The Prevalence and Psychological Costs of Household Violence by Family Members Against Women With Disabilities in Cambodia

Abstract

Women with disabilities (WWDs) are at triple jeopardy due to the combined risks associated with gender, disability, and violence. Not only are WWDs marginalized socially but the violence against them in their own homes is largely neglected in domestic violence research. Evidence from developing countries is particularly sparse. A cross-sectional survey conducted in Cambodia found rates of violence by household members besides intimate partners were significantly higher among WWDs than non-disabled women. This violence engendered increased levels of psychological distress and higher rates of physical injury but low rates of disclosure to health workers and other formal sources of potential support. Community-based strategies are recommended to radically change social and cultural attitudes, beliefs, and responses to WWDs who are victims of household violence to reduce negative social reactions toward them and to make it safer for them to disclose and receive psychosocial, legal, and other necessary support for this underreported type of violence.

Keywords

vulnerability to abuse domestic violence women with disabilities predictors of household violence cultural contexts

In the last decade, views of disability have undergone a fundamental change. No longer conceptualized as primarily a medical condition or a matter of charity or dependency, disability is now recognized as a human rights issue that demands a social understanding (United Nations, 2006; World Health Organization [WHO] and World Bank, 2011).

According to The World Report on Disability (WHO and World Bank, 2011), just more than 15% of adults worldwide have a disabthat demands a social understandingility with higher rates in low-income compared with high-income countries (18% vs. 12%). Across countries, women have higher rates of disability (19%) than men (12%) and are exposed to the combined risks associated with female gender and disability, including gender-based discriminatory attitudes and unfair social treatment (WHO and World Bank, 2011).

Women with disabilities (WWDs) are vulnerable to violations of their human rights as set out in the Convention on the Rights of Persons with Disabilities (CRPD) and its Optional Protocol (United Nations, 2006) including the right to receive an education (Article 24).

Recent evidence, predominantly from high-income countries, indicates WWDs have higher rates of interpersonal violence than non-disabled women (K. Hughes et al., 2012; R. B. Hughes, Lund, Gabrielli, Powers, & Curry, 2011). This violence is a leading cause of physical and psychological morbidity (WHO and World Bank, 2011) and reflects inter-linked violations of the rights identified in the CRPD. It shows contempt for the inherent dignity, autonomy, and self-determination of WWDs and the principle of equality between men and women (Article 3), subjects WWDs to cruel, inhuman, or degrading treatment or punishment (Article 15), to exploitation and abuse (Article 16) and it undermines their right to physical and mental integrity (Article 17) and the highest attainable standard of health (Article 25).

WWDs’ need for personal assistance with daily living and reduced physical or emotional defenses heightens their risk of interpersonal violence (Nosek, Clubb Foley, Hughes, & Howland, 2001). Violence can involve the removal of or deliberate damage to mobility or other aids and equipment (Saxton et al., 2001).

Although 80% of the world’s disabled population live in low-income countries, very little research has been conducted on the magnitude and health consequences of violence against WWDs in these settings (K. Hughes et al., 2012). No large-scale research studies from low-income countries were identified for inclusion in a review of research on violence against adults with disabilities living in the community (R. B. Hughes et al., 2011), or in a global systematic review and meta-analysis of studies on violence against people with disabilities (K. Hughes et al., 2012).

A few small-scale studies on violence against WWDs in low-income countries have reported high levels of violence (Human Rights Watch, 2010; Mohapatra & Mohanty, 2005; Stubbs & Tawake, 2009) but are methodologically weak (K. Hughes et al., 2012).

Pressing questions remain about the level of interpersonal violence and other rights’ violations and their relationship to the mental health of WWDs in low-income settings where women’s mental health has long been neglected in research and development programs (Miranda & Patel, 2005). In high-income countries. robust evidence attests to the strong association between interpersonal violence and poor mental health including high rates of depression, anxiety, post-traumatic stress disorder (PTSD),and suicidal thoughts and actions (Campbell, 2002; Dutton et al., 2006; Ellsberg et al., 2008; Nathanson, Shorey, Tirone, & Rhatigan, 2012; Vos et al., 2006). However, this research has been focused on intimate partner violence (IPV) and on women without disabilities. Its findings may not apply to WWDs, perpetrators besides intimate partners or to low-income settings with different socio-cultural arrangements where women are more likely to live in extended family situations.

The Current Study

This participatory research project in Cambodia was designed to meet the urgent need for robust research from low-income countries on violence against WWDs and its impact on their mental health.

Cambodian women occupy a subordinate social position compared with their male counterparts marked by higher rates of illiteracy, lower rates of pay, and higher levels of gender-based violence (GBV; United Nations Development Program [UNDP], 2010). Cambodia’s Gender Inequality Index is in the lowest tertile of all countries globally (99/145 countries; UNDP, 2011).

Rates of violence against WWDs in Cambodia are unknown. Neither of the two national studies on violence against women deliberately recruited WWDs (Ministry of Women’s Affairs, 2010; National Institute of Public Health, National Institute of Statistics Cambodia and ORC Macro, 2006).

The focus of the study is on household violence, operationally defined as violence or abuse perpetrated in the home by family or other household members, excluding an intimate partner. A number of reasons informed this decision. WWDs are much less likely to marry than non-disabled women and therefore fewer WWDs are exposed to IPV (WHO and World Bank, 2011). WWDs report a greater variety of perpetrators than non-disabled women (Saxton et al., 2001) and in low-income countries that lack a social safety net or alternative sources of support, the vast majority of WWDs live with their families rather than in community or residential care (WHO and World Bank, 2011).

This mixed-method study involved a survey, in-depth interviews, focus groups with members of a self-help group for WWDs, and key informant interviews and was a collaborative project between Australian and Cambodian partners. The findings reported here are derived from the analysis of data from a semi-structured interview used in the survey component of the study.

