The Prevalence of Interpersonal Violence (IPV) Against Women and its Associated Variables: An Exploratory Study in the Rongo Sub-County of Migori County,Kenya

Abstract

Interpersonal violence (IPV) within intimate partner relationships and gender-based violence remain major public health problems worldwide; 44.8% of Kenyan women have reported experiencing IPV beginning after the age of 15 years (National Bureau of Statistics Nairobi, Kenya, 2015). Combatting IPV and its sequelae is integral to promoting gender equality, a key target of the Sustainable Development Goals. We quantify the lifetime prevalence of IPV among women in two rural Kenyan communities, as well as factors associated with IPV in this area, such as educational attainment and severe depression. We conducted a cross-sectional population-based survey of households in the North and East Kamagambo wards of Migori County, Kenya in May 2018. A questionnaire regarding IPV was given to female respondents. Group-wise comparisons and multiple logistic regression analyses were performed to describe community prevalence and factors associated with IPV against women. A total of 873 women completed questions about IPV, representing a population estimate of 11,252 women in the study area. Lifetime IPV prevalence in the study area was 60.3%. Variables associated with IPV included involvement in a polygamous marriage (adjusted odds ratio [aOR]: 1.81; 95% confidence interval [CI]: [1.13, 2.88]; p = .013), attaining six or fewer years of education (aOR: 1.84; 95% CI: [1.27, 2.66]; p = .001), and IPV exposure in girlhood (aOR: 1.59; 95% CI: [1.12, 2.28]; p = .011). IPV was independently associated with experience of emotional abuse (OR: 11.22; 95% CI: [7.02, 17.95]; p < .001) and severe depression (OR: 3.51; 95% CI: [1.03, 11.97]; p = .045). Violence against women is a public health emergency in Migori County, Kenya. Low educational attainment, IPV exposure in girlhood, and polygamy were significantly associated with experience of IPV. Our results provide hyper-local data necessary for targeted interventions and generalizable data with sampling methods for use by other implementing organizations in sub-Saharan Africa.

Keywords

interpersonal violence resource-limited setting gender equity Kenya

Introduction

In recent decades, increasing awareness of interpersonal violence (IPV) against women and its consequences have become a larger foci of global and public health interventions (Brown et al., 2007; Mikton et al., 2017). The Sustainable Development Goals (SDGs) have prioritized gender equality and freedom from violence alongside, and as a function of, other public health crises (United Nations, 2015). Women who experience IPV are more likely to attempt suicide and have a higher risk of death by suicide (Devries et al., 2011). Globally, nearly 40% of murdered women, compared to only 6.3% of murdered men, are killed by an intimate partner (Stöckl et al., 2013). Women who experience IPV within intimate partner relationships are more likely to have unwanted pregnancies and abortions, are less likely to deliver in a skilled facility (Goo & Harlow, 2012; Hindin, 2008), and are more likely to give birth to low birth weight babies (Ferdos & Rahman, 2017; Plichta, 2004). Furthermore, children of women who are abused are less likely to be fully vaccinated and more likely to experience nutritional stunting than those who grow up in violence-free households (Hindin, 2008).

Clear links have been established between HIV risk and the experience of IPV (Kouyoumdjian, Calzavara, et al., 2013; Kouyoumdjian, Findlay, et al., 2013; Olowookere et al., 2015; Vest et al., 2002). In South Africa, women who experience IPV have shown to be 48% more likely to contract HIV than women who do not experience partner violence (R. K. Jewkes et al., 2010). IPV perpetuates societal norms that disadvantage and disempower women, normalizes both relationship violence and forced sex, and decreases a woman’s ability to negotiate condom use. Women in violent relationships often lack the autonomy to manage their own health care, implicating their ability to seek HIV testing (Kouyoumdjian, Findlay, et al., 2013). Furthermore, the experience of IPV may contribute to women engaging in risk behaviors associated with HIV transmission, such as transactional sex or intravenous drug use (Joyner et al., 2015). Male partners of women in violent relationships have been shown to be less likely to be involved in antenatal care, which may pose consequences for prenatal service utilization and mother-to-child transmission of HIV (Kiarie et al., 2006).

Kenya is a country with an estimated 1.5 million people living with HIV and was ranked 137 of 159 countries on the United Nations Development Program 2017 Gender Inequality Index (United Nations Development Programme, 2017). Forty-four percent of Kenyan women nationwide experience physical violence at or after the age of 15 years. In Migori County, 57% of women have experienced physical violence (National Bureau of Statistics Nairobi, Kenya, 2015). While Kenyan men are also victims of IPV, most report their offenders are other men. Neither the 2006 Kenyan Sexual Offenses Act nor the 2015 Kenyan Protection Against Domestic Violence Bill address or criminalize spousal rape (United Nations Committee on the Elimination of Discrimination Against Women, 2011).

Migori County (Figure 1) in south-western Kenya has historically been under-resourced and experiences some of the country’s poorest health outcomes (National Bureau of Statistics Nairobi, Kenya, 2015). In both greater Kenya and Migori County, women are less educated than men and those in rural communities are less educated than those in urban areas. Migori County women have a median of 6.8 years of education, compared with 7.2 years for men in the same area, and with 11 years for women in Nairobi (National Bureau of Statistics Nairobi, Kenya, 2015). The Nyanza region, in which Migori County is located, sees the second highest rates of polygyny in Kenya, with 11.8% of men having multiple wives (National Bureau of Statistics Nairobi, Kenya, 2015). In 2017, Nyanza had the highest HIV prevalence in the country at 15.1%, and 70% of residents living below US$2 per day (African Institute for Development Policy, 2017). Maternal and under-5 mortality rates in the area are higher than national averages at 673/100,000 births and 82/1,000 live births, respectively (African Institute for Development Policy, 2017). Rongo sub-county, one of six official divisions of Migori County, is further divided into four wards, with North Kamagambo and its population of approximately 30,000 people receiving services from the Lwala Community Alliance (Lwala) hospital and community health worker program since 2007. In 2018, Lwala expanded service delivery and programming to neighboring East Kamagambo (population approximately 35,000), with similar rollouts planned for South Kamagambo 6 months following data collection and in Central Kamagambo in 2020.

