Understanding Domestic Violence Among Older Women in Ukraine: A Secondary Analysis Using Gender-Based Violence Screening Data.

This secondary descriptive analysis sought to understand Gender-Based Violence (GBV), with a focus on Domestic Violence (DV), among older women in Ukraine's conflict setting. Analysis was conducted on a subsample of 150 women aged 60+ from GBV-Information Management System intake data of 12,480 GBV survivors. Fisher's exact tests were used to compare differences in GBV incidents among women who experienced DV compared to other types of GBV. Using United Nations humanitarian and aging frameworks, qualitative analysis was completed following two rounds of coding. Sixty percent of women aged ≥60 experienced DV. Local women were more likely to experience DV versus displaced women (85.6% vs. 48.3%, p <.001). Six core themes emerged: experiencing versus witnessing violence, intergenerational conflict, livelihoods, alcohol, humiliation, and neglect. Deeper understanding of DV among older women in humanitarian settings is needed, strengthening a call to action to prioritize protection against, and prevention of, GBV more broadly among this marginalized group.

Keywords: domestic violence; cultural contexts; anything related to domestic violence; elder abuse; alcohol and drugs

Violence against women is common in humanitarian settings and is underreported ([15]). The region of eastern Ukraine has experienced armed hostilities since 2014. The first years of the protracted armed conflict had cost the lives of 10,000 civilians and forced 1.4 million persons to be displaced by 2020 ([29]). Owing to a number of vulnerabilities, older women (over 60 years of age) have been among the most impacted demographic in this crisis and have suffered from significant unmet need. Moreover, displaced older women in eastern Ukraine have faced considerable barriers to healthcare, housing, and social protection (OCHA, 2022).

The conflict also has had profound effects on local (i.e., non-displaced) older women's financial stability. Financial dependence places older women at increased risk for domestic violence (DV). A recent secondary review showed that the gendered retirement and consequent pension gap has rendered many older women vulnerable and reliant on family members ([9]), and this dependency increases the likelihood of DV in a conflict context. Our findings can inform policies and programs to prevent violence against older women; of particular importance given the escalation of armed conflict since February 2022.

DV encompasses intimate partner and intrafamilial violence, and incidents of DV and sexual violence have spiked since the onset of the conflict in 2014 ([4]; [12]). The ongoing crisis has exacerbated gender disparities and reported increases of DV ([23]). Although official statistics for DV are not reliable, Amnesty International had reported a 76% increase in cases in the Donetsk region and a 158% increase in Luhanska region in 2018 when compared to average case reports from the previous 3 years ([3]). Although the increase in DV cases is stark, DV is generally commonplace in Ukrainian society with 26% of ever-partnered women between 18 and 74 years having experienced intimate partner violence (IPV) in 2022 ([17]). One of the drivers of DV in eastern Ukraine is the lack of Mental Health and Psychosocial Support services (MHPSS) available to combatants. Furthermore, the violence by partners (who are overwhelmingly men) reported by women is taking on more severe forms with the ongoing conflict ([3]). Another driver is changes in family dynamics due to the impact of the conflict. As of 2022, 59% of internally displaced persons (IDPs) are women and older persons over 60 years of age live in 57% IDP households ([22]). Older women are likely more vulnerable to DV due to increased dependency and reduced family support in a new environment as indicated by IDPs self-reporting on their needs ([22]). In the context of humanitarian settings, the drivers of GBV and DV are similar, and among the multitude of survival challenges that conflict brings, the likelihood of DV being neglected is high.

Despite the increasing risk, there is a dearth of data on the incidence and prevalence of DV against older women in Ukraine and other humanitarian settings. A global systematic review of violence against older women indicated limitations with current data "including that it is predominantly derived from high-income countries, often does not address context [...] and does not disaggregate by age group" ([13]). A recently published study on risk of IPV by age group in conflict-affected Nigeria collected data for older women which were classified as women of reproductive age—between 25 and 49 years ([16]). Thus, global evidence on older women over 60 years is scarce concerning their unique vulnerabilities and risk factors for different types of violence in conflict contexts, and based on our review there is no literature in Ukraine regarding DV against women over 60 years. Our analysis aimed to understand factors associated with DV experiences by older women in Ukraine's conflict setting.

