Chap 6: Epidemiology of Intimate Partner Violence PDF

# Chapter 6: The Epidemiology of Intimate Partner Violence Heidi Stöckl, Karen Devries, and Charlotte Watts ## Introduction to the Epidemiology of Intimate Partner Violence * Violence against women in its many forms has been recognized as a highly prevalent human rights and public health issue (G...

# Chapter 6: The Epidemiology of Intimate Partner Violence Heidi Stöckl, Karen Devries, and Charlotte Watts ## Introduction to the Epidemiology of Intimate Partner Violence * Violence against women in its many forms has been recognized as a highly prevalent human rights and public health issue (Garcia-Moreno and Stöckl 2009). Over the last 20 years there has been a recognizable increase in the research on the prevalence of violence against women, especially intimate partner violence. * While the number of studies on the prevalence of intimate partner violence and its social costs and long-term effects, such as child maltreatment and economic costs, has increased over the past decade, there has been less research on the health effects of exposures to different forms of intimate partner violence (Krug et al. 2002). * Existing literature nevertheless suggests that intimate partner violence might lead to a wide range of potential health effects, including physical, sexual, reproductive, and mental health problems. * The pathways linking intimate partner violence with adverse health outcomes are known to be both direct, through injuries resulting from the violent acts, and indirect, through increased stress, reduced mobility, and limited access to resources and healthcare. * For example, sustained and acute elevated stress levels, often an immediate and long-term consequence of intimate partner violence, have been linked to cardiovascular disease, hypertension, gastrointestinal disorders, chronic pain, and the development of insulin-dependent diabetes (Miller 1998). * In addition, some women also try to manage the stress and trauma caused by intimate partner violence through the use of alcohol, prescription medication, tobacco, or other drugs (Campbell 2002). * Intimate partner violence is therefore of critical importance, not only in its own right, but also in terms of its long-term implications for the burden of disease. This chapter presents an overview of the body of scientific data on the prevalence of intimate partner violence experienced by women and it's risk and protective factors, providing additional information on intimate partner violence during pregnancy and intimate partner homicide. ## Intimate Partner Violence around the World * The 1993 Declaration on the Elimination of Violence against Women defined violence against women as: > any act of gender-based violence that results in, or is likely to result in, physical, sexual or psychological harm or suffering to women, including threats of such acts, coercion or arbitrary deprivation of liberty, whether occurring in public or in private life (United Nations General Assembly 1993). * Intimate partner violence, as one form of violence against women, has been defined as physical, sexual, emotional, or economic abuse, including controlling behavior committed by an intimate partner. The definition of intimate partner also includes same-sex partners. * While emotional abuse, often defined as being humiliated, insulted, intimidated or threatened and controlling behaviors (including not being allowed to see friends or family) is also known to significantly impact the well-being and health of women (Coker et al. 2000; Jewkes 2010), a lack of agreement on standard definitions and measures means that the evidence for this particular form of intimate partner violence is still patchy. Most of the existing research on intimate partner violence is therefore still focused on physical and sexual intimate partner violence. ## Intimate Partner Violence Prevalence * Intimate partner violence surveys commonly measure physical violence through asking participants if they have ever (lifetime prevalence), or in the last year, experienced at least one act of physical violence. * Acts of physical violence typically listed range from being slapped or pushed to being choked or having a gun, knife, or other weapon used on them. * Acts of sexual violence often capture experiences such as being physically forced to have sexual intercourse to being forced to do something sexual that they found humiliating or degrading. ### Global Prevalence Rates The 2012 Global Burden of Disease study included intimate partner violence as a risk factor for poor health in women (Devries et al. 2013; Lim et al. 2013). Estimates for men were not computed because of a lack of global data about health effects. In brief, to estimate the global prevalence rates for women, a systematic review of the existing literature was conducted; inclusion criteria were representative population-based studies with prevalence estimates for intimate partner violence among ever-partnered women aged 15 years or older; these criteria identified 250 studies. In addition, analyses of four major multi-site surveys that included questions on intimate partner violence were conducted. In total, data from 141 studies in 81 countries were entered into a random effects meta-regression that was fitted to produce prevalence estimates for all Global Burden of Disease regions and age groups, as well as global estimates. These estimates for intimate partner violence were corrected for differences in definitions of violence, time periods of measurement, severity of violence, and whether a study was national or sub-national. The global prevalence of lifetime exposure to physical or sexual intimate partner violence, or both, among all ever-partnered women worldwide, established by the Global Burden of Disease study, was 30.0 per cent (95 per cent CI 27.8 per cent to 32.2 per cent) (Devries et al. 2013). ### Variations by Region * Regional breakdowns of the prevalence of intimate partner violence showed that the prevalence was highest in the WHO (World Health Organization) Southeast Asian, Eastern Mediterranean, and African regions. In these three regions, approximately 37 per cent of ever-partnered women reported that they had experienced