In preparation for service expansion, Lwala conducted a population-based household survey in 2018 to provide ongoing collection of priority health metrics in their original catchment area of North Kamagambo, as well as baseline data prior to service expansion in East Kamagambo. Defining IPV as experiencing violence at the hands of “a partner or community member with whom they had a close relationship,” our objective in this exploratory study was to estimate the IPV prevalence among women in two rural areas of Rongo sub-county across the life course. Furthermore, we, examine variables related to IPV victimization such as education and involvement in polygamous marriages to inform potential programmatic interventions.

Method

Study Setting and Sampling

This cross-sectional population-based household survey was conducted in the North and East Kamagambo wards of Rongo sub-county, Migori County, Kenya in May 2018. Details of the sampling methodology and household selection are presented elsewhere (Henderson et al., 1973; Starnes et al., 2018). Briefly, given the emphasis on maternal and child health, the survey’s sample size was determined using under-5 mortality as a binary outcome using a binomial test to compare one proportion to a reference value. Using a power of.85 and alpha of .05, a sample size of 270 households in North Kamagambo and 868 in East Kamagambo was needed to detect a reduction of 40 per 1,000 live births from the rate of 82 per 1,000 live births reported in Nyanza Province (National Bureau of Statistics Nairobi, Kenya, 2015). Given that this survey represents baseline data collection for East Kamagambo, the number of households was oversampled.

A proportional stratified sampling method was used to give all households equal chance of selection despite differing population densities. The geographic boundaries of North and East Kamagambo wards were mapped. The wards were then subdivided into 39 grid cells in North Kamagambo and 124 grid cells in East Kamagambo, with a target sampling goal of seven households to be surveyed per grid cell.

Figure 1.

Map of Migori County, Kenya, showing the location and divisions of Rongo Sub-County.

Survey Questionnaire

The survey contained more than 300 questions across multiple domains and was modeled on validated tools, including the Kenya Demographic and Health Survey (Kenya DHS; National Bureau of Statistics Nairobi, Kenya, 2015). The survey was designed to obtain population-based data for indicators across multiple areas, including child mortality, nutrition, family planning, HIV, water and sanitation, economics, education, and IPV. Items for the IPV section were adapted from two clinically validated screening tools for partner violence: the Abuse Assessment Screen (AAS; Rabin et al., 2009) and the Partner Violence Screen (PVS; Soeken et al., 1998), as well as from the Spousal Violence questionnaire from the Kenyan DHS. Mental health was scored on the PHQ-8 and categorized by cut-off values into five groups of depression status/severity (Kroenke et al., 2009). Demographic, health, and socioeconomic data were captured about the respondent and household. The survey was piloted with small groups of respondents in both North Kamagambo and East Kamagambo to ensure applicability and comprehension.

Study Population

This study was conducted in partnership with local people in Rongo sub-county. One interview per household was conducted with the designated head of household, with preference given to the female head-of-household when available. Surveys were administered by trained enumerators employed by Lwala for the survey period. All were hired from the community and were fluent in English, as well as Dholuo, and Swahili. The survey tool was developed using a REDCap electronic data capture tool hosted at Vanderbilt University (Harris et al., 2009). Surveys were administered on tablets during a face-to-face interview using the REDCap mobile application to reduce data entry errors and ensure appropriate skip logic. Before beginning the IPV questionnaire, enumerators separated women from all other members of the household and conducted the remainder of the survey privately. IPV questions were not visible on the surveying tablet until the enumerator was prompted to confirm the woman was alone and willing to answer IPV questions. All women aged 18 years and older at the time of data collection, for whom the IPV instrument was completed, were included in our analysis.

IPV Outcome and Covariates

We reviewed the literature to determine variables associated with IPV. These variables included younger age, marital status and involvement in polygamy, religion, educational attainment, depressive symptoms, and HIV status and testing history (Kouyoumdjian, Findlay, et al., 2013; Stöckl et al., 2014; Sutton & Dawson, 2018; Vest et al., 2002). Respondents were positively indicated for experience of IPV if they answered “yes” to one or more questions regarding experience of physical and/or sexual violence by a partner or community member with whom they had a close relationship (Rabin et al., 2009; Soeken et al., 1998). A positive emotional abuse screen was constituted by an affirmative response to questions about being intentionally humiliated or threatened by a partner or loved one. Exposure to IPV during childhood was indicated during a question about having seen violence in one’s household growing up, such as between parents. Depression was scored on the Personal Health Questionnaire Depression Scale (PHQ-8; Kroenke et al., 2009). The PHQ-8 is a shortened version of the Patient Health Questionnaire (PHQ-9; Spitzer et al., 1999) which has been validated in rural African settings in previous studies (Audet et al., 2018). It is scored from 0 to 24 and categorized into 5 levels of depressive symptoms. The validated cut-off point for detecting depression is 10. If the total score is equal to or more than 10, the participant is screened as positive for depression (Kroenke et al., 2009; Smith et al., 2010). The PHQ-8 omits the ninth question, an item regarding suicidal ideation, as emergency care could not be guaranteed for those indicating suicidal behavior (Audet et al., 2018).

