Intimate partner sexual violence and violent victimisation among women living homeless in Madrid (Spain)

The purpose of this paper is to discover the extent to which women living homeless in Madrid (Spain) have been victims of intimate partner sexual violence (IPSV), analysing the connection between stressful life events (SLE) relating to violent victimisation and having suffered sexual assault by a partner. This study took place using a sample of women living homeless in Madrid ( n = 136). The results obtained show that a high percentage of the women living homeless who were questioned had been victims of IPSV. The women interviewed, who had been victims of IPSV, had suffered, to a greater extent, SLEs relating to situations of violent victimisation, both during childhood and adolescence as well as in adulthood. The episodes of violent victimisation that seem to be the most common forerunners to IPSV are incidents of sexual abuse in childhood and adolescence.


| INTRODUCTION
The United Nations' Sustainable Development Goal 1 states to end poverty in all its forms everywhere. People living homeless throughout the world are one of the most extreme manifestations of the global phenomenon of poverty and social exclusion Vázquez, Suarez, et al., 2019). There are complex factors that contribute to the phenomenon of homelessness, both individually and socially (Busch-Geertsema et al., 2010;Esen, 2017;Fazel et al., 2014;Vázquez et al., 2018) and these are the responses to a range of life experiences and processes (Matulic-Domandzic et al., 2019). Of those living homeless, women are, in certain aspects, especially vulnerable (Rodríguez-Moreno, Roca, et al., 2020;, with a particularly high risk of suffering violent victimisation (Vet et al., 2019).
Women tend to be underrepresented in research on people living homeless and issues of gender in this collective, as a whole, are underresearched (Mayock & Bretherton, 2016). There has been little research, specifically focusing on women without a home , mainly due to the difficulty in accessing large enough sample groups of women in this situation. Unfortunately, women have been invisible in both popular and academic portrayals of homelessness (Rodriguez-Moreno, Roca, et al., 2020). In the view of Matulic-Domandzic et al. (2019), certain forms of exclusion associated with the specific characteristics of women tend to remain hidden or linked to the private sphere. The reality for people living homeless is complex and multicausal and, according to Herrero (2003), the family environment of women in this situation and violence suffered at the hands of their partners lead to particular vulnerability.
Some research suggests that women living homeless suffer multiple and serious experiences of violence (including physical, psychological and sexual abuse) from an early age; they largely suffer physical and sexual assaults from partners and others (Browne, 1993;Guillén et al., 2020;Rodríguez-Moreno et al., 2020b;Rodríguez-Moreno, Roca, et al., 2020;Vázquez, Suarez, et al., 2019;Wenzel et al., 2000). Specifically, sexual violence has been defined as any unwanted sexual act through coercion, in any setting and regardless of the relationship between aggressor and victim (Krug et al., 2002). Victimisation often precedes the situation of homelessness, increasing the risk of suffering other forms of poly-victimisation as a result of the vulnerability women face when they arrive on the streets (Taylor, 2008). Women living homeless also show a significant deterioration in their overall health (Rodríguez-Moreno et al., 2020b), which in certain cases could be the result of intense exposure to psychological, physical and sexual abuse over the course of their lives (Tinland et al., 2018). Roodman and Clum (2001) estimate that between 15% and 79% of women in the general population had suffered some form of victimisation from an early age, with a correlation between experiences of victimisation in childhood and revictimisation in adulthood (Ecott et al., 2020). According to the theory of the intergenerational transmission of violence (Crombach & Bambonyé, 2015;, repeated exposure to situations of abuse and violence in the family environment would have a negative influence on minors, who would learn to exercise and/or tolerate violence, even considering violence in everyday relationships to be normal (Greene et al., 2018). Finkelhor (1999) notes that within those suffering sexual assault in childhood, the suffering of multiple stressful life events (SLEs), high frequency and duration of sexual abuse and the presence of a relationship to the abuser are factors that can predict a poor evolution of dysfunctional behaviours in the adult years. In this regard, the consequences can manifest in adulthood through symptoms of depression and anxiety, sexual dysfunction, difficulty in establishing intimate relationships, mistrust and/or substance abuse (Sarasua et al., 2013). Young et al. (2017) found that sexual assault in childhood can double the likelihood of suffering physical assault in adulthood, the triple the likelihood of suffering sexual assault.