There is also a dearth in literature on the mental health impact of violence on older women in Ukraine, as the conflict has deprioritized the provision of mental health services and psychosocial support ([23]). This is exacerbated by limited policy protections in place and a lack of psychosocial services, particularly for older survivors of violence ([2]; [21], [23]). Between 2015 and 2020, Ukrainian Foundation for Public Health (UFPH), HealthRight International, and the United Nations Population Fund (UNFPA) deployed mobile MHPSS teams in eastern Ukraine to partially address this service gap. The secondary analysis we conducted covers the period 2014 to 2017. In order to develop policies and programs that reduce vulnerability and improve access to psychosocial services for older women, we need to better understand the unique context and patterns of violence experienced by older women in conflict settings. This was a mixed-methods study seeking to gain a deeper understanding of older women's experience of DV in conflict-affected Ukraine.

Our primary research question was to better understand the vulnerabilities to DV among a sample of older women seeking mental health services in a conflict setting, particularly how residency status (i.e., local vs. displaced) influenced their experiences. The definitions of the terms we will be using in this article are highlighted in Table 1.

Graph

Table 1. Table of Definitions.

This was a mixed-methods study using quantitative and qualitative data from a secondary descriptive analysis. We sought to understand the phenomenon of DV among older women in conflict-affected eastern Ukraine from 2014 to 2017. More specifically, we examined differences between conflict-affected older women who were displaced and those who were not displaced to better understand the role that displacement plays in DV vulnerability. Data for this analysis came from UN standardized case management intake forms collected from women accessing crisis intervention counseling through our mobile MHPSS teams. We focused on older women specifically because we conducted mixed-methods analysis of a subsample of 150 older women across an age range between 60 and 80 years. We intentionally used the term DV rather than IPV because IPV restricts violence to the partner, whereas DV encompasses violence inflicted by partners, children, direct or indirect relatives, and other persons living in the same residence but not an intimate partner.

For the purposes of this article, we define DV as per Table 1. Family members were defined as "persons married to each other; persons living as family but not married to each other; their children; persons under guardianship or care; are direct or indirect relatives living together" in accordance with Ukrainian law ([10]) (see Table 1).

This retrospective study was a secondary descriptive analysis of responses to a semi-structured assessment form developed by the UN GBV Information Management System (GBV-IMS) Global Team to harmonize GBV data collected during service delivery in humanitarian settings to assess older women's vulnerabilities in different regions of Ukraine. One open-ended question was summarized by the social service providers and the qualitative accounts were the summary notes of the provider from the mobile team. Cross-sectional data were gathered through the intake forms.

To ensure a comprehensive understanding of their lived experiences, we conducted quantitative analysis of a randomized subsample of 150 older women across an age range between 60 and 80 years from UN GBV-IMS form intake by HealthRight's mobile teams. The GBV-IMS form was adapted for the humanitarian context in conflict-affected Ukraine. A group of stakeholders under the leadership of UNFPA, including UFPH, HealthRight International and International Medical Corps who piloted mobile teams in eastern Ukraine, translated, adapted and validated the tool for the context (e.g., the education, employment, and residence status categories were tailored).

The purpose of the secondary analysis was to better understand the met and unmet GBV vulnerabilities experienced by older women in Ukraine. We analyzed the responses to one open-ended question within the UN GBV-IMS form. Recognizing DV as a primary finding, additional analyses were conducted to compare DV versus non-DV cases among displaced versus local older women.

Between February 2016 and June 2017, data for the GBV-IMS was gathered by HealthRight's mobile teams comprised of trained psychologists and social workers. As part of completing the GBV-IMS forms, providers were asked to briefly describe the incident. These brief descriptions were the basis for our qualitative analysis. Informed consent was obtained from clients at intake following standard procedures ([28]). Administration of the intake forms ranged in duration from 20 min to 1.5 hr.

The purpose of primary data collection was for crisis intervention counseling and referrals. A total of 12,480 women sought services from the mobile teams in Luhanska, Donetska, Dnipropetrovska, Kharkivska, and Zaporizka oblasts whose self-reported residency status at the time of the incident was either local or displaced. Out of 12,480, 1,522 women were aged 60 and over (12% of the original sample). Out of 1,522, 150 (10% of women aged 60 or over) women were selected at random using a random-number generator for the qualitative analysis which works by using an algorithm to generate numbers in a pseudo-random fashion ([19]). The subsample of 150 women contained a 70/30 split between older local and displaced women, respectively.