Statistical Analysis

Descriptive statistics and group-wise comparisons were used to delineate characteristics of residents in North Kamagambo and East Kamagambo, as well as to compare women who had experienced IPV to those who had not. The group-wise comparisons included Pearson’s chi-square for categorical variables and t-tests for continuous variables. Logistic regression was used to model factors associated with experience of IPV. To avoid overfitting, the reported model of five variables chosen by clinical expertise was trimmed from an initial model of 10 variables. Emotional abuse was excluded from the regression model to avoid violating the assumption of multicollinearity. To generalize the information to the population of East and North Kamagambo, weights were developed and used for the regression models. The weights were calculated for each community using the inverse of the probability of being sampled: [number of households sampled/total number of households]-¹. Data were managed in R version 3.5.2 (R Core Team, 2018) and analyzed using “svy” commands in Stata version 15.1 (StataCorp LP, College Station, TX) to account for the survey design.

Ethical Approval and Consent to Participate

This study was approved by the Ethics and Scientific Review Committee at AMREF Health Africa (No. P452/2018) and the Institutional Review Board at Vanderbilt University Medical Center (No. 161396). Written informed consent was obtained from all participants. Approval was also obtained from the local area chief through the Office of the Deputy County Commissioner and Ministry of Health.

Results

The survey was administered to 1,088 individuals living in North or East Kamagambo, Migori County, Kenya, of which 913 women fit the inclusion criteria. In total, 873 women, representing a population estimate of 11,252 women over the age of 18 years, living in North or East Kamagambo completed the IPV instrument and were included in this analysis. The average age of respondents was 32.54 years with the majority of women (73.8%) being in monogamous marriages. Just less than 71% had completed six or fewer years of education, and 99.3% of these women reported ever having been tested for HIV. The majority of the sample (60.3%) reported having experienced IPV in their lifetime. There was a higher proportion of women who were in a polygamous marriage, had 6 or fewer years of education, been exposed to IPV in girlhood, and who screened positive for depression among the group that had experienced IPV versus those who had not; 20.9% versus 11.2% (p = .001), 76.4% versus 62.5% (p < .001), 66.9% versus 59.9% (p = .035) and 27.2% versus 19.6% (p = .024), respectively (Table 1).

Table 1.

Demographic Characteristics for Female Head-of-Household in North and East Kamagambo by IPV Status, Unweighted.

Characteristics	IPV−	IPV+	Combined	p value^a
Characteristics	347 (39.7%)	526 (60.3%)	873	p value^a
Age (M in years ± SD)	32.97 ± 12.18	32.26 ± 10.66	32.54 ± 11.29	.367
Age group (n, %)				.118
18–24	89 (25.6%)	124 (23.6%)	213 (24.4%)
25–29	214 (61.7%)	355 (67.5%)	569 (65.2%)
50+	44 (12.7%)	47 (8.9%)	91 (10.4%)
Marital Status (n, %)				.001
Married—Monogamous	270 (77.8%)	374 (71.1%)	644 (73.8%)
Married—Polygamous	39 (11.2%)	110 (20.9%)	149 (17.1%)
Other	38 (11.0%)	42 (8.0%)	80 (9.2%)
Highest level of education (n, %)				<.001
0–6 years	217 (62.5%)	402 (76.4%)	619 (70.9%)
7+ years	130 (37.5%)	124 (23.6%)	254 (29.1%)
Religious Group (n, %)				.024
Seventh-Day Adventist	150 (43.2%)	172 (32.7%)	322 (36.9%)
Catholic	54 (15.6%)	85 (16.2%)	139 (15.9%)
Protestant	53 (15.3%)	108 (20.5%)	161 (18.4%)
Roho Church	73 (21.0%)	127 (24.1%)	200 (22.9%)
Others	17 (4.9%)	34 (6.5%)	51 (5.8%)
Ever HIV tested (n, %)				.246
No	1 (0.3%)	5 (1.0%)	6 (0.7%)
Yes	346 (99.7%)	521 (99.0%)	867 (99.3%)
Girlhood exposure to IPV (n, %)				.035
No	139 (40.1%)	174 (33.1%)	313 (35.9%)
Yes	208 (59.9%)	352 (66.9%)	560 (64.1%)
Has ever experienced emotional abuse (n, %)				<.001
No	316 (91.1%)	291 (55.3%)	607 (69.5%)
Yes	31 (8.9%)	235 (44.7%)	266 (30.5%)
Feels safe in current relationship (n, %)				.174
Yes	271 (78.1%)	395 (75.1%)	666 (76.3%)
No	66 (19.0%)	102 (19.4%)	168 (19.2%)
No response	10 (2.9%)	29 (5.5%)	39 (4.5%)
PHQ-8 (depression score; n, %)				.024
None/minimal (0–4)	134 (38.6%)	169 (32.1%)	303 (34.7%)
Mild (5–9)	145 (41.8%)	214 (40.7%)	359 (41.1%)
Moderate (10–14)	42 (12.1%)	83 (15.8%)	125 (14.3%)
Moderately Severe (15–19)	22 (6.3%)	38 (7.2%)	60 (6.9%)
Severe (20–24)	4 (1.2%)	22 (4.2%)	26 (3.0%)
Has ever used Lwala Services (n, %)				.566
No	238 (68.6%)	351 (66.7%)	589 (67.5%)
Yes	109 (31.4%)	175 (33.3%)	284 (32.5%)

Note. IPV = interpersonal violence; PHQ = Personal Health Questionnaire Depression Scale. Frequencies rounded to nearest whole number; percentages rounded to nearest tenth.

