Maternal-Fetal Attachment and Domestic Violence in Iranian and Afghan Pregnant Women: A Cross-Cultural Study

Background: Domestic violence against pregnant women is becoming more frequent and intense. Studies have shown that women who were under more physical and psychological violence experienced a lot of stress and depression and had less ability to communicate with their fetuses. International statistics show a high prevalence of violence against Afghan women and the communities migrated to Afghanistan. Therefore, the present study aimed to compare the maternal-fetal attachment and domestic violence between Iranian and Afghan pregnant women. Methods: This study has a descriptive-comparative cross-sectional design. The study population consisted of Iranian and Afghan pregnant women referred to Kerman Health Centers in 2018-2019. One hundred-forty-six Iranian pregnant women and 142 Afghan pregnant women were enrolled in the study by the quota sampling. Data collection tools were demographic and clinical information questionnaire, questionnaires of domestic violence against women, and maternal-fetal attachment scale. Results: The mean score of maternal-fetal attachment in Iranian and Afghan pregnant women was higher than the average. Also, scores of maternal-fetal attachment and its dimensions in Afghan pregnant women were signicantly higher than that of Iranian pregnant women (P < 0.001). The mean score of domestic violence was low in Iranian and Afghan pregnant women. There was no signicant difference between Iranian and Afghan pregnant women in the total score of domestic violence (P > 0.05). There was a signicant inverse correlation between maternal-fetal attachment and domestic violence in Iranian pregnant women (P < 0.05). In Afghan pregnant women, there was a positive and signicant correlation between maternal-fetal attachment and sexual violence in while there was a signicant and reverse relationship between other dimensions of attachment and violence (P < 0.05). Conclusion: This study showed that the rate of violence in the studied populations was low, and maternal-fetal attachment rate is appropriate. The relationship between violence and mother-fetal attachment had a different pattern in Iranian and Afghan women. Therefore, further studies are needed to investigate the factors affecting maternal-fetal attachment, especially in Afghan women. between Iranian and Afghan pregnant women in the total score of domestic violence. In conclusion; the results showed that the rate of violence in the studied populations was low, and maternal-fetal attachment rate is appropriate. The relationship between violence and mother-fetal attachment had a different pattern in Iranian and Afghan women. Therefore, further studies are needed to investigate the factors affecting maternal-fetal attachment, especially in Afghan women.


Introduction
The maternal-fetal bonding begins before birth and after the rst fetal movement and encourages the mother to adjust her behavior and moods to promote the safe development of the fetus. The fetus also feels and is affected by the mother's behaviors and actions [1]. Maternal-fetal attachment involves behaviors and actions that indicate maternal-fetal emotional bonding. These behaviors lead to continued parental care, proper nutrition, sleep and exercise, abstinence from alcohol and drugs, and a desire to recognize the fetus during pregnancy, which ultimately lead to the desired pregnancy outcome and improve maternal-fetal health [2]. Maternal-fetal attachment behaviors play an important role in health, acceptance of parental identity, future mother-infant relationship, child's development, and emotional stability [3,4]. A mental image of the fetus helps the mother conceive of the fetus as a living being within herself. According to studies, 30% of mothers in the rst trimester, 63% in the second trimester, and 92% in the third trimester conceive the fetus as a real human [1,5,6]. Pregnant women who have a strong attachment to their fetus believe that the fetus is in contact with them and is a separate individual. These mothers think about the way the fetus moves, its similarity, and growth in later years. These emotions, which have a positive effect on attachment, help them develop a sense of protection, sensitivity, and communication with the fetus [7]. In general, the relationship between a pregnant woman and her fetus has immediate and long-term effects on maternal healthcare measures during pregnancy, neonatal and health outcomes, maternal-infant interaction, and infant attachment security [8]. Various studies have shown that age, education of wife and husband, unwanted pregnancy, economic conditions, employment status of the mother and her husband, antenatal education are factors that can have either a positive or negative impact on maternal-fetal attachment [3,[9][10][11]. Also, desired marital relationships play an important role in increasing maternal-fetal attachment. Because mothers tend to see their fetus under the patronage of father and somehow guarantee its future. Husband's violence is one of the factors that lead to lower marital quality and maternal-fetal attachment [11,12].
Another factor affecting maternal-fetal attachment is domestic violence. Violence is a pattern of behavior that is imposed through fear, intimidation, abusive behavior to apply power and control over a close individual, and includes physical, sexual, economic abuses, verbal threats, and divorce [13]. About 40-50% of women experience domestic violence in some parts of their lives. As a result, their physical and mental health is affected and sometimes leads to suicide [14]. Violence against pregnant women is a very sad event and a common phenomenon in most societies. The prevalence of domestic violence is higher than the prevalence of any pregnancy-related medical disorders such as preeclampsia, gestational diabetes, etc. [13]. Pregnancy imposes a lot of physical and psychological stress on the individual and along with other stressors such as violence can have adverse effects on the fetus, and the mother and these effects can increase maternal and infant mortality [14]. Studies in different countries show that domestic violence against pregnant women varies between 25-86% [13,15,16]. Other studies in different countries have shown that one in three pregnant women is subjected to physical, emotional, or sexual violence by their partner [17][18][19]. Domestic violence has seriously negative effects on pregnant women. Domestic violence has been reported to be associated with abortion, preterm labor, low birth weight, antenatal care, early placental separation, membrane rupture, and fetal injury [15,16,20]. Various studies considered husbands' violence against their pregnant wife as the main cause of decreased marital quality and maternal-fetal attachment. Also, women who are subjected to physical violence cannot pay enough attention to the dimensions of the maternal-fetal attachment process because they are struggling to cope with the hostile environment in the home. As a result, the attachment rate declines in these women [21][22][23]. Domestic violence is not only a serious public health issue during pregnancy but also endangers the maternal-fetal health. On the other hand, regarding the impact of violence on the level of maternal-fetal attachment and the importance of maternal-fetal attachment for the health of mother and fetus in future, this study aimed to compare maternal-fetal attachment and domestic violence in Iranian and Afghan pregnant women.