The present study initially intended to focus on GBV more broadly, but since one of the primary analyses found a high prevalence DV, a second round of analysis was conducted. The GBV-IMS forms recorded two types of information related to GBV. One was related to the type of violence experienced, specifically, physical, emotional, or economic GBV, as well as an option for non-GBV. In addition, characteristics related to the perpetrator were collected. To construct the DV variable, those instances of violence in which the perpetrator was a family member (i.e., intimate partner, children, in-laws, or grandchildren) were coded as DV, whereas those in which the perpetrator was not a family member (i.e., neighbor, service provider, stranger, etc.) were coded as non-DV. Supplemental Table S1 presents the distribution of selected characteristics of the sample of older women randomly included in the qualitative analysis to all older women in the sample. This study reported findings from a qualitative analysis of one specific, open-ended question in the GBV-IMS form about the experiences of older women survivors. The inclusion and exclusion criteria are summarized in Table 2.

Graph

Table 2. Inclusion and Exclusion Criteria.

Only data from participants who consented to releasing their de-identified information were included in this analysis. To ensure confidentiality, each response was anonymized by the trained psychologists and social workers to ensure no personal identifiers. As secondary analysis of de-identified data, the redacted ethical review board does not consider this study as one which involves human subjects research.

Descriptive statistics (n and %) were used to explore survivor, perpetrator, and incident type characteristics among 150 women selected for qualitative analysis. Differences in the socio-demographic and violent incident characteristics between older women who did and did not experience DV were assessed using Fisher's exact tests (given the small sample size). Generalizability of the findings to the entire sample was assessed by using Fisher's exact tests to compare the distribution of selected characteristics among the group of 150 women included in the qualitative analysis to the sample of all older women. A p-value below 0.05 was considered significant. All analyses were conducted in Stata 15.1 (Stata Corporation LP, College Station, TX, USA).

Qualitative results from one open-ended question in the UN GBV-IMS form were coded using Dedoose 4.3 (Los Angeles, CA, USA: SocioCultural Research Consultants, LLC, 2019). Thematic analysis was conducted by identifying emergent themes through inductive coding, allowing codes and themes to be derived based on the qualitative data itself, with no preconceived notions. Six steps to thematic content analysis provided the guide for data analysis with the following specific processes. First, all transcripts were read by the research team members to ensure all were familiar with and had obtained a sense of the depth of the data. Second, following two rounds of coding, 14 codes were generated related to the research question and included supporting quotes. Then, a detailed codebook emerged from the data and code frequency reports were examined to explore relationships between codes to develop themes using a coding matrix. Since we were unable to speak to respondents directly, an interpretation meeting with the primary data collectors (i.e., mobile team crisis intervention counseling providers) was held to review the data, ensure accuracy in the interpretation, and verify if the themes resonated with them. Member checking was consistent with their observations and we cross-checked the results with the published literature, including HelpAge resources. The themes then underwent thematic analysis and were revised into six core themes, with subthemes identified. Finally, the themes were solidified and the scoping review of previous literature in this area was used as a comparator for themes identified during data analysis to assess for similarities.

In order to understand the vulnerabilities that older women face in conflict-affected Ukraine, we sought to compare our data against indicators from humanitarian-specific and aging frameworks that incorporate aging or GBV to assess the gaps in care for older women in Ukraine. Among the identified common themes that emerged, we used the UN Women and Madrid International Plan of Action and Aging (MIPAA) frameworks ([25]; [26]) to contextualize the themes. These frameworks were chosen because they are the only frameworks that address the current issues that this research centers on. Based on our research, there is no framework that focuses on the protection of older women against GBV more broadly (or DV more specifically) in humanitarian settings. Hence, these two frameworks are the closest to a complete roadmap for this population.

The first framework used was the UN Women's GBV emergency response program model that contains three indicators of GBV response management ([25]): (a) survivors of GBV access appropriate services in a safe and timely manner, (b) women and girls face reduced risk of violence, and (c) policies, systems, and funding prioritize women, girls, and survivors.

The second framework was the MIPAA (UNECE, 2018), which focused on these areas: (a) advancing health and well-being into old age, (b) older persons and development, and (c) ensuring enabling and supportive environments. Both frameworks served as a lens for analysis and guidance to better assess the strengths and weaknesses of GBV response management for older women in humanitarian settings. The six indicators from the two frameworks served as a guideline and comparator for our results and data analysis.