Chi-square p values for categorical and t test for M/SD.

Looking at independent associations with IPV using a univariate weighted logistic regression, it was found that women who screened positive for severe depression had more than a threefold higher reported experience of IPV (OR: 3.51; 95% CI: [1.03, 11.97]; p = .045). The multivariate logistic regression aimed to find how possible variables associated with IPV were impacted by other characteristics; it found that women who were exposed to IPV in girlhood had higher odds of IPV compared to the those who did not (aOR: 1.59; 95% CI: [1.12, 2.28]; p = .011) and women in polygamous marriages had higher odds of experiencing IPV compared to women in monogamous marriages (aOR: 1.81; 95% CI: [1.13, 2.88]; p = .013). In addition, women who completed 6 or fewer years of school had higher odds of experiencing IPV than those who had finished 7 or more years (aOR: 1.84; 95% CI: [1.27, 2.66]; p = .001; Table 2).

Table 2.

Logistic Regression Model for Predictors of IPV, Including Sexual Violence, Weighted.

Variables	Unadjusted odds ratios			Adjusted odds ratios
Variables	OR	95% CI	p value	OR	95% CI	p value
Region
East Kamagambo	Ref
North Kamagambo	1.19	[0.82, 1.72]	.368
Age group
18–24	Ref
25–49	1.01	[0.67, 1.51]	.980
50+	0.55	[0.30, 1.00]	.051
Religion
Seventh-day adventist	Ref			Ref
Catholic	1.14	[0.68. 1.93]	.621	0.96	[0.56, 1.65]	.894
Protestant	1.49	[0.93, 2.38]	.098	1.20	[0.74, 1.95]	.461
Roho Church	1.68	[1.07, 2.64]	.023	1.32	[0.83, 2.08]	.238
Others	1.75	[0.84, 3.63]	.133	1.75	[0.79, 3.90]	.168
Experience of emotional abuse
No	Ref
Yes	11.22	[7.02, 17.95]	<.001
Marital status
Married—monogamous	Ref			Ref
Married—polygamous	1.98	[1.26, 3.11]	.003	1.81	[1.13, 2.88]	.013
Other	0.76	[0.41, 1.41]	.380	0.68	[0.37, 1.23]	.201
Highest level of education
7+ years	Ref			Ref
0–6 years	2.07	[1.43, 2.98]	<.001	1.84	[1.27, 2.66]	.001
IPV exposure in girlhood
No	Ref			Ref
Yes	1.69	[1.19, 2.40]	.003	1.59	[1.12, 2.28]	.011
Feels safe in current relationship
Yes	Ref
No	0.98	[0.63, 1.50]	.908
No answer	3.15	[1.40, 7.11]	.006
Ever HIV tested
No	Ref
Yes	0.31	[0.03, 2.82]	.299
Lwala use
Yes	Ref
No	0.74	[0.52, 1.06]	.101
PHQ 8 (depression score)
None/minimal	Ref			Ref
Mild	1.36	[0.92, 2.00]	.121	1.26	[0.85, 1.85]	.251
Moderate	1.63	[0.95, 2.80]	.076	1.44	[0.82, 2.52]	.203
Moderately severe	1.46	[0.75, 2.84]	.262	1.13	[0.56, 2.26]	.736
Severe	3.51	[1.03, 11.97]	.045	2.38	[0.65, 8.69]	.190
Afraid partner would refuse condom use
No	Ref
Yes	1.32	[0.93, 1.87]	.121

Note. IPV = interpersonal violence; OR = odds ratio; CI = confidence interval; PHQ = Personal Health Questionnaire Depression Scale. Lwala: Lwala Community Alliance.

Discussion

This work provides a much-needed examination of IPV in Migori County, an area in which metrics of gender inequality are much higher than that of Kenya at large (United Nations Office for the Coordination of Humanitarian Affairs, 2020). Data collected by large national research projects are unable to explore the nuanced relationship of these variables in context, rendering them less useful to design and implement targeted programmatic responses. While this population is largely understudied, the relationship between the researchers and a local community-led NGO allows these data to be translated into meaningful and hyper-local intervention that draws not only on the issues uncovered by this research but also on the talents and resources of the community itself.

Overall, the prevalence of having experienced IPV in one’s lifetime was 60.3%, higher than women in Kenya and across SSA (Ahinkorah et al., 2018). The Kenya Demographic Health Survey revealed that 45% of women aged 15 to 49 years have experienced either physical or sexual violence during their lives, while 31% of women were living with active violence in their homes (National Bureau of Statistics Nairobi, Kenya, 2015). While we do not know why this region has a higher prevalence of IPV, western Kenya is an extremely poor region with higher rates of HIV, potentially increasing the stress within intimate partner relationships. As in other countries, being in a polygamous marriage and lower educational attainment were associated with increased odds of experiencing IPV, likely given women’s dependence on their male partners due to financial or cultural expectations (Tanimu et al., 2016; Vest et al., 2002). As found in other studies, women who reported an experience of IPV were more likely to report the experience of emotional abuse and were more likely to suffer from severe depression (White & Satyen, 2015).

