV. Cognitive reappraisal was associated with IPV perpetration in one study (Wei et al., 2020a, b), while expressive suppression was not.

Depression was positively associated with IPV perpetration among sexual minority people in six studies (Miltz et al., 2019; Sharma et al., 2021; Stults et al., 2015a, 2021a; Wei et al., 2020a, b; Zavala, 2017). However, in the research by Sharma et al. (2021), the association was not significant among couples who stipulated a sexual agreement. In the study by Stults et al. (2015a), this relation was no longer significant when controlling for childhood maltreatment. Bacchus et al. (2017) did not find a significant association between depression and IPV perpetration.

In addition, a marginal association between symptoms of mild anxiety disorder and any negative behaviors in the past 12 months (i.e., any abusive behaviors, which include physical abuse, frightening, forcing sex, and controlling behaviors perpetrated in the past 12 months) was found in Bacchus et al. (2017). Symptoms of mild anxiety disorder were not associated with physical abuse, frightening, forcing sex, or controlling behaviors in this study (Bacchus et al., 2017).

PTSD was positively associated with IPV perpetration at a bivariate level in one study (Stults et al., 2015a); however, this relation was not significant in the regression model when controlling for childhood maltreatment. Also, Stults et al. (2021a) did not find a significant association between PTSD and IPV perpetration.

In the study by McKenry et al. (2006), non-perpetrating females reported higher psychological adjustment compared with non-perpetrating males, and male and female perpetrators. A positive association between psychological maladjustment and psychological IPV perpetration was confirmed in Pepper and Sand (2015), although psychological maladjustment was not related with sexual or physical IPV in this study.

Stults et al. (2015a) identified a positive association between IPV perpetration and loneliness at a bivariate level. However, this relation was no longer significant in the regression analysis when controlling for childhood maltreatment. In the study by McKenry et al. (2006), IPV perpetrators experienced more family stress than non-perpetrators, and the relation between perceived stress and IPV perpetration was fully mediated by insecure attachment in the study by Craft et al. (2008). Similarly, in Whitton et al. (2021) economic stress was associated to physical, sexual, and severe psychological IPV perpetration, while this association was not significant when considering minor psychological IPV. Any IPV perpetration and controlling behavior were positively related to suicidality in one study (Wei et al., 2020a, b). These results were not supported in relation to physical, emotional, controlling, and sexual IPV perpetration.

Finally, Pistella et al. (2022) did not highlight an association between COVID-19 psychosocial impact and IPV perpetration.

Personality Traits An association between at least some personality traits and IPV perpetration was found in seven studies (Chong et al., 2013; Fortunata & Kohn, 2003; Landolt & Dutton, 1997; Pepper & Sand, 2015; Reuter et al., 2015; Stults et al., 2015a; Zavala, 2017). Specifically, the results found by Landolt and Dutton (1997) showed that both actor’s and partner’s abusive personality (i.e., constituted by borderline personality organization [BPO], anger, fearful attachment, preoccupied attachment, maternal rejection and paternal rejection) was associated with physical and psychological IPV perpetration. More specifically, each constituent of the abusive personality of both the actor and the partner were associated with psychological IPV perpetration. For physical IPV perpetration both actor and partner effects were significant for BPO, fearful, and preoccupied attachment, while neither actor nor partner effects were significant for anger and paternal rejection, and only actor effects were significant for maternal rejection (Landolt & Dutton, 1997).

In the study by Fortunata and Kohn (2003), a relation between personality traits and IPV perpetration was confirmed: batterers had higher scores on the aggressive (sadistic), antisocial, avoidant, passive-aggressive, self-defeating, borderline, paranoid, and schizotypal personality scale scores and higher alcohol-dependent, drug-dependent, bipolar (manic syndrome), and delusional clinical syndrome scale scores. However, no significant differences between batterers and non-batterers emerged in the scores on compulsive, dependent, depressive, histrionic, narcissistic, schizoid, anxiety, dysthymia, PTSD, somatoform, major depression, and thought disorders scales. When controlling for desirability and debasement, group differences for the avoidant, bipolar (manic syndrome), dependent, passive-aggressive, schizoid, schizotypal, and self-defeating personality were no longer significant (Fortunata & Kohn, 2003).

In addition, hostility was positively associated with perpetration of IPV in two studies (Pepper & Sand, 2015; Reuter et al., 2015), as was impulsivity in the study by Stults et al. (2015a), even after controlling for the effect of childhood maltreatment. However, impulsivity was not associated with IPV perpetration in Stults et al. (2021a). Self-control (Zavala, 2017) and anger management (Chong et al., 2013) were found to be negatively associated to IPV perpetration. However, in the study by Chong et al. (2013), the relation between physical IPV perpetration and anger management was fully mediated by psychological IPV perpetration. Emotional instability was positively related with physical and psychological IPV perpetration, but not with sexual IPV, in the study by Pepper and Sand (2015). Furthermore, these authors found a significant association between negative worldview and psychological IPV perpetration, while this relation was not significant when physical or sexual IPV were considered as dependent variables (Pepper & Sand, 2015). Emotional unresponsiveness was not associated to IPV perpetration (Pepper & Sand, 2015).

Self-Esteem and Self-Efficacy Self-esteem was negatively associated to IPV perpetration in four studies (Causby et al., 1995, McKenry et al., 2006; Wei et al., 2020a, b; Wei et al., 2021). However, negative self-esteem was not associated to IPV perpetration in Pepper and Sand (2015).

Self-efficacy was negatively associated to emotional IPV perpetration in the study by Wei et al., (2020a, b). In contrast, two studies did not find significant association between physical and sexual IPV perpetration and self-efficacy or self-adequacy (Chong et al., 2013; Pepper & Sand, 2015).

Attachment In the study by Longares et al. (2018a), the results highlight a significant association between insecure adult attachment and psychological IPV. Outness moderated this relation: at low levels of overall outness, the relationship between insecure attachment and psychological IPV was not significant. Similarly, at low and high levels of outness to religion, this association was not significant. Outness to the family did not moderate the association between insecure attachment and psychological IPV (Longares et al., 2018a). According with these findings, in the study by McKenry et al. (2006), perpetrators had a less secure attachment style than non-perpetrators. Fontanesi et al. (2020) identified a negative association between confidence (i.e., a dimension that represents a secure attachment style) and commitment defection and manipulation (i.e., two of the three dimensions of sexual abuse in this study), and, surprisingly, a positive association between confidence and emotional abuse. No significant relationship between confidence and coercion of resources and violence (i.e., the third dimension of sexual abuse) was detected (Fontanesi et al., 2020).