Aims of the Present Study

The present study sought to determine whether and to what extent WWDs differed from non-disabled women on socio-demographic factors that function as proxy variables for human rights violations such as being denied the right to education, in lifetime prevalence rates of different types of household violence, coercive control, and injury and in the extent of psychological distress associated with such violence. An additional aim was to identify whether WWDs compared with non-disabled women exhibited different patterns of disclosure and access to health care following injury as a result of household violence.

Method

Ethical Clearance

The study was approved by the Monash University Human Research and Ethics Committee (MUHREC) on November 1, 2010.

Eligibility Criteria

To be eligible to participate in the study, both WWDs and non-disabled women had to be aged between 18 and 45 years, able to understand the purpose of the study, what their participation involved including the time they would need to spend being interviewed and able to give informed consent to participate in the research. Informed consent was sought following a plain language explanation of the study by interviewers. Consent could be given verbally by participants with no or low levels of literacy or in writing by participants who were able to read and write. All participants had to be able to communicate sufficiently well to respond to the questions asked in the face-to-face interview. Consequently, women with severe communication disorders, intellectual disabilities, or who were unable to hear at all were not eligible to participate.

Participants

A total of 354 Cambodian women aged between 18 and 45 years participated in the survey. Half the sample comprised WWDs (177) and half were non-disabled women (177). A power analysis revealed a total sample of 330 was sufficient to detect a significant difference, if there was one, in the rates of violence between WWDs and non-disabled women. Slight oversampling occurred to allow for possible sample attrition.

Design and Procedure

A cross-sectional survey involving face-to-face semi-structured interviews was conducted in Cambodia between March 2011 and October 2011. Interviewers worked in pairs, to optimize their own safety and the safety of participants. All interviews were administered in Khmer by bi-lingual Cambodian interviewers that worked for the two Cambodian partner organizations in the research, Banteay Srei, a women’s empowerment organization and the Cambodian Disabled People’s Organization (CDPO). Both organizations ran programs in the field sites chosen for this study. Participants were asked to nominate a place they believed offered privacy and safety for the interview to be conducted.

Participants were recruited from five field sites selected to reflect the geographic distribution of the population in Cambodia where around one quarter of the population lives in urban areas and three quarters live in rural areas. Selection of field sites with existing services for women including WWDs was designed to meet participants’ safety and referral needs. Just more than a quarter of the sample (26.3%) was drawn from the urban sites of Phnom Penh and urban Siem Reap and nearly three quarters (73.7%) was drawn from the rural sites of Battambang, rural Siem Reap and Kampong Speu.

Definition of disability

A two-stage process was used to define WWDs. In the first stage, women believed to have a disability were identified in each field site by the research interviewers working in collaboration with local disability organizations, the partners in the research, Banteay Srei and CDPO that worked in each field site and the local commune. In the second stage, the validity of this initial method of selecting WWDs for inclusion in the study was cross-checked against their responses to the Washington Short Set of Questions on Disability (Washington Group, 2006). Only participants reporting significant difficulty with a function such as seeing, hearing, mobility, cognition, self-care, and communication, were defined as WWDs.

Recruitment took place at the household level. Throughout the process of recruitment, a WWD was recruited first followed by a non-disabled woman. This meant that once a WWD was identified through the multi-pronged process described above, the interviewers visited her at home to determine if, after having explained the purpose of the study, she was interested in participating. Once a potential participant agreed to participate, a time convenient to her was made to conduct the semi-structured interview. After each WWD was interviewed, the field researchers then identified a non-disabled woman in the same street through knocking on the doors of adjoining houses until they found a woman who met the inclusion criteria, did not have any discernible disability and was willing to participate in the research. This method allowed for broad socio-economic matching of the two groups of participants based on shared geographical location and represents a nested case-control study within the survey.

Before going into the field, a weeklong workshop based on the training used in the WHO Multi-Country Study on Women’s Health and Domestic Violence (Jansen, Watts, Ellsberg, Heise, & Garcia-Moreno, 2004) was held in Phnom Penh to equip interviewers to collect high-quality data and understand the ethical and safety aspects of doing violence research.

The study followed the ethical and safety guidelines developed for the WHO Multi-Country Study on Women’s Health and Domestic Violence Against Women (Garcia-Moreno, Jansen, Ellsberg, Heise, & Watts, 2005). Interviewers were trained to offer a break and provide a drink and a snack to any participant who appeared fatigued or said she needed a break, to inform participants that they could elect not to answer any question they did not wish to answer and to check at several points in the interview that the participant wished to and gave consent to continue with the interview. Interviewers also offered a break or to reschedule the interview for another time if any participant became upset but expressed a wish to continue rather than withdraw from the study. They ensured that the interview took place in privacy, and informed each participant before the interview started that if someone did interrupt the interview that the topic would be changed to a pre-agreed neutral topic on women’s health. This latter strategy was designed to protect participants’ safety and reduce the risk of retaliation that might ensue if a perpetrator became aware that violence was being disclosed to an interviewer.

Materials

Survey questionnaire

The semi-structured questionnaire used in the face-to-face interviews was adapted from the one employed in the WHO Multi-Country Study on Women’s Health and Domestic Violence Against Women (Garcia-Moreno et al., 2005). Topics in the questionnaire included the socio-economic and demographic characteristics of participants, gender norms and beliefs, mental health, types of violence, patterns of disclosure, and help seeking by women who reported violence as well as reproductive health and financial autonomy. In addition, the Washington Short Set of Questions on Disability were asked (Washington Group, 2006).

The questionnaire used here was shorter than the WHO questionnaire to reduce the burden of participation for WWDs. Even so, it was estimated that interviews would last between 1 hr and 15 min and 2 hr.

Household violence

Operational definitions of physical, psychological, and sexual violence by a household member employed in the multi-country study (Garcia-Moreno et al., 2005) were used in the current research. All questions on violence were phrased to elicit a “yes” or “no” response.