National data indicate that Nyanza Province has the second highest rates of polygamy in Kenya, with 11.8% of men reporting having two or more wives (National Bureau of Statistics Nairobi, Kenya, 2015). This figure was higher in our population, in which 17.1% of respondents indicated being in a polygamous marriage. Despite our findings indicating that women in polygamous marriages have an almost doubled likelihood of experiencing IPV, few studies examine polygamy as associated with IPV (Bostock et al., 2009). This gap is consequential as it suggests a possible underestimation of IPV prevalence, not only in our findings, but in population-based surveys more broadly. While not directly explored in our study, one can assume that if a male partner is committing violence against one wife, then there is a high probability the other wives are also experiencing similar violence, though this needs to be investigated further. Moreover, our survey questions were designed to ascertain exposure to IPV across the life course and allowed for possibilities that the experienced violence was from other close relations or partners and did not specifically ask if the experienced violence came from one’s current spouse. As such, more in-depth understanding of this association is crucial going forward and for the appropriate development of interventions to address the problem. As regional marriage tradition in the region encourages women to move to their husband’s village, thus removing them from the support system of their family, future work should address the role of land ownership and isolation from support systems in women’s vulnerability to abuse.

While the vast majority of women in our population who reported experiences of IPV were married (90.9%), most women (64.3%) in a 2014 national study who reported experiencing violence were divorced or separated (National Bureau of Statistics Nairobi, Kenya, 2015). This suggests a potential gap in support services and resources in Migori County to help women make decisions about whether or not to stay in an abusive relationship. Local leadership’s historic inaction regarding IPV may lead affected women to feel that there is no support from the traditional “first line” of response in issues of IPV. Furthermore, families’ frequent unwillingness to accept married daughters back into their homes may leave women facing violence with little support and no place to seek safety. Nationally, this phenomenon may be compounded by policy and legislation that delegitimizes violence within relationships, as is reflected in the country’s lack of criminalization of spousal rape and continued allowance of sexual offenders to prosecute their victims (United Nations Committee on the Elimination of Discrimination Against Women, 2011).

Although this survey was heavily modeled on the Kenya DHS, there are some pertinent differences. The Kenya DHS does not use experience of sexual violence as a contributor to overall IPV, while our survey constituted IPV experience as either sexual or physical violence. Furthermore, the Kenya DHS examines IPV starting at the age of 15 years, while this project examined IPV across the life course. Given Migori County women marry, on average, earlier than women in other parts of Kenya (National Bureau of Statistics Nairobi, Kenya, 2015), we thought it is necessary to assess IPV with no minimum age to allow for experiences of physical and sexual violence at the hands of partners that occurred younger than the age of 15 years.

For women in our study, lower educational attainment was associated with reported experience of IPV, consistent with other literature that cites education as a mechanism of social support, social empowerment, and economic independence (R. K. Jewkes et al., 2010). Higher education has been associated with more liberal attitudes about women and less tolerance of female subservience (R. Jewkes, 2002). The relationship between education and IPV becomes further complicated in contexts where young adolescents are coerced into sexual relationships with older men to fund their education (Ziraba, et al., 2018; Njue, et al., 2011). As current school-aged girls who may be affected by this phenomenon were not captured in our data collection, further study of coerced sex as a means of funding education should examine potential associations with educational attainment and IPV.

We found that more than two-thirds of women who had witnessed IPV in girlhood had experienced some sort of IPV themselves. This is consistent with literature that suggests girls who observe violent relationships in childhood are more likely to be involved in violent partnerships later in life (Abramsky et al., 2011; Kouyoumdjian, Calzavara, et al., 2013).

Existing literature documents educational attainment as protective against IPV later in life (R. Jewkes, 2002; R. K. Jewkes et al., 2010; R. Jewkes et al., 2002). At the individual level, educated girls are more likely to become financially independent and have more liberal ideas about their roles and positions in society. Furthermore, schooling provides powerful social networks for girls, as well as promoting self-confidence and providing access and use of relevant information and skills (R. Jewkes, 2002). At the societal level, areas that prioritize girls’ education place higher value and social power on women create environments that are less permissive of IPV (Dunkle et al., 2004; R. Jewkes et al., 2002). Thus, our findings regarding higher rates of IPV among less educated women present a significant opportunity to further reduce risk of IPV experience. By educating girls and providing them skilled employment opportunities, women would have reduced reliance on unhealthy relationships as adults. Further work should approach education as a potential twofold prevention mechanism against violent partnerships: not only is education itself is protective, but IPV can be explicitly addressed with male and female students to break the cycle of violence.

We found a significant, independent relationship between experience of IPV and severe depression. Furthermore, our study indicates that 65.3% of women in this population suffer from depressive symptoms. While the cross-sectional design of our survey limits us from studying direction of causality between these variables, it is consistent with other studies that cite high rates of depressive symptoms among women who have reported experience of IPV (Joyner & Mash, 2012; Wong et al., 2008). Given the concerning lack of depression data from the African continent at large (Mitchell et al., 2016), this study highlights the largely unaddressed issues of IPV and mental health in low-income settings.

The primary limitation of this study is that it is cross-sectional in nature. Also, given the sensitivity of our IPV questions, we must acknowledge the potential for underreporting due to social desirability response bias (Saunders, 1991), as well as a means of self-deception (Riggs et al., 1989; Visschers et al., 2017). Finally, while every attempt was made to create a safe, private environment with culturally sensitive questions to elicit responses about one’s personal IPV experiences, another limitation may have been the format of the questioning itself. Such questioning may be more effective if done away from the home and utilizing more open-ended style questions, in which each woman can better describe her own experiences.

Conclusion

Given the established high HIV prevalence in our study population, our findings mirror literature that demonstrates a “triple epidemic” of IPV, HIV, and depression among African women (Mitchell et al., 2016; National Bureau of Statistics Nairobi, Kenya, 2015). Given the established relationships between IPV and poor maternal/child health outcomes, the results of this work can inform Migori County officials and development actors like Lwala to integrate IPV educational and support services into their current programming.