In addition, four studies found a significant association between attachment anxiety and IPV perpetration (Bartholomew et al., 2008a, b; Derlega et al., 2011; Gabbay & Lafontaine, 2017a; Tognasso et al., 2022). The association between attachment anxiety and sexual IPV perpetration was fully mediated by dyadic trust and sexual intimacy in a serial mediation model in the study by Gabbay and Lafontaine (2017a). In the study by Derlega et al. (2011), only the relation between attachment anxiety and pursuit behaviors was significant, while the relationship between attachment anxiety and aggressive behaviors was not. In Tognasso et al. (2022), attachment anxiety was related to any and psychological IPV perpetration, and this relation was partially mediated by internalized homonegativity. Attachment anxiety was not associated to physical and sexual IPV in this study (Tognasso et al., 2022). Furthermore, only attachment anxiety assessed through interview, and not self-reported anxious attachment, was still associated to physical and psychological IPV perpetration in the study by Bartholomew et al., (2008a, b) when controlling for IPV victimization (i.e., bidirectionality of abuse). Gabbay and Lafontaine (2017b) did not identify a significant relation between self-reported attachment anxiety and physical or psychological IPV perpetration.

Four studies highlighted a significant association between attachment avoidance and IPV perpetration (Bartholomew et al., 2008a, b; Gabbay & Lafontaine, 2017a, b; Tognasso et al., 2022). However, in the study by Bartholomew et al., (2008a, b), only attachment avoidance assessed through interview was associated to physical and psychological IPV perpetration (even after controlling for IPV victimization), while self-reported avoidance was not. Attachment avoidance was only associated to physical, and not psychological IPV perpetration in the study by Gabbay and Lafontaine (b), and this relation was no longer significant when controlling for receipt of violence. In Tognasso et al (2022), the association between attachment avoidance and sexual IPV was not significant. However, they highlighted a direct relation between attachment avoidance and physical IPV, and a positive association between attachment avoidance, and any and psychological IPV, partially mediated by internalized homonegativity. Furthermore, the relation between attachment avoidance and sexual IPV perpetration was partially mediated by dyadic trust and sexual intimacy in a serial mediation model in the study by Gabbay and Lafontaine (2017a). Derlega et al. (2011) did not find any association between attachment avoidance and pursuit or aggressive behaviors. In addition, discomfort with closeness, need for approval, and preoccupation with relationships were all related with some dimension of sexual coercion or emotional abuse (see Table 2 for more details) in the study by Fontanesi et al. (2020). The relationship being secondary was not associated to emotional abuse or sexual coercion in this study (Fontanesi et al., 2020).

The proximity dimension of caregiving (and not sensitivity, compulsive caregiving, and controlling caregiving) was negatively associated to physical and psychological IPV perpetration in the research by Gabbay and Lafontaine (2017b), although this relation was not significant when controlling for receipt of violence. In addition, these authors identified a significant association between psychological IPV perpetration and both hyperactivation of the attachment and caregiving systems and deactivation of the attachment and caregiving systems, even in the presence of each other. Regarding physical IPV perpetration, only hyperactivation was still associated to physical couple violence when controlling for the effect of deactivation strategies. None of these findings were significant when receipt of violence was controlled for (Gabbay & Lafontaine, 2017b).

Relational Factors

Couple-Level Demographic Factors Cohabitation was correlated with increased risk of IPV perpetration in the study by Suarez et al. (2018), while cohabitation with a same-sex partner was not associated to physical or psychological IPV perpetration in Chong et al. (2013). Length of relationship was not associated to IPV perpetration in four studies (Chong et al., 2013; Pistella et al., 2022; Sharma et al., 2021; Turell et al., 2018), and neither was having a child in the studies of Fortunata and Kohn (2003) and Turell et al. (2018). Relationship status (i.e., single, married, having boyfriend or other) did not predict IPV perpetration as well (Wei et al., 2021).

Open Relationships, Monogamy, and Infidelity Being in an open relationship and infidelity were both associated to abuse perpetration in one study (Turell et al., 2018). In contrast, batterers and non-batterers did not differ regarding monogamous relationships in Fortunata and Kohn (2003).

Couple Dynamics Three studies found a negative association between dyadic adjustment or relationship satisfaction, and IPV perpetration (Balsam & Szymanski, 2005; Li et al., 2019; Stephenson et al., 2011a, b). However, in the studies by Li et al. (2019) and Stephenson et al., (2011a, b), only psychological IPV was associated to relationship satisfaction, while physical, and physical or sexual IPV respectively were not. In addition, in two studies relationship satisfaction was not associated to IPV perpetration (Derlega et al., 2011; McKenry et al., 2006).

Sexual satisfaction was negatively associated to IPV perpetration in one study (Pistella et al., 2022). In the study by Poorman and Seelau (2001), perpetrators had lower expressed and wanted inclusion, and expressed and wanted affection compared with non-perpetrators. However, expressed and wanted control did not differ between perpetrators and non-perpetrators, and there were no differences between the groups in the differences between expressed and wanted inclusion, expressed and wanted affection or expressed, and wanted control (Poorman & Seelau, 2001). Perpetrators of emotional or physical violence showed lower levels of communal coping, couple efficacy, and couple outcome preferences in the study by Stephenson et al., (2011a, b). In addition, perpetrators of emotional abuse (not those who perpetrated physical or sexual abuse) had lower degree of concordance with the partner lifestyle topics. Perpetrators of sexual violence had lower communal coping scores compared with non-perpetrators, while they did not differ in couple efficacy and couple outcome preferences (Stephenson et al., 2011a, b). Furthermore, higher scores in investment size (i.e., personal investment in the relationship), not in poor quality of alternatives or commitment in relationships, were related with unwanted pursuit (and not with aggressive behaviors) in the study conducted by Derlega et al. (2011).

Finally, while dependence was not related with IPV perpetration in three studies (McKenry et al., 2006; Pepper & Sand, 2015; Telesco, 2003), jealousy (Telesco, 2003) and fusion/intrusiveness (i.e., enmeshment in one’s couple relationship; Causby et al., 1995; Mason et al., 2016; Milletich et al., 2014) were found to be both positively associated to IPV perpetration. In the study by Causby et al. (1995), while share fusion was associated to physical aggression, physical/more severe violence, and psychological violence, time fusion was only associated to physical aggression and psychological violence.

Power Dynamics In the study by Landolt and Dutton (1997), perpetration of psychological IPV by abusers was higher when victims perceived to be in a divided-power couple compared to when victims perceived to be in an egalitarian couple. No other differences regarding psychological IPV perpetration emerged when comparing victims’ perception of being in a divided-power, egalitarian, or self-dominant couple. Perceived power differentials or power imbalances were not associated to IPV perpetration in two studies (McKenry et al., 2006; Telesco, 2003). In addition, in the study by Landolt and Dutton (1997), couples that disagreed in their perception of relationship power dynamics (i.e., non-congruent couples) did not differ from congruent couples in their levels of IPV perpetration. Finally, dominance was positively associated to IPV perpetration in the study by Chong et al. (2013). However, this relation was no longer significant when controlling for demographic variables. Milletich et al. (2014) did not identify a significant relation between dominance/accommodation and IPV perpetration. Nevertheless, these authors found an indirect influence of accommodation on IPV perpetration through fusion: accommodation was positively related with fusion, which in turn was positively associated to IPV perpetration (Milletich et al., 2014).

