Intimate partner violence against pregnant women in East Gojjam Zone, Ethiopia


 Abstract Background: Though underestimated, intimate partner violence (IPV) against women in Ethiopia is among the highest in the world. The overall aim of the study was to assess the prevalence and forms of intimate partner violence during pregnancy, associated factors, and recommendations to reduce the act. Methods: A cross-sectional study was conducted from January 25 to February 15, 2018 in government health facilities, North West Ethiopia. Married women who seek antenatal care service were participated in the study. Ethical clearance and approval was obtained from Amhara Regional Health Bureau research and publication committee. Informed consent was obtained from women and facility managers before collecting the data. Odds ratio (OR) with 95%CI was computed to determine the presence and strength of association between the outcome of interest and key independent variables. Results: Nearly half (46.4%) of the study participants were victims of at least one episodes of intimate partner violence in the recent pregnancy. Psychological violence 141 (44.2%) was the most common form of violence encountered followed by sexual violence 137 (42.9%). Significant proportion of women were insulted, forced to have sex, pushed/shoved, kicked/beaten, and slapped/thrown something at them (30.1 - 39.4 %) in the recent pregnancy. Lack of formal education, arranged marriages, lack of decision-making autonomy in household matters and wider spousal age difference were significantly associated with intimate partner violence. Conclusions: Higher proportion of pregnant women experienced intimate partner violence. Sociodemographic factors were associated with intimate partner violence. We suggest interventions focusing on education for raising community awareness through different strategies. Strengthening health facilities is also important to early identify and manage pregnant women with intimate partner violence, and prevent adverse birth outcomes of the act.

partner violence in the recent pregnancy. Psychological violence 141 (44.2%) was the most common form of violence encountered followed by sexual violence 137 (42.9%). Signi cant proportion of women were insulted, forced to have sex, pushed/shoved, kicked/beaten, and slapped/thrown something at them (30.1 -39.4 %) in the recent pregnancy. Lack of formal education, arranged marriages, lack of decisionmaking autonomy in household matters and wider spousal age difference were signi cantly associated with intimate partner violence.
Conclusions: Higher proportion of pregnant women experienced intimate partner violence.
Sociodemographic factors were associated with intimate partner violence. We suggest interventions focusing on education for raising community awareness through different strategies. Strengthening health facilities is also important to early identify and manage pregnant women with intimate partner violence, and prevent adverse birth outcomes of the act.

Background
Violence against women is a global public health concern and affects everyone in the world [1,2].
Globally, one-third of women experienced some form of physical, psychological/emotional and/or sexual violence [2]. The prevalence of the act is wider in Africa, Middle East and Latin America [3]. World Health Organization's (WHO's) multi-country study revealed that Japanese women experienced the lowest rates of domestic violence while the highest is in Ethiopia, Peru, and Bangladesh. [1] Studies showed that the prevalence of IPV was 30.3% in Tanzania, 53.2% in Nepal, and 67% in Kenya [4][5][6]. A systemic review of African studies showed that intimate partner violence during pregnancy ranged from 2.3% to 57.1 %; most of the studies showed a prevalence of 27% and above [7].
Violence against pregnant women results in fatal and non-fatal health outcomes for the mother and her baby [8,9]. The adverse health outcomes of IPV during pregnancy is higher in women with low socio economic status [9]. Adjusted for socioeconomic status of women, IPV was associated with depressive episodes in women [10]. The period of pregnancy by itself is stressful due to many uncertainties associated with it. Moreover, pregnant women are not immune for IPV; the violence may begin or worsen during pregnancy. The level of stress during pregnancy affects the severity of birth outcome [3]. Women experiencing IPV are predisposed to medical, gynaecological, and mental health problems [11]. IPV results in anxiety, depression, and stress which may lead to a higher risk of adverse obstetric outcomes; hypertension, anaemia, premature rupture of membrane, preterm delivery, low birth weight, vaginal bleeding, urinary tract infection, vomiting and dehydration, and low APGAR scores of 1 & 5 minutes. A mother with stress will have abnormally high level of cortisol and adrenaline in her blood. This high level of stress hormones will reach the baby and affect the baby's brain development. It also cause changes in the blood ow to the baby, making it di cult to carry oxygen and other important nutrients to the baby's developing organs, and make the mothers feel overwhelmed and fatigued which might impact their diet and sleep habits and consistency of prenatal care. [3,12,13] Despite the due attention given by the Ethiopian government for the protection of women's' right, 20% to 78% of women experienced lifetime domestic violence by their husband or intimate partner [14][15][16][17]. In Ethiopia, there is a deep-rooted traditional norm that supports men to discipline their wives in which wife beating is considered as normal [18]. Furthermore, IPV is often protected by family secrecy, cultural norms, fear, shame, community's reluctance on domestic affair and social stigma made the women remain silent, only few defend themselves [19]. As a result, evidence on the magnitude of IPV during pregnancy and its effect on adverse birth outcomes is scarce [20]. Therefore, the purpose of this study was to assess the magnitude of IPV against pregnant women and key factors associated with the act. All married /cohabited pregnant women who seek ANC service at health centres were included in the sample. We recruited women from six health centres who visited the health centres during the study period. The sample size has been estimated to be 319. It was determined using single population proportion formula n = [DEFF*Np(1-p)]/ [(d 2 /Z 2 1-α/2 *(N-1)+p*(1-p)] based on the following assumptions: 72.5% past 12 months prevalence of intimate violence in Ethiopia [17], 5% tolerable errors, 95% level of con dence and 10% non-response rate. All women who ful l the inclusion criteria and come to the health centre during the study period were included. Based on the client ow one month before the data collection period, proportional allocation of the total sample size was done to get the required sample from each health centre.
Trained health extension workers collected the data by face-to-face interview using structured and pretested questionnaire adopted from WHO multi-country study on women's health and domestic violence against women [21]. The interview was conducted privately in a separate room. Ethical clearance and approval was obtained from Research and Publications Committee of Amhara Regional Health Bureau. Support letter was written to the health centres. Women were informed about the objective, procedures, potential risks and bene ts of the study and written consent was obtained before the interview. STATA IC 12.1 software used for statistical analysis. Odds ratio with 95%CI computed to determine the presence and strength of associated factors with IPV