Table 3 presents the distribution of selected characteristics of the sample of older women (n = 150). A majority of the women in our sample (62.6%) were aged in their 60s; most women were widowed (46.0%), local (70.7%), and were not in the workforce (77.7%). Almost three out of four incidents happened at home (74.7%) and half of women experienced psychological violence in a domestic setting (50.3%). There are some key differences among women who experienced DV and those who did not. Local women were more likely to experience DV compared to displaced women (85.6% vs. 48.3%, p <.001). Most (93.3%) incidents of DV occurred at home. Marital status, disability status, occupation, oblast at the time of incident, and type of violence experienced did not significantly differ by whether the violence was domestic or not (Table 3). Among 150 older women, we found that 20% of displaced women who were able to access care through the mobile teams had a physical disability compared to 13% local or non-displaced women.

Graph

Table 3. Descriptive Characteristics of 60+ Year-Old Women Receiving Psychosocial Support in Eastern Ukraine by Violence Type, N = 150, 2016 to 2017.

1 Notes. GBV = gender-based violence; IDP = internally displaced person.

2 p-value based on Fisher's exact tests * Violence total is 149 because one person did not experience GBV.

As shown in Table 4, displaced women were more likely than local women to experience economic violence (45.5% vs. 17.1%, p =.001) and were marginally less likely to experience psychological violence in a domestic setting (38.6% vs. 55.2%, p =.074).

Graph

Table 4. Types of GBV Experienced and Referral Patterns Among ≥60+ Year-Old Women Receiving Psychosocial Support in Eastern Ukraine by Residency Status, N = 149, 2016 to 2017.

3 Note: Total is 149 because one person did not experience GBV; p-value based on Fisher's exact tests.

Table 5 presents the distribution of perpetrator relationship (which in the context of DV includes family and non-family members living together) and incident type among women aged 60 and over receiving psychosocial support. In almost half of the cases (49.3%) women experienced violence at the hands of family members other than their intimate partner, such as children and grandchildren. In about 1 in 5 cases (21.3%), women experienced psychological trauma due to war-related experiences (e.g., shelling of their home). The most common types of incidences experienced by older women were humiliation (18.7%) and economic violence (18.7%). The vast majority of the women directly lived the violence (89.8%) rather than witnessed the violence (10.2%).

Graph

Table 5. Description of Perpetrator Relationship and Incident Type Among Women Aged 60 and Over Receiving Psychosocial Support in Eastern Ukraine, N = 150, 2016 to 2017.

• 4 Note. GBV = gender-based violence.
• 5 May add to over 150 because of co-occurring abuse types.
• 6 Among those experiencing GBV, not war-related experiences without specific perpetrators.

Through the secondary analysis, six conceptual themes emerged that illustrated challenges faced by older women. Below, we present the perceptions of violence incurred and explore the drivers of violence.

Nearly three-quarters of respondents reported experiencing DV. As highlighted in Table 1, we defined experiencing DV as the intentional exertion of physical, psychological, or economic abuse by family members (including direct or indirect relatives) living together. According to the data, the overwhelming majority of violence reported by older women was experiencing DV. Comparing local versus displaced older women as it pertains to the most common type of violence experienced, over half of local older women experienced psychological violence in a home setting. In contrast, economic violence was the most common type of violence experienced by displaced older women.

In terms of the perpetrator, nearly all local older women reported experiencing DV by an intimate partner or family member, whereas some displaced older women reported experiencing physical violence perpetrated by militia or non-armed state militants at the checkpoints. One participant reported an incident of experiencing psychological violence in a domestic setting: "Psychological violence inflicted by brother who claims he has right to inherit part of the house."

Witnessing DV was almost exclusively reported by local older women and all incidents were instigated by family or extended family members. The respondents expressed a feeling of helplessness in witnessing these incidents. An example of an incident described by a respondent: "Son attacked daughter-in-law, the woman is very worried and scared."

Intergenerational conflict emerged as an added stressor for older women and the vast majority of respondents who reported incidents related to intergenerational conflict were local older women. The data illustrated this theme as interpersonal discord that is triggered mostly by a combination of older women's dependencies on family members, and the impact of family members depriving older women from taking care of their grandchildren. A participant recounted an incident of intergenerational conflict: "The woman's son-in-law [...] forbids her from seeing her grandchildren. The woman is [...] very upset."