Conflict, Conflict Resolution Skills, and Communication Relationship conflict was positively associated to physical and psychological IPV in two studies (Chong et al., 2013; Pistella et al., 2022), while having assertiveness abilities reduced the probability to perpetrate sexual coercion in one study (Toro-Alfonso & Rodríguez-Madera, 2004). Finally, Krahé et al. (2000) found a significant association between token resistance (i.e., one of the two dimensions of ambiguous communication during sexual encounters, which describes the tendency to refuse sex when actually it is what one desires) and sexual violence, while the relation between sexual violence and compliance (i.e., the second dimension of ambiguous communication during sexual encounters, which describes the tendency of having sex with someone when one does not want to) was not significant.

Social- and Community-Level Factors

Characteristics of the Social Network and Social Support Perceived instrumental support by family, friends, and colleagues was negatively associated to IPV perpetration in two studies (Wei et al., 2020a, 2020b; Whitton et al., 2021). However, in Whitton et al. (2021), this relation was significant only when considering physical, sexual, and severe psychological IPV perpetration, while it was not supported when considering minor psychological IPV. In contrast, social support was not related with IPV perpetration in three studies (Edwards et al., 2021; Reuter et al., 2015; Zavala, 2017). Similarly, the number of gay friends was not associated to physical IPV perpetration in Stephenson et al. (2011a), while being involved in a male network composed by perpetrators of violence was positively associated to dating or sexual violence only among lesbian women and not among gay men in the study by Jones and Raghavan (2012).

Involvement in LGB Communities and Support Agencies Involvement in social activities within the LGB community (Wei et al., 2020a, b) and involvement in LGB + support agencies (Stults et al., 2015a) were both positively associated to IPV perpetration. Furthermore, in the study by Turell et al. (2018), the analysis of variance showed that bisexual participants involved in local or online bisexual communities were at increased risk of IPV perpetration than those not involved in bisexual communities. However, in the path analysis, involvement in bisexual communities was not associated to IPV perpetration (Turell et al., 2018). Pistella et al. (2022) did not find a significant association between LGB community involvement and IPV perpetration.

Ethnic Discrimination Two studies identified a positive association between ethnic discrimination and IPV perpetration (Swann et al., 2021; Whitton et al., 2021).

Religiosity Pistella et al. (2022) did not find a significant association between religiosity and IPV perpetration.

Feminine and Masculine Gender Expression and Sexism

Masculinity was positively associated to IPV perpetration in three studies (Jacobson et al., 2015; McKenry et al., 2006; Oringher & Samuelson, 2011). However, in Oringher and Samuelson (2011), only some dimensions of masculinity were associated to physical IPV perpetration: suppression of vulnerability and aggressiveness were both positively related to physical IPV perpetration, while avoidance of dependency on other was negatively related with physical IPV perpetration. In contrast, the association between self-destructive achievement and dominance, and physical IPV perpetration was not significant, and no dimensions of masculinity were associated with sexual IPV perpetration (Oringher & Samuelson, 2011). In the studies by Telesco (2003) and Balsam and Szymanski (2005), the relationship between gender expression and IPV perpetration was not significant.

In the study by Li and Zheng (2021), both benevolent or hostile sexism toward women and hostile sexism toward men were positively associated to cold violence perpetration. These associations were not significant when considering any IPV and controlling violence perpetration. Benevolent sexism toward men was not associated to any IPV, cold violence, or controlling behaviors perpetration in this study (Li & Zheng, 2021).

Intimate Partner Violence

In 17 studies, IPV victimization was positively associated to IPV perpetration (Bartholomew et al., 2008a, b; Edwards et al., 2021; Gabbay & Lafontaine, 2017b; Lewis et al., 2017; Li & Zheng, 2021; Longares et al., 2018b; Miltz et al., 2019; Oringher & Samuelson, 2011; Pepper & Sand, 2015; Pistella et al., 2022; Stults et al., 2015a, b, 2021a; Swan et al., 2021; Waterman et al., 1989; Wei et al., 2021). However, while Pepper and Sand (2015) found a significant relation between physical IPV victimization and perpetration, and between psychological IPV victimization and perpetration, they did not find a significant association between sexual IPV victimization and perpetration. In Lewis et al. (2017), while physical violence perpetration and victimization were each other associated, only the association between psychological IPV perpetration and psychological IPV victimization was significant, but the opposite directional path was not. In addition, in the study by Waterman et al. (1989), the association between sexual IPV victimization and perpetration was significant only among sexual minority men, and not among sexual minority women. The association between physical IPV victimization and physical IPV perpetration was significant for both genders in this study (Waterman et al., 1989).

In six studies (Bartholomew et al., 2008a, b; Chong et al., 2013; Finneran & Stephenson, 2014; Wei et al., 2020a, b, 2021), different forms of IPV perpetrated by participants were all significantly associated to each other (see Table 2 for more details). However, Finneran and Stephenson (2014) found a significant association only between sexual and psychological IPV perpetration, and between psychological and physical IPV perpetration, while the association between sexual and physical IPV perpetration was no longer significant in the logistic model when controlling for the effect of other variables.

Family of Origin-Related Factors

Witnessing Violence in the Family of Origin Witnessing IPV between parents (Messinger et al., 2021; Schilit et al., 1991; Whitton et al., 2021) or siblings (Messinger et al., 2021) was positively associated to IPV perpetration in three studies. In Craft and Serovich (2005), only witnessing violence from mother-to-father was associated to sexual coercion perpetration, while witnessing violence from father-to-mother was not. No significant associations were found between witnessing violence (both from mother-to-father and from father-to-mother) and psychological IPV, physical assault, or physical injury perpetration in this study (Craft & Serovich, 2005). Similarly, five other studies did not identify a significant association between these variables (Bartholomew et al., 2008a, b; McKenry et al., 2006; Milletich et al., 2014; Reuter et al., 2015).

Childhood Maltreatment and Harsh Parenting Childhood maltreatment was positively associated to IPV perpetration in six studies (Fortunata & Kohn, 2003; Messinger et al., 2021; Schilit et al., 1991; Stults et al., 2015a; Toro-Alfonso & Rodríguez-Madera, 2004; Whitton et al., 2021). However, only sexual victimization in the family of origin was associated to sexual IPV perpetration in the study by Toro-Alfonso and Rodríguez-Madera (2004), while suffering physical and psychological victimization in the family of origin were not. In addition, in the study by Bartholomew et al., (2008a, b), a positive association was found between IPV perpetration and mother-to-teen violence, while the relation between father-to-teen violence and IPV perpetration was not significant. The association between mother-to-teen violence was no longer significant when controlling for IPV victimization. Similarly, in other studies, childhood maltreatment (Chong et al., 2013; McKenry et al., 2006; Milletich et al., 2014; Stults et al., 2021a) or harsh parenting (Taylor & Neppl, 2020) were not related with IPV perpetration.

There were no differences between abusers and non-abusers in having a family member during childhood who abused substances in the study by Fortunata and Kohn (2003), while perpetrators of IPV grew up in families with a lower socio-economic status (SES) than non-perpetrators in McKenry et al. (2006).

Substance Use

Addictive behaviors (Toro-Alfonso & Rodríguez-Madera, 2004) and substance use (i.e., both alcohol and other drugs use; Chong et al., 2013) were found to be both positively associated to IPV perpetration. However, in the study by Chong et al. (2013), substance use was related only to physical IPV, while the association with psychological IPV was not significant.