Results
Socio-demographic and behavioural characteristics of study participants Three hundred and nineteen pregnant women from six health centres participated in the study; yielding a response rate of 100%. Two hundred twenty three (69.9%) of women were below the age 30 years old andthe mean age of women was 28 years (28.   pregnancy. In addition, a quarter of the women were humiliated in front of others and threatened or frightened by their husband. Moreover, their husband intimidated one in three women on purpose.

Discussion
Intimate partner violence against pregnant women though underestimated; it has serious health consequences for the mother and the baby. The current study aimed to assess the prevalence and forms of IPV during pregnancy, factors associated with it and forward recommendations to reduce the act in the study area.
The overall prevalence of all forms of IPV against women (psychological, physical and sexual) during the recent pregnancy was unacceptably high. Nearly half of the study participants encountered at least one episodes of IPV in the recent pregnancy. The result of the current study (46.4%) was comparable with ndings of a study done in Debre Markos, Ethiopia (41.1%) [22] and cape town, South Africa (44%) [23],but lower compared to ndings from Bale zone, Ethiopia(59.0%) [24], Abay Chomen district, Ethiopia (55.5%) [20] and Kenya (66.9%) [4]. The difference could be explained by socio-cultural differences in the study populations. Furthermore, the prevalence of IPV in this study was lower than the lifetime prevalence of IPV in Ethiopia (60.6%), and low and lower-middle income countries (55.8%) [15,16].
In contrast, the nding of the current study was higher than a study done in South Africa (15%) [25]and Tanzania (30.3%) [5]. The likelihood of IPV against Ethiopia women was three fold of the rate in South Africa [25]. Similarly, Ethiopian women were victims of IPV at a rate of 16% higher than Tanzanian women were [5]. The possible explanations for the high prevalence of IPV in the current study could be explained by differences in socioeconomic status and the presence of deep rooted traditional norms that support wife beating as the rights of men to discipline their wives [26]. Being dominantly rural, such traditional norms are common and accepted in the community. Thought no statistical signi cant association observed, higher proportion of rural women encountered IPV.
Regarding factors associated with IPV, women's and husband's educational status, women's age (25 to 34 years), decision-making autonomy in household matters, marriages arranged by families and or relatives and spousal age difference were signi cantly associated with IPV. The prevalence of IPV was lowest among women of age 35 years and above, lack formal education or married to a man with no formal education, lack decision-making autonomy in household matters, had wider spousal age difference, and had no role to choose their partner. However, residence, occupation of the women, number of children and marital life were not associated with IPV perpetration. (Table 4) In this study, women 35 years of age and above were 70% (COR = 0.3, 95%CI; 0.1, 0.8) less likely to encounter IPV compared to young adults (15 to 24 years of age). This nding is comparable with a ndings from other studies; the probability of IPV perpetration is higher among adolescents and decreases starting in the late twenties [11,27].
The study corroborated that educational literacy has an inverse relation with IPV perpetration.
Educational status of the woman and her husband was signi cantly associated with IPV among pregnant women. Women who had no formal education experienced IPV 7 times (COR = 7.0, 95%CI; 3.1, 17.8) the rate of IPV among women who attended primary education or above. Similarly, women married to an illiterate husband had a 3 times increased risk of experiencing IPV compared to those married to men who attended primary education or above. This nding was supported by other studies done in Ethiopia [15,28] and Kenya [4]. The possible explanation can be, when individuals and community got educated, they will have knowledge on human rights and skills to negotiate and solve disagreements through discussion. Interventions to improve awareness of couples through different approaches will help to prevent and reduce the occurrence of violence against women.