Violent victimisation thus becomes a key predictive market for the subsequent suffering of violence. In the opinion of Gagnier and Collin-Vézina (2016), sexual abuse as a child is particularly serious in this regard, as it often involves long-term situations due to the asymmetrical relationship between victim and aggressor. To this should be added the low rate of reporting due to the difficulty involved in naming perpetrators (Stiller & Hellmann, 2017) and the fact that child sex abuse often involves multiple aggressors, although little research has been conducted on this issue (Da Silva & Woodhams, 2019).
Research on women living homeless has largely focused on studying behaviours and lifestyles at the time of the interview and there are very few studies on early childhood experiences (Stein et al., 2002), despite the fact that adverse experiences experienced at early ages appear to have long-term negative consequences (Green et al., 2012). Anderson et al. (1988) reported that around 80% of women living in shelters reported having suffered child abuse and that they presented high levels of sexual assault (55%), physical assault (67%) and emotional violence (90%). Later research (e.g., Wong et al., 2016) has found that over 70% of women living homeless suffered abuse in childhood and that 78% had experienced violence in adulthood. These adverse experiences seem to have a significant impact, making it necessary to address these circumstances with a view to meeting specific needs.
It is estimated that 33% of women in the European Union have experienced sexual violence at least once in their lifetime (European Union Agency for Fundamental Rights, 2014). In Spain, the Macroencuesta de violencia contra la mujer 2019 (Macrosurvey of Violence against Women 2019) estimated that 8.9% of Spanish women over the age of 15 had suffered sexual violence from a current or former partner (Delegación del Gobierno contra la Violencia de Género, 2020). It has been found that 8.8% of women in the United States of America have been raped by their partners (Breiding, 2015), indicating that IPSV is a serious problem across different cultural contexts. Given that sexual violence can exist alongside physical and psychological violence, the health repercussions can become highly dangerous (Alsaker et al., 2011;Jaffe et al., 2017). However, though the prevalence and consequences of physical violence by partners have been widely documented, little is known about IPSV (Pengpid et al., 2018). Pengpid and Peltzer (2014) report that women victims of sexual violence also suffered high levels of physical and emotional violence. They are also at a greater risk of femicide, higher attempted suicide rates and have a greater risk of harmful substance abuse than women who had suffered other types of violence. Given the above, the study of IPSV is of particular relevance (Pengpid et al., 2018).
IPSV is a type of violence with a high incidence around the world (Maniglio, 2009;Roy et al., 2014), which may be affecting women living homeless due to their special vulnerability (Luce et al., 2010). There have been very few studies in Spain on violent victimisation suffered by women living homeless (Herrero, 2003), despite the fact that experiences linked to violence may bear a strong correlation to women's homelessness (Fernández-Rasines & Gámez-Ramos, 2013;Marpsat, 2000). The purpose of this paper is to discover the extent to which women living homeless in Madrid (Spain) have been victims of IPSV, analysing the connection between SLEs relating to violent victimisation and sexual assault by a partner. Other specific objectives arising from the previous one are to know which events related to victimisation have led to an increased risk of suffering IPSV, as well as examining which contexts (family or nonfamily) have been able to influence the suffering of IPSV in a greater extent.

| Participants
The study was carried out with a sample of women in a homeless situation in Madrid, made up of 136 women, of legal age, who had spent the night preceding the interview in a shelter or other care facility for homeless people, on the street or in places not initially designed for sleep (abandoned buildings, underground tunnels, cash points, etc.).
The main characteristics of the sample can be seen in Table 1.