This was the least represented of all of the themes, but the intergenerational differences are exacerbated by the conflict which can result in violent incidents as described in an example of a son-in-law who also forbade the woman from "seeing her grandchildren" and "is often drunk, displays aggressive behavior while communicating with woman, humiliates, verbally offends."

Many women highlighted how their livelihoods had been compromised by war-related experiences and struggles from the conflict. There were no differences to highlight between local and displaced older women. One respondent recalled: "People fled in whatever they were dressed in. Only collected the necessary documents and no clothes." The livelihood that stood out the most among both local and displaced older women was economic abuse by family members. Economic abuse often was in the form of taking away pensions, which older women described as their only source of income. One example of pension abuse cited was a man who was trying to "limit the jointly earned money" and kept the pension to himself. Another example included a grandson who took the woman's pension and "threatened to put her in an elderly home."

Our findings showed that alcohol use played a role in exacerbating DV incidents against older women. The overwhelming majority of alcohol-related DV incidents were reported by local older women. The most common perpetrators reported according to the data were intimate partners and male family members. All local respondents revealed violent DV incidents including physical, psychological, or economic violence occurring at home that involved the perpetrator consuming alcohol. One example of alcohol-related aggression was reported by one respondent who indicated her son came back home and used tactics such as humiliation and intimidation to get money and exploit their dependency. Another participant described an alcohol-related DV experience:

The woman, who is currently retired, is a pensioner and has a son [...] The man [...] abuses alcohol, talks back to mother, humiliates her, maybe even hits her. On the day of the incident, the man took all of the pension fund. After that, he went into the house to his mother and pushed her into the yard, the woman fell into the snow, began to scream. The neighbors ran outside after hearing the cry and called the police.

Humiliation was commonly reported by older women in the form of offensive remarks and alarming language during incidents of psychological violence in the home. A large majority of respondents who reported feeling humiliated during quarrels were local older women. Humiliation was a core tactic used as violence against local older women, particularly in the context of DV as perpetrated by family members and intimate partners. The feeling of humiliation was explicitly expressed by older women during incidents where they were humiliated because of "economic disadvantages" or being wrongfully accused by family members.

This theme emerged and is overlaid by the multiple and intersecting forms of discrimination that older women faced. There were only few cases of older women describing neglect, and the majority of these cases were reported by local older women. Neglect was represented in the form of disregard of older women's vulnerabilities and was experienced at the individual (e.g., from family members) and institutional level (e.g. from hospital staff). One participant described an incident of neglect by family members:

The woman lived with her daughter and 5 grandchildren. Two months ago, the daughter married a serviceman and moved with him to Chernihiv, and took only 2 young children. The grandmother lives in poverty.

Neglect was noted as an abuse of trust, as well as a need for care that was either not fulfilled despite the urgency, or imposed on older women. At the institutional level, a respondent described an incident at a hospital: "While in the hospital for hypertension, the client experienced neglect by medical staff. She was insulted and humiliated." Older women relayed their experiences of neglect as an abandonment through failure to receive care and assistance when it was most needed.

Common recurring cross-cutting themes that arose out of the analysis included alcohol, economic abuse (including abuse of pension) and humiliation in order of frequency. Economic abuse was most frequently coupled with alcohol-related DV incidents and humiliation equally co-occurred with economic abuse and intergenerational conflict.

This secondary descriptive analysis sought to understand the phenomenon of DV among older women in conflict-affected Ukraine. The analysis also helped identify potential barriers and facilitators to accessing GBV care. Our results identified DV as a primary type of violence among older women's experience. It was seen to be a major threat for older women seeking care which is in alignment with the literature. Contrary to what would be expected based on prior literature ([1]), our findings indicated that local older women were more likely to report DV than IDPs. This finding is not necessarily a clear indication of DV prevalence among local and displaced older women. Instead, our analysis contributes to the understanding of the forms of violence experienced by an already marginalized group, how this violence may be exacerbated in a conflict setting, and how these multi-faceted vulnerabilities render older women's broader experiences of GBV neglected in humanitarian settings.