IPV is a public health emergency in rural Kenya. Data from Migori County illuminate polygamy, girlhood IPV exposure, and low educational attainment as factors associated with women’s experience of IPV. Furthermore, there is a robust relationship between experience of IPV and severe depression. This hyper-local data equip Lwala to address the variables related to, as well as the consequences of IPV in their wraparound community health model, positioning them as a pioneer in holistic women’s health.

Footnotes

Acknowledgements

We would like to acknowledge Lwala Community Alliance for facilitating this research, as well as the Vanderbilt Institute for Global Health for their support of this study. Additionally, we are thankful for the technical support of the REDCap team at Vanderbilt University Medical Center and the Vanderbilt Department of Biostatistics. Our most sincere thanks to the enumerators who conducted data collection, the Lwala Monitoring and Evaluation team who oversaw the surveying project, and the community members of North and East Kamagambo.

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) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Troy D. Moon

Author Biographies

Miller Morris, MA, MPH, received her Master’s in the Social Determinants of Health from Vanderbilt University, followed by a Master’s of Public Health with a concentration in Global Health at the Vanderbilt University School of Medicine. Her research focuses on interpersonal violence and its implications on community health, as well as implementation science of community-led, multidisciplinary efforts. She is interested in examining durable public health challenges as a function of power dynamics and gender inequity. She hopes her research will inform not only community-level practices but also support policy-level change and improvement and in identifying mechanisms to translate research into practice.

Vincent Okoth is currently the monitoring and evaluation manager at Lwala Community Alliance. He designs, operationalizes, and manages all of Lwala’s data systems, from the individualized data Community Health Workers collect at households to the program evaluation data Lwala collects through robust surveys. He loves that his work creates impact through data.

Heather L. Prigmore, MPH, is a biostatistician in the Department of Biostatistics at Vanderbilt University Medical Center. Her research currently focuses on Global Health, specifically the HIV epidemic in Africa. She is interested in applying statistical concepts to medicine and health services to influence policy and improve practice by helping researchers understand their data.

Daniele J. Ressler, MS, MPA, is impact director of Lwala Community Alliance, responsible for organizational research, learning, M&E, and impact. Her research interests include identifying and improving drivers of intermediate and long-term change associated with Lwala’s innovative programming in community health and clinical health systems delivery. She is also interested in the role of m-health and electronic data within community-based public health and strategies of effective scaling of community-driven programs.

Julius Mbeya, MA, is managing director at Lwala Community Alliance. In this role, he overseas innovative community-based programming that help address challenges affecting children and women including gender-based violence. He is passionate about the transforming power of community-driven development and the lasting change that this brings about.

Ash Rogers, MPA, is the executive director of Lwala Community Alliance, leading vision, strategy, and development. Prior to Lwala, she was the Director of Operations at Segal Family Foundation, overseeing a US$12 million portfolio of more than 180 grantee organizations in Sub-Saharan Africa. She has worked with organizations including U.S. State Department, Komo Learning Centers, and HELP International—with a common focus on maternal, child, reproductive health, and rights.

Troy D. Moon, MD, MPH, is an associate professor in the Department of Pediatrics, Division of Infectious Diseases, and of the Vanderbilt Institute for Global Health. His work has focused on building medical, public health, and research capacity in sub-Saharan Africa and he has 11 years of experience implementing large population-based household surveys in resource-limited settings.

Carolyn M. Audet, PhD, is an assistant professor in the Department of Health Policy and the Assistant Director of the Center for Clinical Quality and Implementation Research at Vanderbilt University. Her research focuses on developing community and family-based interventions to improve health outcomes among those living with HIV in sub-Saharan Africa.

References

Abramsky

Watts

C. H.

Garcia-Moreno

Devries

Kiss

Ellsberg

… Heise

(2011). What factors are associated with recent intimate partner violence? Findings from the WHO multi-country study on women’s health and domestic violence. BMC Public Health, 11(1), Article 109. https://doi.org/10.1186/1471-2458-11-109

African Institute for Development Policy. (2017). Reproductive, maternal, neonatal and child health: Migori County. https://www.afidep.org/publication/reproductive-maternal-neonatal-and-child-health-migori-county/

Ahinkorah

B. O.

Dickson

K. S.

Seidu

A. A.

(2018). Women decision-making capacity and intimate partner violence among women in sub-Saharan Africa. Archives of Public Health, 76, Article 5. https://doi.org/10.1186/s13690-018-0253-9

Audet

C. M.

Wainberg

M. L.

Oquendo

M. A.

Blevins Peratikos

Duarte

C. S.

… Moon

T. D.

(2018). Depression among female heads-of-household in rural Mozambique: A cross-sectional population-based survey. Journal of Affective Disorders, 227, 48–55. https://doi.org/10.1016/j.jad.2017.10.022

Bostock

Plumpton

Pratt

(2009). Domestic violence against women: Understanding social processes and women’s experiences. Journal of Community & Applied Social Psychology, 19(2), 95–110. https://doi.org/10.1002/casp.985

Brown

Butchart

Harvey

(2007). Third milestones of a global campaign for violence prevention report 2007. Scaling up. World Health Organization Press. https://apps.who.int/iris/bitstream/handle/10665/43671/9789241595476_eng.pdf; jsessionid=31729AC0390A5F2D41B9395741841312?sequence=1

Devries

Watts

Yoshihama

Kiss

Schraiber

L. B.

Deyessa

… Garcia-Moreno

(2011). Violence against women is strongly associated with suicide attempts: Evidence from the WHO multi-country study on women’s health and domestic violence against women. Social Science & Medicine, 73(1), 79–86. https://doi.org/10.1016/j.socscimed.2011.05.006

Dunkle

K. L.

Jewkes

R. K.

Brown

H. C.

Gray

G. E.

McIntryre

J. A.

Harlow

S. D.

(2004). Gender-based violence, relationship power, and risk of HIV infection in women attending antenatal clinics in South Africa. The Lancet, 363(9419), 1415–1421. https://doi.org/10.1016/S0140-6736(04)16098-4

Ferdos

Rahman

M. M.

(2017). Maternal experience of intimate partner violence and low birth weight of children: A hospital-based study in Bangladesh. PLOS ONE, 12(10), Article e0187138. https://doi.org/10.1371/journal.pone.0187138

10.