As can be seen in Table 1, the vast majority of those interviewed were of Spanish origin, with a mean age of 45 years. Although 60% were single, 39.9% had been married at some point in their life, although at the time of the interview fewer than 6% were still married. More than 60% had children. It should be noted that the women in the sample had their first child at an early age. 12.5% of the interviewees had not finished primary studies, whereas 18% had finished some type of university studies. The mean length of time of the homeless situation of the participating women was 6 years and the vast majority stated that they had slept in a shelter during the month before the interview.  , created from the review of the instrument, was used. de Brugha and Cragg (1990) and previous studies analysed groups in social exclusion and in contexts of poverty Rivas-Rivero et al., 2020;Vázquez et al., 2015;Vázquez, Suarez, et al., 2019). The instrument consists of 26 items (10 events suffered before the age of 18 and 16 events suffered at some point in life). The interview required dichotomous responses (i.e., yes/no answers) to the list of SLEs. For the present study, SLEs suffered before and after the age of 18, related to violent victimisation, were selected. The instrument also collects the age at which SLEs were first experienced. The items had a dichotomous response regarding the occurrence of such events. In the present work, Cronbach's alpha (α = 0.830) indicates an acceptable level of internal consistency.

| Procedure
Homeless women were accessed in shelters for the homeless and in other care facilities for this group. After explaining to the participants, the object of the research and the treatment that would be given to the data, their informed consent was requested, guaranteeing them that their anonymity would always be respected. Considering the results obtained in previous studies , it was presumed that a relatively high number of interviewees could have a low or very low cultural level and that the number of women in a situation of homelessness in Madrid of foreign origin-with significant difficulties understanding the Spanish language-would be relatively high. For this reason, a structured interview was used to collect information, which made it possible to correct possible problems derived from the interviewees' reading and/or comprehension difficulties. The interviews lasted between 40 and 80 min.

| Data analysis
In the study, with a quantitative approach and ex post facto design, the independent variables (SLEs related to victimisation before and after 18 years of age) were compared with respect to having suffered IPSV.

| RESULTS
Of the 136 women interviewed, 34.6% had suffered IPSV at some point in their lives; 58.8% had not suffered IPSV and 6.6% did not report either way on this matter. Table 2 shows different SLEs related to the violence suffered by the women interviewed and the average age at which they experienced each one for the first time.
As shown in Table 2, around 40% of respondents suffered from abuse and problems with violence in their family environment from a very early age. It is of note that the different forms of psychological, physical and sexual violence suffered by those interviewed from early ages were largely perpetrated by people in their family RIVAS-RIVERO ET AL.
| 5 environment. Table 2 also shows high rates of physical and sexual violence after age 18. One-fifth of those interviewed suffered sexual assault, mainly by persons outside their family or by their partners, on average before the age of 30. Table 3 shows the differences in the SLEs suffered by those interviewed relating to violence before the age of 18 in terms of whether they were victims of IPSV.
As can be seen in Table 3, among women who had been victims of IPSV, a greater incidence of SLEs related to violence in childhood and adolescence was observed, mainly sexual abuse by family members and/or individuals outside the family.
No statistically significant differences were observed between those who were victims of IPSV and those who were not in regard to having suffered psychological abuse ("Physical abuse by someone in the family" and/or "Psychological abuse by someone outside the family") (38.3% vs. 24.4%) (χ 2 = 2.734, p = 0.098) (OR = 1.927; 95% CI = 0.881-4.217) or physical abuse ("Physical abuse by someone in the family" and/or "Physical abuse by someone outside the family") (29.8% vs. 21.8%) (χ 2 = 1.005, p = 0.316) (OR = 1.522; 95% CI = 0.667-3.472). However, a difference was found between those who were victims of IPSV those who were not in relation to having suffered sexual abuse ("Sexual abuse by someone in the family" and/or "Sexual abuse by someone outside the family")  Table 4 shows the differences in the suffering of SLEs related to violence after the age of 18, depending on whether or not they were victims of IPSV. It can be seen that the women interviewed who were victims of IPSV had suffered a larger proportion of physical and sexual assault after the age of 18. A larger proportion of those women who were victims of IPSV had suffered sexual abuse after the age of 18 ("Sexual assault by someone in the family" and/or "Sexual assault by someone outside the family") than those who not suffer IPSV (31.1% vs. 12.8%) (χ 2 = 6.421, p = 0.011) (OR = 3.079; 95% CI = 1.260-7.526). Similarly, those who were not victims of IPSV To determine which variables most accurately predict having suffered IPSV, a binary logistic regression analysis was carried out using the backward step regression method: Wald (Table 5). The predictor variables included in the analysis were the SLEs suffered that were statistically significant and for which there was a strong association with having suffered IPSV. The resulting model to explain the suffering of IPSV allows a correct estimate in 59.6% of cases (χ 2 = 35.292; p = 0.000). On the contrary, the Hosmer and Lemeshow test provides a significance of p = 0.305, which shows an acceptable adjustment of the model. The Cox and Snell R 2 estimates an adjustment value of 0.249 and the Nagelkerke R 2 = 0.339. Therefore, having suffered sexual abuse before and after the age of 18 in different contexts (perpetrated by people inside and outside the family) appears to be a risk factor for becoming a victim of IPSV.