The study's qualitative analysis enabled us to better understand these vulnerabilities. Six themes were identified that were consistent with the MIPAA and UN Women frameworks (UNECE, 2018; [25]). Our qualitative analysis overall indicated insufficient implementation of these frameworks through policies and programs in conflict-affected settings where older women need greater assistance and protection. The neglected needs of older women experiencing and/or witnessing DV impact their livelihoods. Cultural barriers, including stigma and shame, make it less likely for survivors to seek care. In addition, DV is often perceived as a problem to be solved at home with no need for external support, including at the judiciary level with 38% of Ukrainian judges having this mindset ([21]). This increases the risk of violence against older women in fragile settings.

Comparing our findings against the UN Women framework, our study's core themes highlighted the gaps in the protection of older women in fragile settings, specifically in reducing the risk of violence. Also, there is not enough data to assess access to appropriate services or policies and systems prioritizing older women, so we are not able to draw conclusions on these issues. Our study is one of the few that examines the effects of GBV on older women in humanitarian settings. To our knowledge, there are no DV-specific frameworks in humanitarian settings although there is evidence of IPV-specific interventions for humanitarian settings ([6]). The six key findings below are in alignment with our research question and the DV literature:

A major finding in our secondary analysis was that DV significantly impacted older local women. This result emerged separately to the core themes and we can speculate that older women's dependent position in the family increased their vulnerability to experiencing violence within the home. This is supported in the literature in the context of humanitarian settings ([25]; [9]).

Almost all local women experienced rather than witnessed violence, and this is indicative of their residency status that leaves them more vulnerable to experience DV. In relation to type of violence experienced, local women were more likely to experience psychological violence (55%), which is in line with it being the most frequent form of DV used against older women in Ukraine, exacerbated by the mental toll of the conflict ([22]). Being separated from families and often living on their own explains why displaced older women were not subjected to DV in the same way local older women were, but they experienced other types of violence as indicated by our finding that displaced older women experienced economic violence the most (46%). The inability to access financial support was an issue particularly among older displaced people. We can infer that this may lead to increased vulnerability to GBV in the form of economic violence among displaced older women than local women due to more barriers to access pensions ([9]).

It is important to identify the nuances in the experiences of displaced versus local older women through further research, and since our findings showed DV (in particular) largely impacts local older women, this strongly points to the limited understanding of how older women experience abuse, not only in research but in implementation as evidenced by both the UN Women and MIPAA frameworks not capturing the manifestations of abuse of older women versus GBV (UNECE, 2018; [25]).

A sub-indicator as part of UN Women's rapid response framework during crises is safe access to GBV health services ([25]). Older women's low knowledge about the availability of these services poses a barrier to access that is an additional sub-indicator within the framework. However, our findings on channels of communication for older women revealed that the most popular mode for both local (34.9%) and displaced (31.8%) older women was self-referral and public advertisement when learning about crisis intervention services. This suggests that health promotion advertising is an important component in knowledge dissemination about the services and requires more attention to address the service access barrier. Our analysis also indicates a gap in referrals, as more women should have learned about the services from other providers and social services.

Intoxication with alcohol is a known risk factor for DV and has also been documented in Ukraine ([11]). Alcohol is often used to deal with symptoms of post-traumatic stress disorder as a result of conflict-involvement and displacement, further increasing the risk of violence against women living with family members ([20]; [21]). These mechanisms are exacerbated by financial instability in conflict-affected settings ([21]), and alcohol was evidently present in the majority of violent incidents toward older women in our analysis.

Our analysis also showed that alcohol-triggered physical incidents tend to occur against vulnerable older women who typically have the most stable income in the household in the form of a pension. Alcohol appears to play a role in DV through a manifestation of the broader context of the conflict leading to economic uncertainty, which fuels the habit of perpetrators using older women's income to consume alcohol, and there is a tendency for that to lead to physical and psychological violence toward older women.

The quantitative analysis of our subsample showed two significant findings: older women have a multifaceted experience of GBV, and the majority of older women reporting psychological violence in a home setting were also experiencing DV. The six core themes generated from our findings provided further insight into older women's vulnerabilities, and they can be broadly categorized into two categories: interpersonal/individual factors such as experiencing/witnessing violence, intergenerational conflict, humiliation, and neglect and external factors such as alcohol and livelihoods.