intimate partner violence at some point in their life. * In the Americas approximately 30 per cent of women reported lifetime experiences of intimate partner violence, while the prevalence was around 25 per cent in all other regions (see Table 6.1 for details) (Devries et al. 2013). ### Variations by Age A further variation is the prevalence rates of intimate partner violence among ever-partnered women across different age groups. The findings in Table 6.2 show that prevalence rates only vary slightly by age, with young women having nearly the same exposure rates to intimate partner violence as older women (Devries et al. 2013). This suggests that intimate partner violence commonly starts early in women's relationships. ## Intimate Partner Violence during Pregnancy Because of its health consequences for both the mother and the unborn child, research on intimate partner violence in the last decade has paid special attention to the time of pregnancy. Three multisite surveys-the Demographic and Health Survey (DHS Survey) (Devries et al. 2010), the International Violence against Women Survey (Devries et al. 2010), and the WHO Multicountry Study on domestic violence against women (WHO Study) (García-Moreno et al. 2005) investigated the prevalence of physical intimate partner violence in different countries. The three surveys found the lifetime prevalence of physical intimate partner violence during pregnancy to range between 1 per cent in urban Japan to 28 per cent in rural Peru; with the majority of surveys estimating a prevalence between 4 and 12 per cent (Devries et al. 2010; García-Moreno et al. 2005). In these surveys higher prevalence rates emerged in Latin American and African countries, with lower rates reported in the surveyed European and Asian countries (Devries et al. 2010). The WHO Study on domestic violence against women further found that intimate partner violence during pregnancy is often conducted with the clear aim to also harm the unborn baby, as between 23 per cent (rural Tanzania) and up to 52 per cent (rural Peru) of women who experienced physical intimate partner violence during pregnancy reported that their partner punched or kicked them in the abdomen (García-Moreno et al. 2005). In addition to population-based surveys, numerous surveys have been conducted in antenatal care clinics around the world for which no representative data are available. These surveys yield substantially higher rates. For example, a systematic review of antenatal care studies from Africa found prevalence rates of 23-40 per cent for physical, 3-27 per cent for sexual, and 25-49 per cent for emotional, intimate partner violence during pregnancy (Shamu et al. 2011). ## Intimate Partner Homicide Another way of capturing intimate partner violence is to examine its most serious outcome-homicide. The Global Burden of Disease study also estimated the prevalence of intimate partner homicide across the world. In addition to a systematic literature review, it also consisted of a survey of relevant homepages of country statistics offices, Ministries of Justice, Home offices, or Police Headquarters of the 169 WHO listed countries. If no information was found, the respective agencies were contacted via email. In total 227 different studies and statistics were found, capturing 1,122 estimates across 66 countries from 1982 to 2011 (Stöckl et al. 2013). Findings of this study suggest that the overall median percentage of homicides committed by an intimate partner is above 13 per cent, with a median above 38 per cent for female homicides having been perpetrated by an intimate partner while the median proportion for men was 6 per cent. Given the high number of homicides (approximately 20 per cent) for which the victim-offender relationship is not known, it can be assumed that the results presented are conservative and that the true prevalence is much higher (Stöckl et al. 2013). Regional differences emerged as well. One of the most striking differences was the relatively high number of studies on intimate partner homicide in high-income countries and the lack of information, especially on male intimate partner homicide, in other regions of the world. The overall median percentage of intimate partner homicides among murdered women was above 58 per cent in the Southeast Asia region, above 40 per cent in high-income countries and the African region, and above 38 per cent in the Americas. Percentages were lower in the Western Pacific region with 19 per cent, the lower and middle-income European region with 20 per cent, and the Eastern Mediterranean region with 14 per cent (Stöckl et al. 2013). Among male homicides, the overall median percentage of intimate partner homicide was highest in high-income countries with more than 6 per cent, the African region with 4 per cent, and the lower and middle income European Region with more than 3 per cent. In all other regions the median percentages were less than 2 per cent (Stöckl et al. 2013). ## Risk, Protective, and Maintaining Factors This section provides an overview of the evidence for factors associated with intimate partner violence, relying chiefly on risk factor analyses using multi-country surveys of low, middle, and high income settings in Africa, Asia, Europe, Australia, and America. These studies include analyses based on the WHO Study (Abramsky et al. 2011), the DHS Survey (Hindin et al. 2008), the World Safe study (Jeyaseelan et al. 2004), and a macro analysis of more than 50 countries (Kaya and Cook 2010). The risk, protective, and maintaining factors for intimate partner violence are presented by the strength of existing evidence. Risk and protective factors that have been measured with agreed and standardized measures and that received support across a number of settings are presented before factors that are built on less conclusive support. Several important methodological limitations issues have to be taken into account when interpreting these risk and protective factors. As most risk and protective factor analyses on intimate partner violence are based on cross-sectional survey data it is often impossible to distinguish if certain associations are outcomes of, causes of, or merely associated with, intimate partner violence. Also, risk