Alcohol Use Participants’ (Bartholomew et al., 2008a, b; Davis et al., 2016; Fortunata & Kohn, 2003; Kelley et al., 2014; McKenry et al., 2006; Schilit et al., 1990; Wu et al., 2015) or partner’s (Leone et al., 2022) alcohol use were found to be positively associated to IPV perpetration in eight studies. More specifically, in Davis et al. (2016), alcohol use was associated to physical/sexual and emotional IPV toward both regular and casual partner, and to controlling and HIV-related IPV perpetration toward regular, but not casual, partners. Monitoring IPV perpetration was not associated to alcohol use in this study (Davis et al., 2016). In addition, the relation between alcohol use and IPV perpetration was moderated by outness in Kelley et al. (2014): this association was significant only at high levels of outness. In Bartholomew et al., (2008a, b), the association between alcohol use and IPV perpetration was no longer significant when controlling for IPV victimization (i.e., bidirectionality of abuse). Physical aggression was associated to discrepant drinking between partners at a later time point in Lewis et al. (2018), while it was not related to subsequent physical aggression. Discrepant drinking was associated to subsequent psychological aggression and vice versa in this study (Lewis et al., 2018). Several studies did not find a significant association between alcohol use by the participants (Bacchus et al., 2017; Kelly et al., 2011; Reuter et al., 2015; Sharma et al., 2021; Stults et al., 2015b, 2021a) or their partners (Schilit et al., 1990; Sharma et al., 2021), and IPV perpetration. Alcohol dependence or abuse was associated to IPV in Fortunata and Kohn (2003), while this relationship was not significant in Bacchus et al. (2017).

Drug Use Drug use was related with IPV perpetration in seven studies (Bacchus et al., 2017; Bartholomew et al., 2008a, b; Fortunata & Kohn, 2003; Stults et al., 2015b, 2021a; Wong et al., 2010; Wu et al., 2015). However, in Wu et al. (2015), only methamphetamine use was associated to IPV perpetration, while marijuana, powdered or rock/crack cocaine, or heroin use were not. In addition, in Bacchus et al. (2017), participants who reported frightening and physically hurting their partner were at increased risk of cannabis use compared to those who did not. In contrast, there were no differences in cannabis use between those who perpetrate forcing sex or any abusive behaviors in the past 12 months, or those whose partner needs to ask permission to do activities, and those who did not. Furthermore, physically hurting a partner, but no other forms of abuse, was related with class A drugs (i.e., ecstasy, LSD, cocaine, crack, heroin, and injected amphetamines) use (Bacchus et al., 2017). In Bartholomew et al., (2008a, b), the association between drug use and IPV perpetration was no longer significant when controlling for IPV victimization. Drug use during sex was associated to IPV perpetration in two studies (Miltz et al., 2019; Wei et al., 2020a, b). In contrast, several studies did not identify a significant association between participants’ (Finneran et al., 2012; Kelly et al., 2011; Schilit et al., 1990; Sharma et al., 2021; Wei et al., 2021) or partner’s (Schilit et al., 1990; Sharma et al., 2021) drug use, and IPV perpetration.

Medical Conditions

Participants who reported to be HIV-positive were at increased risk of physical, but not emotional or sexual, IPV perpetration in the study by Stephenson et al., (2011a, b). In contrast, somewhat surprisingly, perpetrators of any abusive behaviors in the past 12 months were at lower risk of having a diagnosis of sexually transmitted infections (STI) than non-perpetrators in the study by Bacchus et al. (2017). Perpetrators of physical abuse, frightening, forcing sex, or controlling behaviors did not differ from those who did not perpetrate these forms of violence in the risk of having an STI diagnosis in this study (Bacchus et al., 2017). HIV status was not related with IPV perpetration in four studies (Bartholomew et al., 2008a, b; Finneran & Stephenson, 2014; Finneran et al., 2012; Stephenson & Finneran, 2016).

Furthermore, participants who perpetrated IPV did not differ from non-perpetrators in thinking their partner would not support their PrEP use or in not knowing if their partner would support their PrEP use, or in their perception of benefits provided by PrEP use in the study by Kahle et al. (2020).

Sexual Behaviors

Years at Anal Sexual Debut and Sexual Partner(s) Miltz et al. (2019) assessed the association between years at anal sexual debut and IPV perpetration, and they did not find significant results. In contrast, an age of 18 or older at sexual debut was positively associated to controlling behaviors and negatively related to emotional IPV in one study (Wei et al., 2020a, b). Wei et al. (2021) supported this latter finding, highlighting a negative association between age at sexual debut and IPV perpetration.

The number of sexual partners was positively associated to IPV perpetration in two studies (Zhu et al., 2021; Wei et al., 2020a, ). In Zhu et al. (2021), this association was moderated by self-efficacy (at high levels of self-efficacy the relation between multiple casual sexual partners, and IPV perpetration was no longer significant), while Miltz et al. (2019) did not find significant results. Behavioral bisexuality was not associated to IPV perpetration in two studies (Stephenson et al., 2011a; Finneran et al., 2012). Having group sex was associated to perpetration of lifetime, but not past year, IPV in one study (Miltz et al., 2019).

Sexual Intercourses Participants who reported two or more instances of anal receptive and insertive sex had a higher risk of perpetrating couple violence compared with those who reported no instances of these behaviors in the study by Stults et al. (2016). Use of lubrification was not associated to physical IPV perpetration in Stephenson et al. (2011a).

In addition, several studies assessed unprotected sex and IPV perpetration (Bogart et al., 2005; Finneran & Stephenson, 2014; Miltz et al., 2019; Stephenson & Finneran, 2017; Stephenson et al., 2011a; Stults et al., 2016). Most of these (five studies) found a relation between these two variables (Bogart et al., 2005; Finneran & Stephenson, 2014; Stephenson & Finneran, 2017; Stephenson et al., 2011a; Stults et al., 2016). However, in the study by Stephenson and Finneran (2017), condomless anal intercourse (CAI) was only associated to physical, sexual, emotional, and controlling IPV, while not to monitoring IPV. Furthermore, in the studies by Finneran and Stephenson (2014) and Stephenson et al. (2011a), perpetrators of physical IPV were more likely to have had unprotected anal intercourse than non-perpetrators of physical IPV. No differences in unprotected anal intercourse emerged between perpetrators of sexual IPV and non-perpetrators of sexual IPV. Furthermore, in Stephenson et al. (2011a), both sexual and physical IPV were higher among participants who have had unprotected anal intercourses (UAI) compared to those who have not, while in Finneran and Stephenson (2014) only for physical IPV the difference between the two groups was significant. Miltz et al. (2019) did not find significant associations between unprotected sex and IPV perpetration.

Inconsistent condom use with regular partner was related to any IPV and controlling violence perpetration in the study by Zhu et al. (2021), while this relation was not significant when considering physical, emotional, and sexual IPV. Inconsistent condom use with casual partners was instead only associated to sexual IPV (Zhu et al., 2021). Finally, experiencing transactional sex was associated to IPV perpetration in one study (Wei et al., 2021).