Six questions were asked about physical violence which was defined by behaviors of increasing severity and included being slapped or having something thrown that could hurt the woman, being pushed or shoved, hit with a fist or something else that could hurt; being kicked, dragged, or beaten; choked or burnt on purpose; and/or threatened with the use or actual use of a gun, knife, or other weapon.

The four questions on psychological violence included being asked about whether a household member had insulted or made the woman feel bad about herself, had belittled or humiliated her in front of other people, deliberately did things to scare or intimidate her, and threatened to hurt her or someone she cared about.

The three questions on sexual violence included being physically forced to have unwanted sexual intercourse, having sexual intercourse she did not want because she was afraid of what the perpetrator might do, and being forced to do something sexual that she found degrading or humiliating. Disability was measured using the Washington Short Set of Questions on Disability (Washington Group, 2006). These questions ask about six functional domains or basic actions including seeing, hearing, mobility, cognition, self-care, and communication. Each question has four types of responses to capture the full spectrum of difficulty functioning from mild to severe: no difficulty, some difficulty, a lot of difficulty, and unable to do it at all. The question on sight is phrased “Do you have difficulty seeing?” and all other questions are phrased with the same introductory phrase, that is, “Do you have difficulty” followed by a specific function or basic action as described previously. A single category to indicate “significant difficulty” was devised by combining the responses of those who reported “a lot of difficulty” with those who reported “cannot do it at all.”

Mental health was assessed using the WHO Self-Reporting Questionnaire (SRQ) (Beusenberg & Orley, 1994). This 20-item, well-validated screening measure of emotional distress was designed for use in developing countries and was utilized in the WHO Multi-Country Study on Women’s Health and Domestic Violence Against Women (Garcia-Moreno et al., 2005). It includes symptoms of psychological distress related to common mental disorders such as depression, anxiety, and somatic complaints. Respondents are asked to indicate with a “yes” or “no” answer whether, within the 4 weeks prior to the interview, they had experienced a series of symptoms such as crying, inability to enjoy life, and thoughts of ending life. For example, depressive symptoms are elicited by questions such as “Do you cry more than usual” and “Do you feel you are a worthless person?”

Relying on the rationale used in the WHO Multi-Country Study, no cut-off score was adopted in the present study because the SRQ has not yet been validated for use among Cambodian women.

Results

Disability Status and Profile

All women initially recruited as a WWD, reported significantly higher levels of difficulty on each of the Washington questions than non-disabled women. Consequently, no re-categorization of disability status was required. For example, with regard to level of difficulty seeing, highly significant differences were found in the extent of difficulty between WWDs and non-disabled women with more WWDs reporting “a lot of difficulty” or “cannot do it at all” (χ² = 26.14, 3 df, p < .001).

WWDs who reported significant difficulty on any of the Washington questions included 18.9% with difficulty seeing, 6.8% difficulty hearing, 45.1% difficulty walking or climbing stairs, 21.7% difficulty remembering or concentrating, and 13.7% difficulty communicating.

Table 1 reports differences in the socio-economic and demographic characteristics of WWDs compared with non-disabled women. Just more than 28% of WWDs had never attended school and three quarters of all WWDs were currently unmarried and lived with their birth families. Chi-squared analyses revealed significantly more WWDs than non-disabled women had never attended school (χ² = 3.9, 1 df, p < .05), never married or had a male partner (χ² = 55.2, 1 df, p < .001), were unmarried at the time of the interview (χ² = 76.1, 1 df, p < .001), lived with their birth family (χ² = 25.7, 1 df, p < .001), and were significantly less likely to be earning money (χ² = 5.6, 1 df, p < .02). Although 71% (126/177) of WWDs lived with their extended families, this living arrangement was not confined to them as 45% (79/177) of non-disabled women also lived with members of their birth family and no significant difference was reported in the length of time each group had lived in their current community. Similarly, no significant difference in the age of the two groups was found. The mean age of the total sample was 31.7 years (SD = 8.3).

Table 1.

Socio-Demographic Differences Between WWDs and Non-Disabled Women.

Socio-Demographic Indicators	WWDs	Non-Disabled Women	p
Number	177	177
M age in years	32.1 (SD = 8.6)	31.4 (SD = 8.1)	ns
Never attended school	28.2%	19.2%	*
Never married/partnered	57.6%	19.2%	****
Unmarried now	75.1%	28.8%	****
Lives with birth family	75.1%	44.6%
Earns money	70.6%	81.4%	**

Note. WWDs = women with disabilities.

p < .05. **p < .02. ****p < .001.

Lifetime Prevalence of Household Violence

As described earlier, several questions were asked about each type of physical, psychological, and sexual violence by a household member (excluding an intimate partner) over a lifetime. Dichotomous variables were formed from the questions asked about each type of household violence and coded as 0 for the absence of a particular type of household violence and 1 for the presence of one or more types of violence. An additional composite variable of “any household violence ever” was created. This variable was based on the summed responses to each type of violence that were recoded with “no” responses to all types of violence coded as 0 and a “yes” response to one or more types of household violence coded as 1.

The lifetime prevalence rates for all measure of household violence are shown below.

Table 2 reveals that the lifetime prevalence for “any household violence ever” was 46.0%. The commonest type of household violence was psychological, reported by 43.8% of the sample. This was followed by physical violence reported by 18.4% and sexual violence, the least common type of household violence, was reported by 4.2% of all study participants.

Table 2.

Lifetime Prevalence of Different Types of Household Violence.