| DISCUSSION
First, 35% of women living homeless interviewed in Madrid stated that they had been the victim of IPSV, four times more than the rate observed among women in the general Spanish population (Government Department against Gender Violence, 2020). To better understand the violence suffered by women living homeless, it is important to look deeper into the different types of violence suffered and the context in which it takes place. Inadequate attention to circumstances such as these can lead to an underestimation of the violence suffered and an incomplete assessment of the violent victimisation these women may be suffering. Different authors (Ecott et al., 2020;Pengpid et al., 2018;Simmons et al., 2015) point out that suffering IPSV is related to subsequent situations of abuse and concurrent suffering of other forms of violence, as well as posing an increased risk for the physical and mental health of its victims (Alsaker et al., 2011;Guillén et al., 2020;Jaffe et al., 2017;Pengpid & Peltzer, 2014;Rodríguez-Moreno et al., 2020a;Sarasua et al., 2013).
A very high percentage of women living homeless interviewed in Madrid had suffered abuse during their childhood and/or adolescence and violence in their family environment. In the opinion of different authors (Fernández-Rasines & Gámez-Ramos, 2013;Guillén et al., 2020;Herrero, 2003;Magnilio, 2019;Marpsat, 2000;Rodríguez-Moreno et al., 2020b;Rodríguez-Moreno, Roca, et al., 2020;Roy et al., 2014;Tinland et al., 2018;Wenzel et al., 2000), violent victimisation suffered by women during childhood and adolescence may be related to subsequent situations of homelessness. In this regard, it is of note that a significant percentage of the women interviewed had been exposed to violence against their mothers from an early age, as well as suffering psychological, physical and/or sexual abuse at the hands of a family member. However, these forms of violence did not occur only in the family environment; a high proportion was also perpetrated by people outside the family.
These notably negative experiences suffered from an early age by the women interviewed (largely before age 10) reveal the exposure of different forms of poly-victimisation, in line with the observation made in other studies that point to a high frequency of these experiences among women living homeless (Simmons et al., 2015, Guillén et al., 2020. In women living homeless in Madrid, exposure to violence by members of the family environment at an early age could also have contributed to their fleeing the abusive environment, leading them along paths towards social exclusion. Among the women living homeless interviewed in Madrid, having suffered a larger number of SLEs is connected to a greater probability of becoming a victim of IPSV. There also appears to be a particularly strong association between having suffered sexual abuse before and after the age of 18, both inside and outside the family environment and becoming a victim of IPSV. It is therefore essential to look at this problem in a collective about which there are few studies of this nature (Fernández-Rasines & Gámez-Ramos, 2013), probing deeper into patterns influencing the suffering of IPSV, an issue as yet understudied (Öberg et al., 2019;Pengpid et al., 2018).
Among women living homeless, the distancing from their family at an early age, a shortage of available social support and the situation of poverty and exclusion that they suffer may contribute to dependence on their partners. This issue, together with a possible greater acceptance of violence as a result of the early experience of violent victimisation (Gagnier & Collin-Vézina, 2016;Greene et al., 2018), could affect the repetition of abusive sexual behaviour in intimate partnerships (Tan et al., 2018). Furthermore, the shortage of social care and disconnection from the family may be keeping women living homeless in a spiral of dependency towards their partner, despite knowing that they are victims of IPSV in that relationship (Luque, 2003).