What we can interpret more specifically from our findings is that the interpersonal themes highlighted the disregard of older women's needs and age was a factor in their exposure to, and experience of, violence. Among the 150 older women, age was significant in women's experience of violence (60% of women aged 60 or older experienced DV), and DV was mainly experienced among local women (85.6%). A possible interpretation for this finding could be that older women are more likely to live with their family (i.e., local), and their pension status can be a contributing factor to experiencing violence ([9]). Experiences of GBV among local versus displaced women in Ukraine has been documented in the literature with internally displaced women experiencing violence three times more frequently than local women ([27]). This does not negate our findings which suggest residency status makes a difference in the manifestation and types of violence described by the respondents. Furthermore, the UNFPA report findings are not specific to older women, and this is indicative of the limited research focusing on older women's vulnerabilities in humanitarian settings and patterns of violence observed according to residency status. Evidence of increased prevalence of DV during the conflict has also been documented (OSCE, 2019; [12]), but this is also not specific to older women. Thus, existing literature elaborates on the types and manifestations of GBV in Ukraine's conflict context but does not directly address the vulnerabilities experienced by older women in relation to abuse and DV.

Our core themes intersected with all three indicators of the MIPAA framework (UNECE, 2018) in that livelihoods of displaced older women and experiences of DV among local older women undoubtedly hindered advancement of their health and well-being as well as development in emergency situations. We believe that the first and second indicators in the framework cannot be accomplished for older women in fragile settings without strengthening resilience, ensuring well-coordinated services and addressing barriers to service access.

While GBV (in the form of DV) was prevalent in our sample, it is not represented in aging frameworks. We recommend policy and program implementation that is specific to older women's risks and vulnerabilities to violence in conflict settings, with a key focus on access barriers to psychosocial services. Safe access to services was difficult to assess based on our findings because we analyzed the experiences of older women who sought care from a mobile crisis intervention counseling service, but this was not representative of women who did not seek care. The sub-indicators of decision-making and cross-sectoral linkages between humanitarian actors need to address GBV risks in the UN Women framework. The frameworks also fail to address issues of diversity among GBV in older women. This is especially the case for older women in Ukraine who comprise a large demographic that is marginalized. Gender, aging, and diversity need to be mainstreamed and streamlined to encourage a multi-sectoral response in GBV response frameworks. While our findings were generally in alignment with both frameworks, there is no broader framework that addresses the nexus of diversity, aging, and GBV, adding to the vulnerabilities of older women in fragile humanitarian settings that are often overlooked.

A GBV gap analysis highlighted the persistent gaps that continue to challenge the GBV sector in humanitarian settings ([14]). The report identifies the referral pathway as a response to address the operational and systematic challenges, while continually acknowledging the complexity and diversity of needs of women and girls. While this pathway adopts a survivor-centered approach, there is no indication for what this accessible service would look like for older women. Furthermore, in the UN Women framework, the sub-indicator of women having access to cash-based support is not applicable to older women, not only due to their income being pension-based but cash also poses a risk of family members taking the cash as a form of economic abuse. This only reinforces the need for a framework that accounts for and is reflective of older women's realities, as highlighted in the recommendation to attend to the vulnerabilities of specific groups when planning GBV response efforts during conflict and war based on lessons from the field ([5]).

Our findings were generally in alignment with both frameworks, but we recommend tailored GBV response programming that addresses diversity concerns among older women experiencing violence (such as DV) in humanitarian settings. The gaps in both frameworks allow for us to advocate for diversity in response and prevention programs. But in order to understand where the gaps lie, an overlay of the two frameworks is needed so that diversity is directly addressed, and older women are included. Also, more research that directly addresses older women's vulnerabilities and experience of violence in humanitarian settings is needed.

The link between neglect and abuse is a gray area in the literature, making it challenging to distinguish between older women experiencing abuse versus GBV. The principles of the MIPAA and the global 2030 agenda involve improving the lives of older women through the right to access health services (advancing health and well-being into old age), older persons development including in emergency situations, and ensuring supportive environments that are enabling and are free from neglect, abuse, and violence under the development indicator (UNECE, 2018). Our findings showed that older women experienced neglect through abandonment of their care-dependency at both the individual and institutional level. At the individual level, instances of neglect were perpetrated by family members rather than intimate partners which corroborates the finding of 46% of older women were widowed. Neglect was aggravated by the fragile livelihood due to the conflict which perpetuates intergenerational conflict, economic abuse, and alcohol-related violent incidents. Hence, a major needs gap based on our findings is the lack of curation of supportive environments for older women that are free from violence and neglect through social protection measures.