Type of Violence	n	%
Household physical violence	65/354	18.4
Household emotional violence	155/354	43.8
Household sexual violence	15/354	4.2
Any type of household violence ever	163/354	46.0

Chi-squared analyses were carried out to detect whether the difference in rates between WWDs and non-disabled women was statistically significant. Odds ratios (OR) and confidence intervals (CIs) are also reported here. Table 3 shows WWDs had significantly higher rates of all types of household violence compared with non-disabled women. Their lifetime prevalence rate for physical violence by a household member was 25.4% compared with 11.4% for non-disabled women. This difference was highly statistically significant (χ² = 11.6, 1 df, p < .002) (OR = 2.6, 95% CI = [1.5, 4.7]). For psychological violence, the respective rates were 52.5% for WWDs and 35.2% for non-disabled women. This difference was equally statistically significant (χ² = 10.7, 1 df, p < .002; OR = 2.0, 95% CI = [1.3, 3.1]). The lifetime prevalence rates for sexual violence by a household member were considerably lower and were 6.6% for WWDs and 1.7% for non-disabled women. This difference was also significant but at a lower level than that found for physical and psychological violence by a household member (χ² = 5.5, 1 df, p < .02; OR = 4.2, 95% CI = [1.2, 15.1]). The higher odds ratio for sexual violence compared with physical or psychological violence by a household member should be interpreted with caution. The absolute numbers of WWDs (n = 12) and non-disabled women (n = 3) who reported sexual violence are small and the confidence intervals associated with the odds ratio for sexual violence are very wide, indicating a low level of precision of the odds ratio. For any type of violence by a household member, the lifetime prevalence rate for WWDs was 57.1% compared with 35.2% for non-disabled women. This difference was highly statistically significant (χ² = 16.9, 1 df, p < .001; OR = 2.4, 95% CI = [1.5, 3.7]). For all types of violence by household members, WWDs had a twofold or higher increase in their odds of experiencing such violence compared with non-disabled women.

Table 3.

Proportion of WWDs and Non-Disabled Women Reporting Lifetime Household Violence .

Lifetime Prevalence of Different Types of Household Violence	WWDs		Non-Disabled Women		p
	n	%	n	%	p
Household physical violence ever	45/177	25.4	20/176	11.4	***
Household psychological violence ever	93/177	52.5	62/176	35.2	***
Household sexual violence ever	12/177	6.8	3/176	1.7	**
Any household violence ever	101/177	57.1	62/176	35.2	***

Note. WWDs = women with disabilities.

p < .02. ***p < .002.

Mental Health

The measure of psychological distress used in this study, the SRQ (Beusenberg & Orley, 1994), assesses common mental disorders including symptoms of depression, anxiety, and somatic complaints. Each of the 20 problems or symptoms is scored as 0 if the symptom is absent and 1 if it is present, resulting in a minimum possible score of 0 indicating no psychological distress and a maximum possible score of 20 indicating severe psychological distress.

Reliability of SRQ

Reliability of the SRQ for the total sample is considered good (α = .84) with minimal difference between WWDs (α = .85) and non-disabled women (α = .83).

Mean SRQ Scores by Disability and Household Violence

The mean score on the SRQ for the total sample (n = 353) was 12.8 (SD = 4.6) and WWDs had a significantly higher mean SRQ score (M = 13.7, SD = 4.5) than non-disabled women (M = 11.9, SD = 4.6; F value = 13.4, with 1 df, p < .001). When “any household violence” was analyzed, the mean SRQ score for women who reported such violence (M = 13.9, SD = 3.8) was significantly higher than that of women who did not report such violence (M = 11.9, SD = 5.1; F value = 15.9, 1 df, p < .001).

To clarify the relationships between SRQ scores, disability status and “any household violence ever” a further analysis of variance was carried out with mean SRQ scores as the dependent variable, two between-subjects factors, disability (absent vs. present) and any household violence ever (no vs. yes) and three covariates. The covariates selected were chosen to control for the socio-economic differences between WWDs and non-disabled women that emerged in earlier bivariate analyses. These were “ever having attended school” (no vs. yes), “ever married” (no vs. yes) and “currently earning money” (no vs. yes).

Table 4 shows the adjusted mean SRQ scores for WWDs compared with non-disabled women according to whether or not they reported any household violence ever. Only one covariate, “ever married” (F value = 5.3, with 1 df, p < .03) exerted a significant effect. Highly significant main effects were found for the two between-subjects’ factors, namely, disability (F value = 11.7, 1 df, p < .002) and household violence (F value = 14.3, 1 df, p < .001). There was no significant interaction effect between the two factors. The group with the lowest adjusted mean SRQ score comprised non-disabled women who reported no experience of any household violence ever (M = 11.3, SD = 4.9), whereas the group with the highest adjusted mean SRQ score comprised WWDs who reported they had experienced household violence at some time in their lives (M = 14.4, SD = 3.8). Intermediate SRQ scores were found for WWDs who reported no household violence ever (M = 12.9, SD = 5.2). This score was almost identical to the mean SRQ score for the total sample and was slightly, but non-significantly lower than that for non-disabled women who reported household violence ever (M = 13.1, SD = 3.8).

Table 4.

Mean SRQ Scores Adjusted for Socio-Demographic Factors by Disability and Household Violence.

Factors		Adjusted Mean SRQ Score	SD	n
Disability	Household Violence	Adjusted Mean SRQ Score	SD	n
No	No	11.3	4.9	114
No	Yes	13.1	3.8	62
Total		11.9	4.6	176
Yes	No	12.9	5.2	76
Yes	Yes	14.4	3.8	101
Total		13.7	4.5	177

Note. SRQ = Self-Reporting Questionnaire.

In addition to higher rates of household violence, WWDs compared with non-disabled women reported higher rates of coercive control by household members.