Sexual Minority-Specific Factors

The frequency of minority stressors experienced was not associated to pursuit behaviors and perpetration of negative behaviors after the breakup of the couple relationship in the study by Derlega et al. (2011).

Experiences of Discrimination Experiencing microaggressions (Taylor & Neppl, 2020) or homophobic discriminations or violence (Balsam & Szymanski, 2005; Finneran & Stephenson, 2014; Li et al., 2022; Swan et al., 2021; Swann et al., 2021; Whitton et al., 2021; Zavala, 2017) were found to be positively associated to IPV perpetration. The relation between microaggressions and IPV perpetration was moderated by sexual orientation (i.e., having a bisexual orientation increased the strength of the association between microaggressions and IPV perpetration) in the study by Taylor and Neppl (2020). Similarly, a moderating effect of commitment was found in the relation between homophobic discrimination and IPV perpetration: only at low levels of commitment in the relation this association remained significant. Furthermore, in Finneran and Stephenson (2014), the relationship between homophobic discrimination and IPV perpetration was no longer significant in the logistic model (only in the ANOVA test the differences between sexual batterers and non-batterers were significant). In the study by Balsam and Szymanski (2005), only lifetime discrimination was associated to psychological and physical/sexual (not LGB-specific abuse) IPV perpetration, while past-year discrimination was not. In Whitton et al., (2021), homophobic violence was associated only to psychological, and not physical or sexual IPV. Experiences of discrimination were not found to be associated to IPV perpetration in five studies (Ayhan Balik & Bilgin, 2021; Edwards & Sylaska, 2013; Finneran et al., 2012; Stults et al., 2021a; Zavala, 2017).

Perceived Stigma Perceived stigma was positively associated to IPV perpetration in four studies (Carvalho et al., 2011; Stults et al., 2015a, 2021a; Wei et al., 2020a, b). However, Stults et al. (2015a) found a positive association only between personal-local stigma and IPV perpetration, while the relation between public-gay related stigma and IPV was not significant. Furthermore, somewhat surprisingly, Stephenson et al., (2011a, b) found a negative association between sexual IPV perpetration and perceived local stigma-couple (i.e., perceived stigma around being in a same-sex relationship), but not with perceived local stigma-individual (i.e., perceived stigma around being a gay or bisexual man). No significant associations between perceived local stigma and physical or emotional IPV perpetration were found in this study (Stephenson et al., 2011a, b). Similarly, four studies did not identify associations between these variables (Edwards et al., 2021; Finneran et al., 2012; Stephenson et al., 2011a; Zavala, 2017); although in Edwards et al. (2021), this relation became significant at high levels of problem drinking, while it was not significant at low levels, social support did not moderate this relation.

Internalized Homonegativity Toward Self and Others Participants’ (Stephenson & Finneran, 2016) and partner’s (Turell et al., 2018) homo- or bi-negativity were found to be positively associated to IPV perpetration. McKenry et al. (2006) did not identify a significant association between IPV perpetration and family of origin’s homonegativity.

In addition, participants’ (Ayhan Balik & Bilgin, 2021; Balsam & Szymanski, 2005; Bartholomew et al., 2008a, b; Edwards & Sylaska, 2013; Finneran & Stephenson, 2014; Finneran et al., 2012; Kelley et al., 2014; Li et al., 2019, 2022; McKenry et al., 2006; Miltz et al., 2019; Pepper & Sand, 2015; Stephenson & Finneran, 2016; Suarez et al., 2018; Tognasso et al., 2022; Ummak et al., 2021; Zavala, 2017) and partner’s (Li et al., 2019) internalized homonegativity were found to be associated to IPV perpetration. However, in the study by Balsam and Szymanski (2005), internalized homonegativity was not related to psychological IPV and LGB-specific abuse, and the association between internalized homonegativity and physical/sexual violence was fully mediated by dyadic adjustment. Furthermore, sexual coercion perpetration was associated only with the religious attitudes toward lesbianism dimension of the Lesbian Internalized Homonegativity Scale (LIHS; Szymanski & Chung, 2001) in the study by Pepper and Sand (2015), while it was not related with any other dimension of the LIHS. Internalized homonegativity was not related with physical and emotional IPV perpetration in this study (Pepper & Sand, 2015) as in Ayhan Balik and Bilgin (2021), where only sexual IPV perpetration was positively associated to internalized homonegativity. In Finneran and Stephenson (2014), internalized homonegativity was associated to sexual IPV, but not with physical IPV perpetration, while Edwards and Sylaska (2013) found a significant relation between internalized homonegativity and physical and sexual IPV perpetration, but not between internalized homonegativity and psychological IPV. Participants’ and partner’s internalized homonegativity were only associated to psychological IPV in the study by Li et al. (2019), while no significant results were found when physical IPV perpetration was considered as the dependent variable. Similarly, Tognasso et al. (2022) identified a positive association between internalized homonegativity and any and psychological IPV perpetration, while the relation between internalized homonegativity and physical and sexual violence was not significant. In Li et al. (2022), the association between internalized homonegativity and partner’s psychological IPV perpetration (not participants’ psychological IPV nor participants’ and partner’s physical IPV) was moderated by commitment in the relationship: at high levels of commitment the relation became not significant. In this study (Li et al., 2022), commitment mediated the relation between internalized homonegativity and participants’ and partner’s physical and psychological IPV perpetration as well. In Finneran et al. (2012), internalized homonegativity was positively associated to sexual IPV perpetration only in the UK, while this relation was not significant in the USA, Canada, Australia, Brazil, or South Africa. Similarly, several studies did not find significant associations between participants’ (Carvalho et al., 2011; Chong et al., 2013; Edwards et al., 2021; Pistella et al., 2022; Whitton et al., 2021) and partner’s internalized homonegativity (Suarez et al., 2018) and IPV perpetration. In addition, in the study by Milletich et al. (2014), internalized homonegativity was not directly related with IPV perpetration. However, these authors found a positive indirect association between these variables that was mediated by fusion (Milletich et al., 2014).

Sexual Identity Concealment Outness was positively related to IPV perpetration in two studies (Ayhan Balik & Bilgin, 2021; Longares et al., 2018a). However, only overall outness, and not outness to religion and outness to family, were positively related with psychological IPV perpetration in Longares et al. (2018a). In Bartholomew et al., (2008a, b), outness was positively related with IPV perpetration when controlling for internalized homonegativity, though this relation became non-significant when controlling for both internalized homonegativity and violence receipt (i.e., bidirectionality of abuse). In contrast, Kelley et al. (2014) found lower levels of outness among IPV perpetrators compared with non-perpetrators. Outness was not related to IPV perpetration in four studies (Balsam & Szymanski, 2005; Carvalho et al., 2011; Edwards et al., 2021; Miltz et al., 2019).

Gay Identity Development Gay identity development was not related to IPV perpetration in the study by Stephenson et al. (2011a).