for violence is multi-causal and probabilistic rather than deterministic. Having a certain risk factor therefore only means that a person is more likely to experience intimate partner violence, not that every person with that risk factor will experience intimate partner violence. Apart from examining the influence of individual risk factors for intimate partner violence, more needs to be understood about the influence of risk factors combined within one woman for her risk of intimate partner violence. A distinction also needs to be made between individual risk and protective factors, and population-based factors, as, for example, individual levels of drinking might increase a particular woman's risk of intimate partner violence, while population-based drinking levels might be irrelevant to her specific risk of experiencing intimate partner violence. Unless stated otherwise, the risk and protective factors outlined in this chapter refer to individual and not population-based analyses. Last, while there might be a connection between a risk factor and intimate partner violence, the association may in reality be indirectly through an unmeasured factor associated with both intimate partner violence and the measured risk factor, rather than the direct association of the two. ### Alcohol Men's alcohol use and abuse has been identified as a clear risk factor by a number of studies worldwide, including the WHO Study (Abramsky et al. 2011), the DHS Survey (Hindin et al. 2008), and the World Safe study (Jeyaseelan et al. 2004), as well as two systematic reviews investigating its association (Foran and O'Leary 2008; Gil-Gonzalez et al. 2006). A recent systematic review of longitudinal studies has also shown a relationship between women's alcohol use and intimate partner violence victimization (Devries et al. 2014). Several pathways may explain the connection between alcohol abuse and intimate partner violence. One potential pathway is that alcohol abuse and intimate partner violence are associated because of other factors related to both drinking and intimate partner violence, such as young age or low socioeconomic status. Another potential pathway builds on the knowledge that alcohol abuse is detrimental for relationship quality as it promotes counterproductive argument styles and aggressive or violent responses, in addition to the higher levels of aggression alcohol use causes due to inference with cognitive abilities (DeMaris et al. 2004; Klostermann and Fals-Stewart 2006). ### Childhood Experiences of Violence Another risk factor that emerged in nearly all countries of the WHO Study and the DHS Survey (Abramsky et al. 2011; Hindin et al. 2008), and which has also been verified by several longitudinal studies in high-income countries (Fergusson et al. 2006; Moffitt and Caspi 1999), is women's and their partner's experiences of abuse during childhood. Experiences of childhood abuse can range from corporal punishment, child sexual abuse, to witnessing parental violence. In more than 10 of the sites the WHO multi-country study the risk for intimate partner violence was especially high when both the woman and her partner were abused in childhood (Abramsky et al. 2011). Social learning theory suggests several connections between childhood experiences of abuse and intimate partner violence. One connection is that children model parental behaviour-children who observe their parents dealing with difficult situations through violence on a regular basis are more likely to perceive violence as an effective conflict solving strategy, and they might lack alternative models of conflict solution. Another connection is that experiences of violence disrupt children's attachment process which impairs their ability to distinguish between love and violence (Renner and Slack 2006). Both connections can influence these children's later choice of partners, and they may be more likely to choose violent partners (Tolan et al. 2006). ### Unfavourable Gender and Violence Norms and Attitudes Despite the difficulties in measurement, unfavourable gender attitudes, especially those that entail being supportive of wife beating, emerged as another clear risk factor for intimate partner violence, in the WHO Study (Abramsky et al. 2011), in the DHS Survey (Hindin et al. 2008), and in a representative analysis of eight African countries (Andersson et al. 2007). Underlying explanations for these associations can be derived from feminist and gender role theories, which maintain that the patriarchal nature of society promotes violence against women by socializing women and men into predefined roles. While women are often educated to be nurturing and understanding, men are socialized to be successful and dominant breadwinners (Fernández 2006). Intimate partner violence can occur if men feel threatened in their role by their partner's demands or gains of more equality, and if they feel their authority, masculinity or control is diminished (Jewkes et al. 2003). Ideas of empowerment and liberal behaviour might thereby increase levels of violence for women whose surroundings consider these ideas or behaviour as deviant and punishable, or in relationships where women have a higher status in society than their partner. This implies that until full gender equality is achieved, women who behave traditionally are in less danger of experiencing intimate partner violence, since they behave according to their roles and do not challenge masculine identities (Pallitto and O'Campo 2005). ### Relationship Status and Quality Relationship status and quality is measured by several indicators, the most prominent being marital status, relationship duration, or number of children. While a few studies, including a few sites in the WHO study (Abramsky et al. 2011), found higher rates of intimate partner violence among cohabiting women, other population-based studies established that short relationship duration or relationship dissatisfaction increases women's risk of experiencing intimate partner violence (Karamagi et al. 2006; Stöckl et al. 2011). Social exchange theory explains these associations with its claim that intimate partner violence occurs when partners believe that the costs of violence are lower than the expected