Table 5 reveals that a higher proportion of WWDs than non-disabled women reported all types of coercive control. Differences between the two groups of women were statistically significant for half (3/6) of the items measuring coercive control. A higher proportion of WWDs (48.0%) than non-disabled women (37.5%) reported having their right to freedom of movement controlled, as indicated by family members insisting on knowing where they were at all times (χ² = 3.9,1 df, p < .05). Significantly more WWDs than non-disabled women (27.1% vs. 17.0%) also reported being ignored and treated indifferently (χ² = 5.2, 1 df, p < .03) and a much higher proportion of WWDs than non-disabled women (48.6% vs. 34.7%) reported being expected to ask permission before seeking health care for themselves (χ² = 7.0, 1 df, p < .02).

Table 5.

Proportion of Women Reporting Controlling Behaviors by Household Members by Disability Status.

Do Household Members	WWDs		Non-Disabled Women		Total
Do Household Members	n	%	n	%	n	%
Try to keep you from seeing your friends	26/177	14.7	19/176	10.8	45/353	12.7
Try to restrict contact with your family of birth	8/177	4.5	3/176	1.7	11/353	3.1
Insist on knowing where you are at all times	85/177	48.0	66/176	37.5	151/353	42.8
Ignore you and treat you indifferently	48/177	27.1	30/176	17.0	78/353	22.1
Get angry if you speak with a man	42/177	23.7	28/176	15.9	70/353	19.8
Expect you to ask permission before seeking health care for yourself	86/177	48.6	61/176	34.7	147/353	41.6

Note. WWDs = women with disabilities.

Injury and Main Perpetrators

Participants were asked “Have you ever been injured as a result of violence/abuse by family members.” Overall, 13.6% of all participants who answered this question (48/352) reported being injured as a result of physical violence by a family or household member. A higher proportion of WWDs reported being injured than non-disabled women. Almost 19% (33/177) WWDs reported having been injured compared with 8.5% (15/177) of non-disabled women. This difference was highly statistically significant (χ² = 7.8, 1 df, p < .006). An odds ratio analysis revealed that for WWDs the odds of experiencing such an injury were 2½ times higher than those of non-disabled women (OR = 2.5, 95% CI = [1.3, 4.7]). Health care for an injury inflicted by a household member was received by less than one in five women who reported injury. Fewer WWDs reported receiving health care than non-disabled women (14.8% vs. 21.4%) but this difference was not statistically significant.

All participants who reported they had been injured as a result of violence by a family or other member of the household were asked to nominate the main person in their life who had hurt them. Table 6 shows the proportion of WWDs and non-disabled women who had been injured as a result of violence according to the person nominated as the main perpetrator in their lives.

Table 6.

Main Perpetrator in Household Responsible for Any Violence by Disability Status.

	WWDs		Non-Disabled Women		Total
Main Perpetrator	n	%	n	%	n	%
Father/mother	17/24	70.8	7/25	29.2	24/48	49.0
Stepfather/mother	2/3	66.7	1/3	33.3	3/48	6.2
Other male family member	10/15	66.7	5/15	33.3	15/48	31.2
Female family member	9/13	69.2	4/13	30.8	13/48	27.1
Other	2/2	100.0	0	0	2/48	4.2

Note. WWDs = women with disabilities.

None of the differences in the proportion of WWDs versus non-disabled women was statistically significant. Furthermore, no one in either group nominated a teacher, police or soldier, male friend of family, female friend of family, boyfriend, stranger, someone at work, or a priest or religious leader as the main perpetrator of violence in their life. Two WWDs did nominate someone “other” as the main perpetrator.

Disclosure and health care related to physical violence by a family or household member

Responses to questions about specific acts of household physical violence indicated that 65/354 women reported such violence. A later single, more general question on “Who have you told about the physical violence by a family or household member?” prompted 91 women to answer in the affirmative. Why more women responded to the general question is unclear, however, previous research has also found that reported rates of violence differ according to the way questions are framed (Garcia-Moreno et al., 2005).

Based on the responses given to the general question about the disclosure of household physical violence, it is apparent that the majority of victims remain silent. Just over 57% of women (52/91) reported they had told no one. Participants who indicated that they had told someone were able to specify more than one confidante. In descending order, 23.1% had told parents (21/91), 20.9% (19/91) had told friends, 20.9% had told neighbors (19/91), 18.7% had told a brother or sister (17/91), 15.4% had told an uncle or aunt (14/91), and 2.2% (2/91) had told children. The only significant difference in the pattern of disclosure between the two groups related to a significantly higher proportion of non-disabled women than WWDs (23.3% vs. 5.0%; χ² = 6.8, 1 df, p < .01) who had told someone in the “other” category. In total, 11.0% of all women (10/91) disclosed to someone in the “other” category.

Of particular interest was the finding that neither WWDs nor non-disabled women tended to disclose to formal sources of potential intervention or psychosocial support. For example, no woman in the study had disclosed to a priest, a counselor, an NGO, or a women’s organization.

Among the small percentage of women who had disclosed, only 1.1% (1/91) had told the police and this one woman was a WWDs. Of the 3.3% (3/91) who disclosed to a local leader, two thirds were non-disabled women. Conversely and as expected, the 3.3% (3/91) of all women who disclosed to a disability organization were WWDs and represented 5.0% (3/60) of all WWDs who reported physical violence by a household member. Less than 5% (4/91) had disclosed to a doctor or health worker and all were non-disabled women. The difference between WWDs and non-disabled women regarding disclosure to a doctor or health worker was highly statistically significant (χ² = 8.1, 1 df, p < .005) and was the only significant difference in disclosure found between the two groups of women.

Discussion

Cambodia is a signatory to multiple human rights conventions. It introduced a Law on the Protection and the Promotion of the Rights of Persons With Disabilities in 2009 and previously introduced a Law on the Prevention of Domestic Violence and the Protection of Victims in 2005.

A recent update on progress toward the achievement of the Cambodia’s Millennium Development Goals (CMDGs; Ministry of Planning, 2010) indicated increased awareness that violence against women is wrong, constitutes a criminal act, and violates women’s basic human rights.