Finally, two studies elaborated conceptual models to understand the mechanisms through which minority stress contributes to IPV perpetration. Lewis et al. (2017) found a complex relation between discrimination, internalized homonegativity, perpetrator trait anger, perpetrator’s and partner’s alcohol problems, perpetrator’s relationship dissatisfaction, and psychological and physical violence. Similarly, in Mason et al. (2016), a complex relation between general life stress, distal and proximal minority stressors, negative affect, hazardous alcohol use, intrusiveness, and physical IPV perpetration was detected.

Discussion

The current paper aimed to review and systematize the available literature on IPV perpetration among sexual minority people and its associated factors. Seventy-eight studies were included in the systematic review.

Several variables were found to be related with IPV perpetration among sexual minority people, and differences and similarities were found between IPV among heterosexuals and sexual minority people. Most of the assessed socio-demographic variables seem to not influence IPV perpetration in most of the included articles. Specifically, age, gender, gender identity, employment, and income were generally found to be unrelated to IPV perpetration. In contrast, when looking at differences across sexual orientations, bisexual people were at increased risk of IPV perpetration in several studies (Bermea et al., 2018; National Intimate Partner and Sexual Violence Survey, 2010). This result further underscores the double stigma associated to bisexual identity. The structural violence and the discrimination that seem to be conveyed by both the heterosexual and the lesbian and gay communities create additional stress and negative affect that can impact individual and relational wellbeing, ultimately leading to the perpetration of couple violence (Turell et al., 2018).

Many psychological factors were found to be related to IPV perpetration among sexual minority people. Depression was found to be related to IPV perpetration in several studies (Miltz et al., 2019; Sharma et al., 2021; Zavala, 2017). Two different theoretical perspectives can explain these findings. On the one hand, symptoms of depression can negatively influence coping and affect regulation mechanisms, which in turn can reduce the ability to manage conflicts and increase the likelihood of using violence toward the partner (Miltz et al., 2019). Accordingly, several studies identified an association between depression and relationship quality in both heterosexual (e.g., Morgan et al., 2018; Roberson et al., 2018) and sexual minority couples (e.g., Vencill et al., 2018; Whitton & Kuryluk, 2014). On the other hand, depression can be considered a consequence of IPV perpetration due to the psychological impact that this experience can entail (Sharma et al., 2021). Similarly, stress perception was associated to IPV perpetration in two studies (Craft et al., 2008; McKenry et al., 2006). Both individual and family stress can impact psychological wellbeing and produce negative affect that needs to be released even through violent behaviors (Zavala, 2017). In contrast, symptoms of anxiety, PTSD, and loneliness were generally unrelated with IPV perpetration (Bacchus et al., 2017; Stults et al., 2015a). Furthermore, several personality traits were found to be associated to IPV perpetration among sexual minority people. In particular, an abusive personality (Landolt & Dutton, 1997), hostility, emotional instability, and a negative worldview (Pepper & Sand, 2015) as well as higher scores on the aggressive (sadistic), antisocial, avoidant, passive-aggressive, self-defeating, borderline, paranoid, and schizotypal personality scale and on the alcohol-dependent, drug-dependent, bipolar (manic syndrome), and delusional clinical syndrome scales (Fortunata & Kohn, 2003) were associated to couple violence perpetration. Although only few studies assessed the association between these variables and other research are needed, these preliminary findings seem to equate those emerged among heterosexual couples (Brasfield, 2014; Brem et al., 2018; Gildner et al., 2021; Spencer et al., 2019), highlighting the need to consider personality traits in clinical settings. Self-esteem was negatively associated to IPV perpetration as well. These results are in line with the disempowerment theory of couple violence (Archer, 1994). According to this perspective, feelings of inadequacy and unworthiness, as well as lack of self-esteem can promote the use of violence to exert control over a partner who is perceived as threatening or who reveals their insecurities (Archer, 1994; McKenry et al., 2006). Between psychological factors, adult attachment seems to take a main role in predicting IPV perpetration among sexual minority people. Attachment theory conceived family violence as the result of dysfunctional strategies of distance and affect regulation (Bartholomew & Allison, 2006; Bowlby, 1984; Fonagy, 1999). Accordingly, attachment anxiety was found to be related to IPV perpetration in several studies (Bartholomew et al., 2008a, b; Derlega et al., 2011; Gabbay & Lafontaine, 2017a). High levels of attachment anxiety entail fears of rejection and loss, which can result in violence toward the partner as a form of exaggerated protest for their unmet attachment needs, driven by the use of strategies of hyperactivation of the attachment system (Bartholomew et al., 2008a, b; Gabbay & Lafontaine, 2017a). In contrast, conflicting results emerged regarding the association between attachment avoidance and IPV (Bartholomew et al., 2008a, b; Derlega et al., 2011; Gabbay & Lafontaine, 2017a, b). Although people with high levels of attachment avoidance can rely on IPV as a means of avoiding closeness and rejection (Gabbay & Lafontaine, 2017a), other studies are needed to confirm these hypotheses.

In addition to psychological factors, several relationship-level variables have also been found associated to IPV perpetration among sexual minority people. Specifically, while couple-level demographic factors (i.e., cohabitation, length of relationship, and having a child) were generally unrelated to IPV (Chong et al., 2013; Fortunata & Kohn, 2003; Sharma et al., 2021; Turell et al., 2018), relationship satisfaction/dyadic adjustment, conflict resolution skills, jealousy, and fusion/intrusiveness were associated to IPV perpetration in several studies (Balsam & Szymanski, 2005; Causby et al., 1995; Li et al., 2019; Mason et al., 2016; Milletich et al., 2014; Stephenson et al., 2011a, b; Telesco, 2003). These findings are in line with the model proposed by Bartholomew and Cobb (2011) to explain heterosexual IPV. As stated by the authors, regardless of personal dispositions to couple violence, those involved in mutually satisfying relationships, characterized by dyadic trust and a positive communication, are at lower risk of experiencing IPV. This theoretical perspective underlines the main role of stress within the couple as a predictive factor for IPV. Furthermore, considering the results found in several studies, lack of boundaries within the relationship (Causby et al., 1995; Mason et al., 2016; Milletich et al., 2014) and high levels of jealousy (Telesco, 2003) can promote IPV as well. In particular, people with high levels of enmeshment in their relationship can resort to abusive behaviors in order to restore a lost sense of oneness in the relationship following a partner’s attempt at separation, or conversely, to create a self-other distance when individuation and separateness are threatened. These data further highlight the role of dysfunctional mechanisms of interpersonal distance regulation in IPV perpetration (Bartholomew & Allison, 2006; Bartle & Rosen, 1994; Bowlby, 1984). In contrast, power dynamics within the couple seem to be unrelated to IPV perpetration among sexual minority people (Chong et al., 2013; Milletich et al., 2014).

Conflicting results emerged regarding the association between IPV perpetration and social- and community-level factors. Involvement in the LGBT community and support agencies was found to be positively related to IPV perpetration among sexual minority people (Wei et al., 2020a, b). Sexual minority people involved in the LGBT community have a greater likelihood to engage in social interactions, which in turn can increase the probability to perpetrate violence toward a romantic or sexual partner (Wei et al., 2020a, b). However, social support was generally unrelated to IPV perpetration (Reuter et al., 2015; Zavala, 2017). These findings are in contrast with several results emerged in studies conducted on heterosexual IPV (Gerino et al., 2018; Okuda et al., 2015; Richards & Branch, 2012), and further highlight the need to consider differences and similarities between these phenomena.