Goo

Harlow

S. D.

(2012). Intimate partner violence affects skilled attendance at most recent delivery among women in Kenya. Maternal and Child Health Journal, 16(5), 1131–1137. https://doi.org/10.1007/s10995-011-0838-1

11.

Harris

P. A.

Taylor

Thielke

Payne

Gonzalez

Conde

J. G.

(2009). Research electronic data capture (REDCap)—A metadata-driven methodology and workflow process for providing translational research informatics support. Journal of Biomedical Informatics, 42(2), 377–381. https://doi.org/10.1016/j.jbi.2008.08.010

12.

Henderson

R. H.

Davis

Eddins

D. L.

Foege

W. H.

(1973). Assessment of vaccination coverage, vaccination scar rates, and smallpox scarring in five areas of West Africa. Bulletin of the World Health Organization, 48(2), 183. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2481004/

13.

Hindin

M. J.

(2008). Intimate partner violence among couples in 10 DHS countries: Predictors and health outcomes [AS18]. www.measuredhs.com

14.

Jewkes

(2002). Intimate partner violence: Causes and prevention. The Lancet, 359(9315), 1423–1429. https://doi.org/10.1016/S0140-6736(02)08357-5

15.

Jewkes

R. K.

Dunkle

Nduna

Shai

(2010). Intimate partner violence, relationship power inequity, and incidence of HIV infection in young women in South Africa: A cohort study. The Lancet, 376(9734), 41–48. https://doi.org/10.1016/S0140-6736(10)60548-X

16.

Jewkes

Levin

Penn-Kekana

(2002). Risk factors for domestic violence: Findings from a South African cross-sectional study. Social Science & Medicine, 55, 1603–1617. https://doi.org/10.1016/S0277-9536(01)00294-5

17.

Joyner

Mash

(2012). Recognizing intimate partner violence in primary care: Western Cape, South Africa. PLOS ONE, 7(1), Article e29540. https://doi.org/10.1371/journal.pone.0029540

18.

Joyner

Rees

Honikman

(2015). Intimate partner violence (IPV) in South Africa : How to break the vicious cycle. http://scholar.sun.ac.za/handle/10019.1/100659

19.

Kiarie

J. N.

Farquhar

Richardson

B. A.

Kabura

M. N.

John

F. N.

Nduati

R. W.

John-Stewart

G. C.

(2006). Domestic violence and prevention of mother-to-child transmission of HIV-1. AIDS, 20(13), 1763–1769. https://doi.org/10.1097/01.aids.0000242823.51754.0c

20.

Kouyoumdjian

F. G.

Calzavara

L. M.

Bondy

S. J.

O’Campo

Serwadda

Nalugoda

… Gray

(2013). Risk factors for intimate partner violence in women in the Rakai Community Cohort Study, Uganda, from 2000 to 2009. BMC Public Health, 13(1), Article 566. https://doi.org/10.1186/1471-2458-13-566

21.

Kouyoumdjian

F. G.

Findlay

Schwandt

Calzavara

L. M.

(2013). A systematic review of the relationships between intimate partner violence and HIV/AIDS. PLOS ONE, 8(11), Article e81044. https://doi.org/10.1371/journal.pone.0081044

22.

Kroenke

Strine

T. W.

Spitzer

R. L.

Williams

J. B. W.

Berry

J. T.

Mokdad

A. H.

(2009). The PHQ-8 as a measure of current depression in the general population. Journal of Affective Disorders, 114(1–3), 163–173. https://doi.org/10.1016/j.jad.2008.06.026

23.

Mikton

C. R.

Tanaka

Tomlinson

Streiner

D. L.

Tonmyr

Lee

B. X.

… MacMillan

H. L.

(2017). Global research priorities for interpersonal violence prevention: A modified Delphi study. Bulletin of the World Health Organization, 95, 36–48. https://doi.org/10.2471/BLT.16.172965

24.

Mitchell

Wight

Van Heerden

Rochat

T. J.

(2016). Intimate partner violence, HIV, and mental health: A triple epidemic of global proportions. International Review of Psychiatry, 28(5), 452–463. https://doi.org/10.1080/09540261.2016.1217829

25.

National Bureau of Statistics Nairobi, Kenya. (2015). Kenya Demographic and Health Survey 2014. www.DHSprogram.com

26.

Njue

Voeten

H. A. C. M.

Remes

(2011). Porn video shows, local brew, and transactional sex: HIV risk among youth in Kisumu, Kenya. BMC Public Health, 8(11), 635. https://doi.org/10.1186/1471-2458-11-635

27.

Olowookere

S. A.

Fawole

O. I.

Adekanle

D. A.

Adeleke

N. A.

Abioye-Kuteyi

E. A.

(2015). Patterns and correlates of intimate partner violence to women living with HIV/AIDS in Osogbo, Southwest Nigeria. Violence Against Women, 21(11), 1330–1340. https://doi.org/10.1177/1077801215594889

28.