rewards, such as increased power and dominance. Potential costs include the reduction in relationship quality and the chance that the abused will retaliate, call the police, or end the relationship (Williams 1992). Abusive partners perceive these costs as smaller if they are sure of their dominant physical, psychological, and economic position in the relationship or if they are less committed to the relationship, marked by cohabiting status or short relationship duration. In addition, it has been argued that cohabiting couples are more likely to have arguments about the boundaries of their relationships, which can lead to stress and trigger violence (Gelles 2007). Children can be both a risk and a protective factor for intimate partner violence. Children might be perceived as a risk factor if they tie women to an abusive partner economically or through the belief that it is important for children to grow up with a father under any circumstance. However, children can also improve relationships by encouraging appropriate behaviour in front of them and by reducing couples' social isolation through connecting parents to schools, neighbours, and broader social networks (Hoffman et al. 1994). ### Education, Employment, and Age As with children, women's education can be both a risk and a protective factor for intimate partner violence. In the WHO study and the DHS survey, women's, and especially both women's and their partner's, higher levels of education versus low or no levels of education were protective against intimate partner violence, as was women's employment and low income (Abramsky et al. 2011; Hindin et al. 2008). Furthermore, countries with high female secondary school enrolment and countries with high female labour force participation in non-agricultural sectors also had lower prevalence rates of last-year physical intimate partner violence (Kaya and Cook 2010). However, there are also some studies, for example a national representative survey study from Germany, which found that high levels of education increased women's risk of experiencing intimate partner violence (Stöckl et al. 2011). These associations are best explained by the resource and the relative resource theories. These theories claim that abusive partners use violence because other resources, including education, employment, job prestige, income, or community standing, are unavailable to them or fail to allow them to achieve dominance and power in their relationships (Goode 1971). The protective aspects of education and employment against intimate partner violence for women include the increased access to wider social networks, information, and support women receive in the pursuit of higher levels of education, and the resultant improved confidence and bargaining position in their relationship. However, in societies with rigid gender roles and in relationships where the woman commands more of these resources than their partners, they might also serve as risk factors for intimate partner violence (McMullan 2007). Closely linked to the argument on education and employment is young age as a risk factor for intimate partner violence, as couples who form a union early are more likely to have early pregnancies and more children, which in turn increases their likelihood to suffer employment instability and financial difficulties (DeMaris et al. 2004). Young age has been found to be associated with intimate partner violence in a number of countries (Abramsky et al. 2011; Hindin et al. 2008). ### Women's Social and Wider Surroundings While there is a clear theoretical explanation for the influence of women's social and wider surroundings for intimate partner violence, there are few tangible measures and empirical studies providing support for them. Factors that are frequently investigated suggest that living in an urban or rural area is significantly associated with intimate partner violence, for example in Uganda and South Africa (Jewkes et al. 2003; Karamagi et al. 2006), and that having a partner who is involved in fights with other men or women's experience of non-partner violence, increase women's risk of intimate partner violence (Abramsky et al. 2011; Stöckl et al. 2011); in Germany, women's social isolation is a risk factor for increased intimate partner violence (Stöckl et al. 2011). Explanations for how intimate partner violence is linked to women's social and wider surroundings are provided by social disorganization theory. This theory claims that neighbourhoods with a positive sense of community and strong social networks have lower rates of intimate partner violence as residents share common values and exert social control upon each other (Almgren 2005). Neighbourhood surveillance and social cohesion not only prevent couples from solving their disputes violently, and provide comfort, encouragement, and financial help to deal with difficult situations, they also reduce stress by lowering crime rates and antisocial behaviour in the neighbourhood (DeMaris et al. 2004). ## Gaps in the Evidence Understanding of intimate partner violence has increased substantially over the last few years, with numerous prevalence studies drawing on population-based surveys covering most of the globe. There are also an increasing number of analyses conducted in different countries that show that many risk and protective factors are similar across the world, while also highlighting important regional and local differences. These studies are important for developing targeted interventions. What is still missing, however, is a deeper understanding of the developmental pathways that connect early negative childhood experiences to women's experiences of intimate partner violence in their later life, especially in respect to how different risk and protective factors interact and combine to increase a woman's risk or resilience to intimate partner violence. While there have been several studies in high-income countries, there is still a lack of longitudinal evidence from middle and low-income countries. Pathways may differ across diverse economic and cultural contexts, where social norms and meanings attached to women's social roles and gendered patterns of behaviour are not necessarily the same as in high-income regions.