In line with these considerations, conflicting results emerged regarding the association between feminine and masculine gender expression, and IPV perpetration. While two studies identified a positive association between masculinity and IPV perpetration (Jacobson et al., 2015; McKenry et al., 2006), two other studies did not highlight significant results (Balsam & Szymanski, 2005; Telesco, 2003). The lack of significant results found in Telesco (2003) and Balsam and Szymanski (2005) demonstrated that the theory most commonly used to explain couple violence among heterosexual people, which conceives IPV as the result of endorsing a traditional masculinity which legitimizes the use of violence toward a subordinate partner (who exhibits feminine traits; Balsam & Szymanski, 2005; Telesco, 2003), may not be applicable within the LGB+ population (Balsam & Szymanski, 2005). However, these results could be influenced by methodological limitations (e.g., in Balsam and Szymanski (2005), only one item was used to assess masculinity and femininity), and other studies are necessary to understand the association between gender expression and gender role stereotypes, and the perpetration of IPV among sexual minority people.

Suffering violence in the family of origin was often found to be positively related to IPV perpetration among sexual minority people (Fortunata & Kohn, 2003; Schilit et al., 1991; Stults et al., 2015a; Toro-Alfonso & Rodríguez-Madera, 2004), although other studies did not confirm these results (Chong et al., 2013; McKenry et al., 2006; Milletich et al., 2014). From a psychoanalytic perspective, experiences of violence in the family of origin can result in feelings of unworthiness and in a lack of emotion regulation abilities, which can contribute to the use of violence within the relationship (Miltz et al., 2019). Furthermore, direct and indirect experiences of violence within the family of origin can serve as a model for conflict resolution that will be applied in future relationships (Zavala, 2017), according to the social learning theory (Felson & Lane, 2009; Gover et al., 2008; Mihalic & Elliott, 1997). However, considering the studies included in the current review, primarily suffering violence within the family of origin, rather than witnessing parental violence, emerged as a risk factor for IPV perpetration. Although further data are needed to confirm these findings, most of the data available to date suggest that only direct experiences of violence in the family of origin contribute to IPV in adulthood (Bartholomew et al., 2008a, b; McKenry et al., 2006; Milletich et al., 2014; Reuter et al., 2015).

Paralleling findings of studies conducted in heterosexual couples (see Cafferky et al., 2018 for a meta-analytic review), several articles included in the current systematic review identified a positive association between substance use and IPV perpetration among sexual minority people (Bacchus et al., 2017; Bartholomew et al., 2008a, b; Davis et al., 2016; Fortunata & Kohn, 2003; Kelley et al., 2014; McKenry et al., 2006; Miltz et al., 2019; Schilit et al., 1990; Stults et al., 2015b; Wei et al., 2020a, 2020b; Wong et al., 2010; Wu et al., 2015). As with other factors associated to IPV perpetration (e.g., mental health and couple satisfaction), two different theoretical perspectives can explain these findings. On the one hand, the psychoactive effect of drug and alcohol use and its neurological and psychological consequences can increase the risk of using violence to manage conflicts and stress within the relationship (Wei et al., 2020a, b). On the other hand, substance use can be conceived as a consequence of IPV perpetration. From this perspective, perpetrators of IPV can use substances to cope with the negative feelings related to the experience of couple violence (Lewis et al., 2018). A reciprocal relationship between these variables can exist as well (Lewis et al., 2018) and longitudinal studies are needed to confirm these hypotheses.

Furthermore, several studies found a positive association between sexual behaviors and IPV perpetration. Specifically, unprotected sex was related to IPV perpetration in five studies (Bogart et al., 2005; Finneran & Stephenson, 2014; Stephenson & Finneran, 2017; Stephenson et al., 2011a; Stults et al., 2016). As stated by several authors (Stephenson & Finneran, 2017; Stults et al., 2016), abusers can endorse a more stereotypical masculinity that promote impulsivity and hypersexuality, which can result in at-risk sexual behaviors. In contrast, medical conditions such as HIV-positive status do not seem to be associated to IPV perpetration in several studies (Bartholomew et al., 2008a, b; Finneran & Stephenson, 2014; Stephenson & Finneran, 2016).

While similarities emerged between IPV in heterosexual and sexual minority couples, as highlighted through the current systematic review, several studies identified factors specifically associated to couple violence among sexual minority people. These seem to be mainly related to the adverse conditions experienced by sexual minority people. Several dimensions of the minority stress model elaborated by Meyer (1995, 2003) were found to be associated to IPV perpetration among sexual minority people. In particular, while conflicting results emerged regarding the relation between IPV perpetration and experiences of discrimination, perceived stigma, and sexual identity concealment (e.g., Balsam & Szymanski, 2005; Bartholomew et al., 2008a, b; Carvalho et al., 2011; Edwards & Sylaska, 2013; Kelley et al., 2014; Longares et al., 2018a; Miltz et al., 2019; Taylor & Neppl, 2020; Wei et al., 2020a, b), internalized homonegativity was generally found to be associated to IPV (e.g., Balsam & Szymanski, 2005; Bartholomew et al., 2008a, b; Edwards & Sylaska, 2013; Kelley et al., 2014; Li et al., 2019; Miltz et al., 2019). Although it is likely that sexual minority people are better able to cope with distal minority stressors (Balsam & Szymanski, 2005), proximal minority stressors and in particular high levels of internalized homonegativity negatively impact self-esteem, self-worth, and self-identity, resulting in internal conflicts, a negative self-image, and feelings of fear and shame (Bartholomew et al., 2008a, b; Frost & Meyer, 2009; Kubicek et al., 2015; Meyer & Dean, 1998; Telesco, 2003). As suggested by several authors (Bartholomew et al., 2008a, b; Byrne, 1996; Cruz & Firestone, 1998), these negative affects toward the self and in particular toward one’s own sexual identity can be projected on the partner, resulting in IPV perpetration in order to destroy those negative parts of the self that have been expelled. In addition, the stress that can be associated to the status of sexual minority, and the accompanying emotional dysregulation (Hatzenbuehler, 2009; Sommantico & Parrello, 2021), seems to be regulated through the body by resorting to violent behaviors toward the partner. Accordingly, exploring the moderating role of mentalization or the mediating effect of emotional regulation abilities in the association between minority stress and IPV perpetration may further shed light on the complex dynamics that shape couple violence among sexual minority people. The application of the psychological mediation framework (Hatzenbuehler, 2009) which highlights the role of emotion regulation in the relation between sexual minorities stressors and wellbeing seems to show promising results in this direction.