This study also found that arranged marriage was signi cantly associated with the likelihood of intimate partner violence during pregnancy. Marriage unions arranged by families and or relatives had 300% (COR = 2.8, 95%CI; 1.2, 7.6) increased likelihood of IPV perpetration as compared to those unions happened by choice. The possible explanation could be romantic relationships may not be common among couples whose marriages are arranged by families and or relatives. Especially it will be true in the early life of the marriage until the couples develop truest and commitment. Arranged marriage is a common social phenomenon in the study area. Typically, the father/ family of the prospective groom will send elders to the father/family of the prospective bride to propose a marriage. If the union is acceptable, the families will negotiate the marital exchange and set the wedding date [29]. Besides, most women got married at an early age, below 15 years old without their choice. When women got married at early, they age are disadvantaged within the marriage. They enter marriage without adequate knowledge and skills to negotiate marital roles [29].
Our study showed that wider spousal age difference was associated with IPV perpetration. Women who had age difference of 10 years or more with their spouses experienced 3 times (COR = 2.9, 95%CI; 1.4, 5.9) higher rate of IPV compared to those who had less than 5 years age difference. The nding emphasizes the need for couples to have same level of maturity and understanding. However, a study done in Nigeria and a meta-analysis on the association between IPV and spousal age showed that IPV perpetration reduced with increased spousal age difference [30,31]. The disagreement in the ndings needs further study.
Findings of this study also showed that decision-making autonomy is an important factor associated with lower likelihood of IPV. The likelihood of IPV perpetration was higher in families where husband is making decision alone compared to woman alone. Husband making decision alone in household matters had a seven times (COR = 6.7, 95%CI, 2.3, 23.3) increased likelihood of IPV perpetration compared to woman decision-making alone. Other studies in Ethiopia substantiated the nding [15]. Similarly, a study in Uganda showed that women decision making in large household purchases and expenditure of husband's earnings was associated with lower probability of IPV. However, unlike the current study women's reporting decision-making autonomy was not signi cantly associated with IPV. [32]. Table 4: Intimate partner violence during pregnancy by sociodemographic and behavioural characteristics, Ethiopia, 2018 IPV exposed IPV non-exposed COR (95%CI)

Limitations Of The Study
In this study, the prevalence of IPV was determined based on the reports of women and social desirability bias might affect the result. In addition, we could not able to compute multivariate analysis due the insu ciency of samples in each category of a variable. Therefore, the observed associations might have been confounded.

Conclusions
The prevalence of intimate partner violence during pregnancy in Machakel woreda, is relatively high. Nearly half of the pregnant women experienced at least one form of intimate partner violence by their husband. The problem seems to be more common among women age 25 to 29 years old, couples having wider spousal age difference, women with no formal education or married to a man with no formal education, women with low or no decision-making power in house hold matters and marriages arranged by families and or relatives. Findings suggest that interventions focused on education and awareness creation activities through various approaches may be protective against IPV. Health facilities should be strengthened to do health education and community awareness activities, identify and counsel and or link pregnant women with intimate partner violence. Ethics approval and consent to participate

Abbreviations
The study got ethical approval from Research and Publications Committee of Amhara Regional Health Bureau and support letter was written to the study sites. Data was collected after informed written consent was obtained from study participants. For women of age less than 18 years old informed written consent was obtained from accompanies (husband, parent). The purpose of the study and the time needed to conduct the interview was explained to the study participants. Con dentiality of the data was ensured and there was no personal identi er attached to the questionnaire.

Consent for publication
The author representing all the authors grants the publisher a full copyright license to publish in whole or part without limitation in all forms of media.

Availability of data and material
The dataset used and/or analysed for the current study is available from the corresponding author on request.

Competing interest
No competing or con ict of interest to declare.

Funding
No funding was obtained for this study