However, high rates of violence against women remain. More than half of all Cambodian men and women believe that even extreme partner violence such as stabbing, shooting, or throwing acid at a wife is justified, if she is argumentative or disagrees with her partner (Ministry of Women’s Affairs, 2010). Reduction in rates of violence depends on meeting the challenge to change the cultural attitudes and gendered beliefs that condone such violence (Ministry of Planning, 2010).

The findings of the current study reinforce the extent of the challenge remaining, particularly for WWDs, who had more than 2½ times higher odds of household physical violence, twice the odds of psychological household violence and more than 4 times higher odds of household sexual violence than non-disabled women. However, it should be noted that the higher odds of WWDs experiencing sexual violence have very wide confidence intervals and therefore lack precision.

If questions on violence by household members that asked about disability-specific acts of violence had been added to the Washington questions used in the survey, a more comprehensive assessment of the nature and extent of violence against WWDs would have been possible. It is a shortcoming of this research that such questions were not included.

A further limitation concerns the study design itself. Although all differences in rates of household violence and psychological distress between WWDs and non-disabled women in the current study were highly statistically significant, its cross-sectional design cannot determine whether the associations reflect causal relationships nor establish any temporal sequence between disability, household violence, and psychological distress. Finally, the study findings would have been strengthened if a random representative sample of WWDs and non-disabled women had been recruited into the survey rather than the methodologically weaker sampling strategy of selecting WWDs and then attempting to broadly match them socio-demographically with non-disabled women.

Future research that seeks to overcome selection bias, will have to grapple with how to meet the safety and support needs of WWDs and non-disabled women both during and after the study as some adverse consequences may not occur until after a study has been completed. The current study attempted to do this by limiting its field sites to areas where services were run by the research partners, Banteay Srei and CDPO. In low-income, low-resource settings, where health and psychosocial support services are scarce, the ethical and methodological demands of doing research may not coincide.

Overall, nearly 60% of WWDs reported one or more types of violence by a family or household member in their lifetime. This implies that the majority of WWDs have had their fundamental human right to live in dignity, safety, and security violated by members of their own families. When it is not safe to let down one’s defenses due to the possibility of violence occurring, no psychological sense of home is possible (Astbury et al., 2000).

Other rights violations experienced by WWDs extended from childhood through their adult lives. These are suggested by the consistent pattern of differences in the socio-demographic characteristics of WWDs compared with non-disabled women that reflect violations of human rights under the CRPD (United Nations, 2006). A significantly lower proportion of WWDs had ever attended school or ever partnered or married and a significantly higher proportion was unable to exercise freedom of movement or enjoy the right to access health care, unless and until, another adult in the family gave them permission to do so.

The finding that only a quarter of WWDs were currently married or partnered, confirms the importance of investigating other perpetrators of violence besides intimate partners when it comes to researching violence against WWDs. If the focus of this study had been on IPV, it would not have been relevant to 75% of WWDs in the current sample.

Previous research has drawn attention to the connection between rights violations including IPV and poor mental health (Astbury, 2010, 2012), but WWDs from low-income countries have been marginalized in most previous research on violence against women (K. Hughes et al., 2012).

This study indicates that household violence is powerfully associated with severe psychological distress including depression as measured by the SRQ (Beusenberg & Orley, 1994) and that WWDs are significantly more likely to be affected than non-disabled women. Depression is not included in the Washington Short Set of Questions on Disability (Washington Group, 2006) but in Global Burden of Disease estimates, depression is the leading cause of disability. Importantly, rates of depression are 50% higher in women than men (WHO, 2008).

WWDs who reported household violence endorsed more symptoms of depression and other forms of psychological distress on the SRQ than either WWDs who did not report household violence or non-disabled women who did. Their mean score of 14 out of a possible 20 symptoms on the SRQ is likely to be clinically as well as statistically significant. A cut-off score of 8 symptoms has been widely used in previous studies in low-income settings to differentiate between clinically and non-clinically psychologically distressed groups (Scholte, Verduin, van Lammeren, Rutayisire, & Kamperman, 2011). Furthermore, the mean SRQ score of Cambodian WWDs is considerably higher than that reported for any country that participated in the WHO Multi-Country Study on Women’s Health and Domestic Violence Against Women (Garcia-Moreno et al., 2005).

In a resource-poor country like Cambodia, there is a severe shortage of psychiatrists, psychologists, and other mental health workers. Services also tend to be concentrated in the capital, Phnom Penh. The rate of psychiatrists working in the mental health sector is 0.23 per 100,000 and no information is available on the number of psychologists or social workers in the same sector (WHO, 2011). Approximately, three quarters of the population lives in rural areas and given the high level of psychological distress experienced by WWDs who reported household violence in the current study, a large unmet need clearly exists for violence prevention and victim support services including increased access to primary care level mental health services in the rural areas where they are needed most. This process has begun in Cambodia but according to the WHO (2011) Mental Health Atlas the majority of primary health care doctors have not received any official in-service training on mental health within the last 5 years.

No woman in the current study had disclosed household violence to a counselor or a women’s organization even though the study was only conducted in areas where services existed. WWDs were also less likely than non-disabled women to have disclosed household violence to a doctor or health worker despite having more than 2½ times higher odds of being injured as a result of this violence.

Even if counseling and mental health services were more readily available, the findings suggest that the existence of services would not be sufficient to assist WWDs. To access services, victims would need to overcome the twin constraints of having to seek permission at home from the very people who have perpetrated violence against them and the fact that non-disclosure of physical violence remains normative in Cambodia, both in the current study and two earlier national surveys (National Institute of Public Health, National Institute of Statistics Cambodia and ORC Macro, 2006; Ministry of Women’s Affairs, 2010).

In Cambodia, reluctance to disclose violence is reinforced in a socio-cultural context where raising one’s voice risks retaliation, where silence about violence is still expected of victims and is regarded as a desirable attribute of the “good” woman (Jack & Astbury, 2013).