Only few studies explored the mechanism through which minority stress influences IPV perpetration. Mediation models have demonstrated a complex relation between minority stress, couple-level variables, negative affect, alcohol problems, and IPV perpetration (Balsam & Szymanski, 2005; Lewis et al., 2017; Mason et al., 2016; Milletich et al., 2014). Specifically, internalized homonegativity seems to negatively affect relationship quality and couple dynamics, and increase negative affect and alcohol problems, increasing the probability to perpetrate IPV. The structural violence experienced by sexual minority people and the lack of social acceptance of non-heterosexual relationships (Balsam & Szymanski, 2005; Frost, 2011) seem to result in lower relationship quality, negative affect, and maladaptive behaviors which in turn promote IPV perpetration. More complex models are needed in order to further confirm these hypotheses.

Limitations and Future Directions

When considering the results found in the current systematic review, several limitations need to be accounted for. First, this is not a meta-analysis, thus no statistical conclusion can be drawn.

Second, only data on the perpetration of IPV have been considered, and results on factors associated to IPV victimization among sexual minority people need to be explored in further reviews.

Third, only quantitative data were considered by design. Exploring results drawn from qualitative studies can provide a broader comprehension of the phenomenon and need to be considered in future studies.

Fourth, the population of our interest consisted of cisgender sexual minority people, while studies mainly conducted on gender minorities were excluded. Future reviews focused on factors associated to IPV among gender minorities are needed.

Finally, only original research papers published in English and indexed in the main psychological databases were included. Exploring results from other kind of sources such as reports from national and international institutions or NGOs, as well as studies published in languages other than English can deepen our understanding of IPV among sexual minorities people. In addition, methodological limitations emerged when considering the studies included in the current systematic review, which need to be considered in future studies to improve our understanding of IPV among sexual minority people.

Differences in the operational definitions of IPV and sexual orientation emerged, which can affect the results found and limit comparability between the studies. Although most of the included articles used validated assessment tools, many others evaluated at least some forms of IPV using items developed by the authors. In addition, only a few studies included questions about LGB+-specific abuse tactics. These methodological limitations negatively influence the opportunity to precisely detect couple violence among sexual minorities and need to be considered in future studies. The development of new tools aimed at assessing IPV among sexual minority people or the adaptation of instruments to date available for their use with this population are recommended.

Furthermore, differences were found between the included articles on criteria to enroll participants in the study. The various groups included under the umbrella-term sexual minority people can experience different forms of violence, and factors associated to IPV perpetration among self-identified LGB+ people can differ from those associated to IPV perpetrators among people who self-identify as heterosexual and report non-heterosexual sexual behaviors. How different definitions of sexual minority influence the results found among this broad population needs to be explored and controlled for in future studies.

Only few studies specified the type of relationship in which the violence occurred. This does not allow for firm conclusion regarding variables specifically associated to IPV in same-sex couples. How the type of relationship moderates or influences the results found within the studies aimed at assessing factors associated to IPV among sexual minority people needs to be considered.

In addition, only few studies used dyadic analysis techniques, and considering the interdependence between partners, future studies are needed to understand how characteristics of both partners can affect the risk of perpetrating IPV.

Moreover, all but one of the included studies have a cross-sectional design, which does not allow for firm conclusions about the causal direction of the associations found within the included studies. While drawing from different theoretical perspectives, many of the identified associated factors (e.g., adult attachment, personality traits, family of origin-related factors, and minority stress) are considered predictors rather than consequences of IPV perpetration. Longitudinal studies are needed to confirm these hypotheses.

In addition, the results of several included studies highlight a strong association between IPV victimization and perpetration, and a high occurrence of mutual violence among sexual minority people (e.g., Bartholomew et al., 2008a, b; Edwards & Sylaska, 2013). For these reasons, future studies need to assess both victimization and perpetration, and control for how they influence each other and the results found.

Only one study employed a cross-cultural design and most of the studies were conducted in the USA. Accordingly, other studies are needed to explore IPV perpetration and its associated factors in other geographic areas to fill these gaps.

Furthermore, only few studies were focused on ethnic minorities or people with a HIV-positive status. Drawing from an intersectional framework (Crenshaw, 1991), future studies should explore how multiple stigmatized dimensions of one’s own personal identity impact the risk of IPV perpetration among sexual minority people.

Finally, more complex models (e.g., mediational, moderation, or structural equation models) are needed to understand the mechanism through which minority stress, and psychological and relational factors are related to IPV perpetration among sexual minority people.

Conclusions

The results of the current systematic review highlight the need to consider couple violence among sexual minority people through a multidimensional approach to account for the multitude of variables associated to IPV perpetration. On the one hand, conflicting results emerged regarding the association between gender expression and IPV perpetration. These findings show that the applicability of theories mainly used to understand IPV among heterosexual couples has not yet been demonstrated when considering IPV among sexual minority people. Other studies are needed to understand the role of adhering to traditional gender roles on IPV perpetration among sexual minority people.

On the other hand, the main role of psychological, relational, and LGB+-specific factors emerged in many of the included studies. Specifically, internal working models and adult attachment style, as well as high levels of stress, couple dissatisfaction and fusion within the relationship seem to play a major role in the perpetration of IPV among sexual minority people. These findings highlight the impact of negative affect, and dysfunctional mechanisms of interpersonal distance and affect regulation, and are in line with the conceptualization of couple violence provided by attachment theory (Bartholomew & Allison, 2006; Bowlby, 1984; Fonagy, 1999) and Bartle and Rosen (1994), which consider IPV as the result of dysfunctional strategies of self-other distance and affect regulation. These theoretical backgrounds allow for the overcoming of a gender-based conception of couple violence, fostering an understanding of violent phenomena beyond those typically perpetrated by men toward women in heterosexual relationships. This approach can guarantee the legitimacy of couple violence perpetrated and suffered by sexual minority people and enables the understanding of this complex phenomenon regardless of the gender or sexual orientation of the people involved.

In addition, the significant association found between minority stressors and IPV perpetration in many studies underlines the necessity to consider the structural violence experienced by sexual minority people and the stress that it entails as a possible explanation for the high levels of IPV identified in this population. In particular, internalized homonegativity and the negative affect it evokes need to be addressed in clinical settings, and prevention programs aimed at reducing social homonegativity and sexual stigma are needed to promote sexual minorities’ individual and relational wellbeing.

The identified results in the current systematic review highlight the importance of appropriate screening processes, able to identify variables that contribute to IPV perpetration for each single case. This allows referral to care-providers who are better suited to address the specific involved factors. A multidimensional approach able to consider the multitude of variables associated to IPV perpetration is necessary to prevent violent behaviors and promote the treatment of perpetrators, with the final aim to reduce relapses. The role of psychological and LGB+-specific factors, as well as relationship dynamics need to be considered for clinical purposes, to reduce IPV perpetration among sexual minority people.

Training of stakeholders working with couple violence or sexual minority people is needed to increase professional skills in dealing with IPV among sexual minority people and increase access to services, which is still limited by lack of awareness regarding this phenomenon and perceived stigma (Santoniccolo et al., 2021). The development of services and interventions based on empirical evidence, addressed to sexual minority perpetrators of IPV and able to take care of these complexities while adopting non-stigmatizing attitudes, is needed as well. The emerging results in the current systematic review can provide an updated guide to develop policies in this direction.