The qualitative and quantitative results of this study should be interpreted cautiously given the small sample cross-sectional investigation of the data. A clear limitation was conducting a secondary analysis based on secondary accounts (i.e., notes of the social service providers summarized by the mobile team) for a different purpose than our research and not being able to fully assess how well the data collection process was conducted. Also, the sample is not generalizable to the population of older women, given the small sample size and that it is solely among women who received services. However, our findings showed that of those who reported and sought services from the mobile teams. Comparing the randomized subsample of 150 women with the original larger sample, there were no significant differences between the selected characteristics. These included disability status, residency status and type of violence experienced, but there were more 70- to 80-year-old women in our subsample (30%) compared with the larger sample (24.6%). Based on these comparisons, we are confident that our sample was representative of the larger sample. However, this is only representative of those who were able to access care through our mobile teams.

Another limitation was that the sample is heavily weighted toward local older women. Although the analysis revealed the DV finding, the over-representation of local older women undoubtedly skewed the prominence of these accounts. Although we identified recurring concepts such as older women witnessing and experiencing violence, we observed interesting trends that were not statistically significant due to the low sample number. An example of an insightful trend was the number of older women with disabilities experiencing violence being higher than older women without disabilities. This limited the generalizability of the sample particularly given the barriers to care among the most vulnerable population. Furthermore, the answers provided by the respondents were short excerpts rather than in-depth interviews which restricts the ability to examine concepts and themes in detail. Since the analysis was based on one question asked by the mobile team, the depth of the answers was limited. Based on our review of literature worldwide, there is a lack of research on older women's context for DV in humanitarian and non-humanitarian settings at the global level. The indicators we analyzed are limited; hence, we advocate for studies with disaggregation variables specific for this population that encapsulate an array of factors such as disability and culture.

Lastly, it is important to note that all respondents reported their GBV experiences in the territories controlled by Ukraine (at the time of data collection), so we have no data from occupied territories. Also, medical care was not provided to the respondents and therefore health consequences for older women who experienced DV were not captured in the data. Despite these limitations, a strength is the sample size being large enough for a qualitative analysis and has allowed us to identify key themes in older women's experience of DV. Therefore, this is a timely study to highlight the impact of the current war on women's rights and protection during armed conflict, as demonstrated by the recent ratification of the Istanbul convention ([8]) which follows on from Ukrainian legislation to criminalize DV in 2019 ([4]). The next step is urgent GBV response implementation that ensures the inclusion of older women and factoring the high likelihood of underreporting of DV incidents as a result of this legislation and the ongoing conflict.

Our study found that compared with older displaced women, older local women have an increased vulnerability to DV manifesting in physical, psychological, and economic violence in Ukraine's humanitarian setting. This demonstrates the need for DV accountability in conflict settings through frameworks that integrate older women within aging frameworks and include violence against older women, particularly in humanitarian settings. Programming should be sensitive to this population's vulnerabilities, as based on our findings alone, residency status impacted the experiences of DV among older women in conflict-affected Ukraine. The importance of this research lies in identifying barriers and gaps experienced by older women, as there is a dearth of literature in this population in humanitarian settings. Older women's experiences are neglected, and a major reason is the lack of evidence to support the need for prioritizing older women, both at the policy and implementation level. This is especially critical in the recent conflict escalation since February 2022. Many cases are not investigated, but there is a reported increase in GBV among local older women particularly in the temporarily occupied territories. Hence, greater commitment is needed to improve data collection and disaggregate data to better understand underlying issues facing older women in conflict settings, as well as policies and social protection programs to support older women and ensure their rights.

Graph: Supplemental material, sj-docx-1-jiv-10.1177_08862605231214594 for Understanding Domestic Violence Among Older Women in Ukraine: A Secondary Analysis Using Gender-Based Violence Screening Data by Sara Rushwan, Halyna Skipalska, Ariadna Capasso, Peter Navario and Theresa Castillo in Journal of Interpersonal Violence

We acknowledge funding support from the United Nations Population Fund (UNFPA) for this work during the period 2015 – 2020.