WWDs rely overwhelmingly on informal rather than formal sources of support within their communities when they experience household violence. More than one in five WWDs disclosed violence to parents, friends, and neighbors. As mental health care services are rare and the number of psychiatrists and other mental health workers working in the Cambodian health care sector is very low, it is essential for formal services including health, disability, and victims’ services to engage with the informal sources of support victims rely on in their communities. This would optimize the likelihood of bringing about attitudinal and behavioral change from the village level up and maximize community efforts to shift the current status quo that silences victims and legitimates violence against them.

Community attitudes that condone violence against women including WWDs or blame them for its occurrence facilitate its continuation and contribute to the psychological distress of victims. U.S. research found that negative social reactions to the disclosure of sexual violence were associated with an increase in the severity of the symptoms of PTSD (Ullman & Filipas, 2001).

Consequently, reducing the stigmatization of and negative social reactions to victims of violence is critical in secondary prevention efforts. Victims are more likely to disclose if they believe it is safe to do so. Moreover, research on survivors of IPV indicate that positive social reactions and social support are related to a higher quality of life and lower levels of depression (Beeble, Bybee, Sullivan, & Adams, 2009).

Impact surveys on the UNDP Community Conversations Project (UNDP Cambodia, Ministry of Justice, & Ministry of Interior, 2010) show it is possible to increase awareness and understanding of domestic violence and to create an environment where the taboo of domestic violence is broken and can be discussed more openly. Unfortunately, the violence experienced by WWDs in their own homes was not specifically addressed in that project. However, there is no reason to suppose the same approach would not work with WWDs. Using village facilitators, this approach involves bringing community members together to identify and analyze their problems, decide on solutions, and plan a course of action. Concerted efforts would need to be made to ensure the safe participation of WWDs in such conversations.

Future research is needed in Cambodia and other low-income settings to determine the effectiveness of low-cost community-based interventions in changing attitudes, beliefs, and responses to WWDs who are victims of violence, in reducing the prevalence of household violence and in improving the psychological well-being of victims.

These community-level interventions need to be framed within a legislative, policy, and programming structure that recognizes the unique vulnerabilities to violence faced by WWDs, and where violence prevention, support, and counseling services are accessible and welcoming to WWDs.

Footnotes

Acknowledgements

We would like to thank AusAID for supporting the Triple Jeopardy project, the women who participated in the survey and shared some painful experiences from their lives, and our co-investigators and colleagues, Jo Crawford, Kathy Oliver, and Nina Vallins from IWDA, Sok Panha, Director of Banteay Srei and Ngin Saorath, Executive Director of CDPO. We would also like to thank the Cambodian research team and field researchers who conducted all the survey interviews, including research coordinators Heng Channtey, Tith Hieng Seka, and Tep Danang and field researchers, Nak Samneang, Touch Siya, Chhay Thida, and Ton Douern.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The study was funded by AusAID, the Australian aid agency, through an Australian Development Research Award (ADRA)

Author Biographies

Jill Astbury is a psychologist with more than 20 years of research experience investigating the social determinants of women’s mental health. Her research has focused on investigating the ways in which gender, social position, and the violation of women’s human rights, particularly that represented by different types of gender-based violence, impact on their mental and reproductive health and emotional well-being. She has more than 100 research publications and was Co-Chair of the Coordinating Group of the Sexual Violence Research Initiative until October 2013 ().

Fareen Walji’s research interests and practice focus on gender, disability, and the interplay between multiple forms of marginalization. Her prior work experience includes anti-violence support work in South Africa, engagement with Aboriginal street youth in Canada, women’s rights research in South India, and the management of disability rights projects with Handicap International in Indonesia. She holds a BA in political science, an MA in development studies, and a master’s in women’s health. From 2008 to 2012, she worked in research for CBM Australia and the Nossal Institute for Global Health at the University of Melbourne where she was involved with research and teaching around disability, gender, and violence in Asia and the Pacific, including the recently completed Triple Jeopardy project. She is currently Programs Manager with the Urgent Action Fund-Africa, a feminist and human rights Pan African fund.

References

Astbury

(2010). Social causes of women’s depression: A question of rights violated? In Jack

D. C.

Ali

(Eds.), Cultural perspectives on women’s depression: Self-silencing, psychological distress and recovery (pp. 19-46). Oxford, UK: Oxford University Press.

Astbury

(2012). Violating the right to health: How partner violence and disability undermine women’s mental health in Cambodia [Special issue on mental health]. Journal of Disability and International Development, 2, 4-11.

Astbury

Atkinson

J. J.

Duke

J. E.

Easteal

P. L.

Kurrle

S. E.

Tait

P. R.

Turner

(2000). The impact of domestic violence on individuals. Medical Journal of Australia, 173, 427-431.

Beeble

M. L.

Bybee

Sullivan

C. M.

Adams

A. E.

(2009). Main, mediating and moderating effects of social support on the well-being of survivors of intimate partner violence across 2 years. Journal of Consulting and Clinical Psychology, 77, 718-729.

Beusenberg

Orley

(1994). A user’s guide to the Self Reporting Questionnaire (SRQ). Geneva, Switzerland: Division of Mental Health, World Health Organization.

Campbell

J. C.

(2002). Health consequences of intimate partner violence. Lancet, 359, 1331-1336.

Dutton

M. A.

Green

B. L.

Kaltman

S. I.

Roesch

D. M.

Zeffiro

T. A.

Krause

E. D.

(2006). Intimate partner violence, PTSD, and adverse health outcomes. Journal of Interpersonal Violence, 21, 955-968.

Ellsberg

Jansen

H. A. F. M.

Heise

Watts

C. H.

Garcia-Moreno

, & WHO Multi-Country Study on Women’s Health and Domestic Violence Against Women Study Team. (2008). Intimate partner violence and women’s physical and mental health in the WHO multi-country study on women’s health and domestic violence: An observational study. Lancet, 371, 1165-1172.