Echeburúa (2020) notes that there are similarities between child sexual abuse and IPSV. In both cases, the sexual abuse tends to be perpetrated by someone with whom the victim has an affective connection, often including financial and/or emotional dependence (e.g., guardianship and protection within the family, life plans in an intimate relationship…), which can lead to the development of insecure attachment and self-blame (Cantón-Cortés, & Cortés, 2015). In addition, sexual abuse in childhood and suffering of IPSV can have serious mental health consequences (Echeburúa, 2020). This is particularly alarming in the case of women living homeless, who are in an especially vulnerable position and in many cases have suffered both these circumstances. Previous studies have indicated that being a victim of IPSV is associated with a higher level of physical and psychological abuse (Tiwari et al., 2014), a higher risk of femicide, more attempted suicides and the conflictive use of psychoactive substances (Pengpid & Peltzer, 2014;Riley et al., 2014;Roze et al., 2020;Simmons et al., 2015).
Echeburua and Amor (2019) point out that traumatic events affect the physical and psychological integrity of the victim and reduce their ability to cope with the violence in different contexts; this is particularly complex in the case of intimate relationships. In the case of women living homeless, the lack of social care networks and family support, in addition to the suffering of physical and sexual assault as a result of the hostile circumstances of not having permanent accommodation (Herrero, 2003), may increase levels of dependence on a partner. For these women living homeless, the high level and repeated suffering of child sexual abuse, particularly which has taken place in the family environment at an early age, may have led to an increase in their tolerance of such behaviour, even seeing sexual assault as normal in day-to-day relationships (Greene et al., 2018). From the logistic regression analysis carried out, it can be seen that among the women living homeless interviewed in Madrid, the variables that best help predict the suffering of IPSV were having been the victim of sexual abuse before the age of 18-both in the family and elsewhere-as well as experiencing sexual assault after the age of 18 committed by people outside the family environment. In line with the observations of other authors, the history of child sexual abuse among women living homeless may be related to subsequent revictimisation in adulthood (Roodman & Clum, 2001) and the fact that they are victims of IPSV (Heslin et al., 2007;Wong et al., 2016). As already mentioned above, these circumstances could entail significant and negative consequences ( Roze et al., 2020;Simmons et al., 2015;, which may have a significant impact on women sliding into a situation of homelessness and how long they remain there. The limitations of this study include the fact that the criteria for the inclusion of the sample in the research were restrictive (assignment to operational categories 1, 2, and 3 of ETHOS), and there was no random or stratified selection of the interviewed women. Consequently, it has not been possible to guarantee the representativeness of the sample. It should also be noted that the study is limited to Madrid (Spain), so it is difficult to generalise its results to other contexts. Furthermore, this is a cross-sectional study design and, therefore, care should be taken when attempting to establish causal relationships. It should be noted that the frequency and duration of SLEs suffered should be specified, to which it must be added that many of the circumstances addressed occurred in the past and there could be a bias in the responses inherent in retrospective studies, so that interpretation of adverse events may have varied over time (Grant et al., 2003). For future work, it would be interesting to delve into SLEs, the frequency and intensity of sexual violence, and analyse whether sexual assaults or other traumatic episodes linked to victimisation have taken place before or during the homeless situation. However, the study adds relevant information on a group with difficult access and shows a complex problem, including the relationship between sexual violence by the partner and violence throughout the lifetimes of women in a situation of homelessness.
The findings of the present work can serve to make visible this problem, sexual violence in homeless women, on which there are not many previous studies, showing that this could be one of the triggers of the situation of female homelessness and health problems and their state of deterioration. For this reason, public policies must carry out an exhaustive examination not only regarding providing safe accommodation that favours women separating from their abusers, but also intervening in overcoming the traumas and adverse situations that seem to have aggravated their deterioration and that keeps them in a situation of dependence on the situation of homelessness and their partners. Addressing SLEs and normalising violence throughout their lives would reduce the psychological distress and self-blame that these events develop in those who suffer them. In this sense, individual and group intervention could favour their emotional recovery, empower them to break the cycle of violence that seems to have marked their lives and provide other models and patterns based on equality.