Plichta

S. B.

(2004). Intimate partner violence and physical health consequences. Journal of Interpersonal Violence, 19(11), 1296–1323. https://doi.org/10.1177/0886260504269685

29.

Rabin

R. F.

Jennings

J. M.

Campbell

J. C.

Bair-Merritt

M. H.

(2009). Intimate partner violence screening tools. American Journal of Preventive Medicine, 36(5), 439–445.e4. https://doi.org/10.1016/j.amepre.2009.01.024

30.

R Core Team. (2018). R: A language and environment for statistical computing. R Foundation for Statistical Computing.

31.

Riggs

D. S.

Murphy

C. M.

O’Leary

K. D.

(1989). Intentional falsification in reports of interpartner aggression. Journal of Interpersonal Violence, 4(2), 220–232. https://doi.org/10.1177/088626089004002006

32.

Saunders

D. G.

(1991). Procedures for adjusting self-reports of violence for social desirability bias. Journal of Interpersonal Violence, 6(3), 336–344. https://doi.org/10.1177/088626091006003006

33.

Smith

M. V.

Gotman

Lin

Yonkers

K. A.

(2010). Do the PHQ-8 and the PHQ-2 accurately screen for depressive disorders in a sample of pregnant women?General Hospital Psychiatry, 32, 544–548. https://doi.org/10.1016/j.genhosppsych.2010.04.011

34.

Soeken

Parker

McFarlane

Lominak

Lominack

(1998, January1). The abuse assessment screen: A clinical instrument to measure frequency, severity and perpetrator of abuse against women. https://www.scienceopen.com/document?vid=b2f76211-69b8-40c4-ac28-e01228394214

35.

Spitzer

R. L.

Kroenke

Williams

J. B.

(1999). Validation and utility of a self-report version of PRIME-MD: The PHQ primary care study. Journal of the American Medical Association, 282, 1737–1744.

36.

Starnes

J. R.

Chamberlain

Sutermaster

Owuor

Okoth

Edman

Moon

T. D.

(2018). Under-five mortality in the Rongo Sub-County of Migori County, Kenya: Experience of the Lwala Community Alliance 2007-2017 with evidence from a cross-sectional survey. PLOS ONE, 13(9), Article e0203690. https://doi.org/10.1371/journal.pone.0203690

37.

Stöckl

Devries

Rotstein

Abrahams

Campbell

Watts

Moreno

C. G.

(2013). The global prevalence of intimate partner homicide: A systematic review. The Lancet, 382(9895), 859–865. https://doi.org/10.1016/S0140-6736(13)61030-2

38.

Stöckl

March

Pallitto

Garcia-Moreno

(2014). Intimate partner violence among adolescents and young women: Prevalence and associated factors in nine countries: A cross-sectional study. BMC Public Health, 14, Article 751. https://doi.org/10.1186/1471-2458-14-751

39.

Sutton

Dawson

(2018). Differentiating characteristics of intimate partner violence: Do relationship status, state, and duration matter?Journal of Interpersonal Violence. Advance online publication. https://doi.org/10.1177/0886260518795501

40.

Tanimu

T. S.

Yohanna

Omeiza

S. Y.

(2016). The pattern and correlates of intimate partner violence among women in Kano, Nigeria. African Journal of Primary Health Care & Family Medicine, 8, Article 1209. https://doi.org/10.4102/phcfm.v8i1.1209

41.

United Nations. (2015). Sustainable development goals. 17 goals to transform our world. World Health Organization.

42.

United Nations Committee on the Elimination of Discrimination Against Women. (2011). Concluding observations of the Committee on the Elimination of Discrimination against Women–Kenya. https://www.refworld.org/publisher,CEDAW,,KEN,4eeb60b12,0.html

43.

United Nations Development Programme. (2017). Gender inequality index. http://hdr.undp.org/en/composite/GII

44.

United Nations Office for the Coordination of Humanitarian Affairs. (2020). Kenya—Gender Inequality Index per county. https://data.humdata.org/dataset/kenya-gender-inequality-index-per-county

45.

Vest

J. R.

Catlin

T. K.

Chen

J. J.

Brownson

R. C.

(2002). Multistate analysis of factors associated with intimate partner violence. American Journal of Preventive Medicine, 22, 156–164. https://doi.org/10.1016/S0749-3797(01)00431-7

46.

Visschers

Jaspaert

Vervaeke

(2017). Social desirability in intimate partner violence and relationship satisfaction reports: An exploratory analysis. Journal of Interpersonal Violence, 32(9), 1401–1420. https://doi.org/10.1177/0886260515588922

47.

White

M. E.

Satyen

(2015). Cross-cultural differences in intimate partner violence and depression: A systematic review. Aggression and Violent Behavior, 24, 120–130. https://doi.org/10.1016/j.avb.2015.05.005

48.

Wong

F. Y.

Huang

Z. J.

DiGangi

J. A.

Thompson

E. E.

Smith

B. D.

(2008). Gender differences in intimate partner violence on substance abuse, sexual risks, and depression among a sample of South Africans in Cape Town, South Africa. AIDS Education and Prevention, 20(1), 56–64. https://doi.org/10.1521/aeap.2008.20.1.56

49.

Ziraba

Orindi

Muuo

Floyd

Birdthistle

I. J.

Mumah

Osindo

Njoroge

Kabiru

C.W.

(2018). Understanding HIV risk among adolescent girls and young women in informal settlements of Nairobi, Kenya: Lessons for DREAMS. PLoS One, 13(5), e0197479. https://doi.org/10.1371.journal.pone.0197479