Garcia-Moreno

Jansen

H. A. F. M.

Ellsberg

Heise

Watts

(2005). WHO multi-country study on women’s health and domestic violence against women. Geneva, Switzerland: World Health Organization.

10.

Hughes

Bellis

M. A.

Jones

Wood

Bates

Eckley

. . . Officer

(2012). Prevalence and risk of violence against adults with disabilities: A systematic review and meta-analysis of observational studies. Lancet, 379, 1621-1629.

11.

Hughes

R. B.

Lund

E. M.

Gabrielli

Powers

L. E.

Curry

M. A.

(2011). Prevalence of interpersonal violence against community-living adults with disabilities: A literature review. Rehabilitation Psychology, 56, 302-319.

12.

Human Rights Watch. (2010). “As if we weren’t human”: Discrimination and violence against women in Northern Uganda. New York, NY: Author.

13.

Jack

D. C.

Astbury

(2013). Overcoming discrimination, persecution and violence against women. In Mustakove-Possardt

Lyubansky

Basseches

Oxenberg

(Eds.), Towards a socially responsible psychology for a global era (pp. 207-230). New York, NY: Springer.

14.

Jansen

H. A. F. M.

Watts

Ellsberg

Heise

Garcia-Moreno

(2004). Interviewer training in the WHO multi-country study on women health and domestic violence. Violence Against Women, 10, 831-849.

15.

Royal Government of Cambodia. (2009). Law on the Protection and the Promotion of Rights of Persons with Disabilities. Phnom Penh, Cambodia: Kingdom of Cambodia.

16.

Royal Government of Cambodia. (2005). Law on The Prevention of Domestic Violence and the Protection of Victims. Phnom Penh, Cambodia: Kingdom of Cambodia.

17.

Ministry of Planning. (2010). Achieving Cambodia’s Millennium Development Goals: An update. Phnom Penh, Cambodia: Cambodian MOP.

18.

Ministry of Women’s Affairs. (2005). Violence against Women: 2005 Baseline Survey. Phnom Penh, Cambodia: Cambodian MOWA.

19.

Ministry of Women’s Affairs. (2010). Violence against women: 2009 Follow-up survey (Final study report). Phnom Penh, Cambodia: Cambodian MOWA. Author

20.

Miranda

J. J.

Patel

(2005). Achieving the Millennium Development Goals: Does mental health play a role? PLoS Medicine, 2(10), e291.

21.

Mohapatra

Mohanty

(2005). Abuse and activity limitation: A study on domestic violence against disabled women in Orissa, India. Orissa, India: Swabhiman.

22.

Nathanson

A. M.

Shorey

R. C.

Tirone

Rhatigan

D. L.

(2012). The prevalence of mental health disorders in a community sample of female victims of intimate partner violence. Partner Abuse, 3, 59-75.

23.

National Institute of Public Health, National Institute of Statistics Cambodia and ORC MACRO. (2006). Cambodia Demographic and Health Survey 2005. Phnom Penh, Cambodia and Calverton, Maryland, USA: Cambodian NIPH, NIS and ORC Macro.

24.

Nosek

M. A.

Clubb Foley

Hughes

R. B.

Howland

C. A.

(2001). Vulnerabilities for abuse among women with disabilities. Sexuality and Disability, 19, 177-190.

25.

Saxton

Curry

M. A.

Powers

L. E.

Maley

Eckels

Gross

(2001). “Bring me my scooter so I can leave you”: A study of disabled women handling abuse by personal assistance providers. Violence Against Women, 7, 393-412.

26.

Scholte

W. F.

Verduin

van Lammeren

Rutayisire

Kamperman

A. M.

(2011). Psychometric properties and longitudinal validation of the Self-Reporting Questionnaire (SRQ-20) in a Rwandan community setting: A validation study. BMC Medical Research Methodology, 16(11), Article 116. doi:10.1186/1471-2288-11-116

27.

Stubbs

Tawake

(2009). Pacific sisters with disabilities: At the intersection of discrimination. Suva, Fiji: United Nations Development Program Pacific Centre.

28.

Ullman

S. E.

Filipas

H. H.

(2001). Predictors of PTSD symptom severity and social reactions in sexual assault victims. Journal of Traumatic Stress, 14, 369-389.

29.

United Nations. (2006). Convention on the Rights of Persons With Disabilities and its Optional Protocol. New York, NY: United Nations General Assembly.

30.

United Nations Development Program. (2010). The real wealth of nations: Pathways to human development (Human Development Report 2010-20th Anniversary Edition). New York, NY: UNDP.

31.

United Nations Development Program. (2011). Sustainability and Equity: A better future for all (Human Development Report 2011). New York, NY: UNDP.

32.

United Nations Development Program Cambodia, Ministry of Justice, & Ministry of Interior. (2010). Talking about domestic violence: A handbook for village facilitators. Phnom Penh, Cambodia: United Nations Development Program Cambodia and Agencia Espanola de Cooperacion Internacional para el Desarrollo.

33.

Vos

Astbury

Piers

L. S.

Magnus

Heenan

Stanley

Webster

(2006). Measuring the impact of intimate partner violence on the health of women in Victoria, Australia. Bulletin of the World Health Organization, 84, 739-744.

34.

Washington Group. (2006). Short question set. Hyattsville, MD: National Center for Health Statistics.

35.

World Health Organization. (2008). The global burden of disease 2004 update. Retrieved from http://www.who.int/healthinfo/global_burden_disease/GBD_report_2004update_full.pdf

36.

World Health Organization. (2011). Mental Health Atlas 2011. Geneva, Switzerland: Department of Mental Health and Substance Abuse, World Health Organization.

37.

World Health Organization and World Bank. (2011). World report on disability. Geneva, Switzerland: World Health Organization.