Study design and settings
This study has a descriptive-comparative cross-sectional design. The study setting was health centers providing perinatal and maternal care services in Kerman city in southeastern Iran. Kerman is one of the most populous cities in terms of the number of Afghan citizens in Iran.

Sampling Method And Sample Size
Quota sampling was used in this study. The city was divided into four districts, and two health centers were randomly selected from each district. Then, convenience sampling was done between Iranian and Afghan clients to reach the appropriate sample size. According to quota sampling and sample size, equal samples were selected from each district. Inclusion criteria were Iranian and Afghan pregnant women over 15 years old. Since no similar study was found in this eld, a pilot study was conducted to estimate the sample size. Fifteen Iranian pregnant women and 15 Afghan pregnant women completed the questionnaires, and then the sample size was estimated using the following formula. The 95% con dence interval and the test power of 1-β = 0.9 were considered. In the pilot study, the mean, standard deviation of domestic violence in Iranian pregnant women were 8.93 ± 9.89, and it was 16.53 ± 21.57 in Afghan pregnant women. According to the above numbers, 127 subjects were calculated in each group, and 150 questionnaires were distributed in each group due to dropout probability. Finally, after collecting 150 questionnaires from each group and eliminating the questionnaires with more than 10% missing data, 146 questionnaires in the Iranian pregnant women group and 142 questionnaires in the Afghan pregnant women were analyzed. Therefore, the response rate of Iranian pregnant women was 97%, and that of Afghan pregnant women was 94.66%.

Data Collection Tools
The tools used in this study were maternal demographic and background information questionnaire, domestic violence against women questionnaire, and maternal-fetal attachment scale. The maternal demographic and background information questionnaire included 16 items about age, education, occupations of mothers and their husbands, and some other information.

Domestic Violence Questionnaire
A researcher-made questionnaire was used to assess domestic violence. An initial draft of the questionnaire was developed using the literature review and expert opinion [13,14,22,23]. Then, eight faculty members of Kerman School of Nursing and Midwifery assessed the questionnaire content validity. Their corrective comments were applied to the questionnaire, and the content validity index was 0.98. The questionnaire was provided to 30 subjects (15 Iranian pregnant women and 15 Afghan pregnant women) for measuring reliability. Cronbach's alpha coe cient was 0.89 for Iranian pregnant women and 0.88 for Afghan pregnant women.
The questionnaire consisted of 27 items in 5 dimensions. Items 1-6 related to verbal violence, items 7-12 related to physical violence, items 13-17 related to sexual violence, items 18-24 related to emotionalpsychological violence, and items 25-27 related to nancial violence. The six-point Likert scale was used for scoring (0 = not at all, 5 = a lot). Questionnaire minimum and maximum scores were 0 and 135, respectively. A higher score showed higher domestic violence.

Maternal-fetal Attachment Scale (mfas)
Cranley rst developed this self-report scale in 1981 and con rmed its reliability by alpha Cronbach's coe cient (α = 85%) [24]. Khoramroodi in Iran translated Maternal-fetal attachment scale, and its validity was con rmed with content validity, and its reliability was con rmed with test-retest (r = 0.85) [25]. In the present study, Cronbach's alpha coe cient was 0.78 in Iranian pregnant women and 0.81 in Afghan women.

Data Collection
The researcher referred to the selected health centers after obtaining permission from the ethics committee and the head of the School of Nursing and Midwifery. Then, she introduced herself to the clinic staff, presented a letter of introduction and explained the research, and obtained their consent and went to the primary research setting, the perinatal and maternal care unit. Before completing the questionnaire, participants were provided with necessary explanations about the objectives of the study and information con dentiality. Then, after providing a private and con dential environment, the domestic violence questionnaire and maternal-fetal attachment scale were completed by Iranian and Afghan women. Data collection lasted from 23.Oct.2018 to 19.Feb.2019. Whenever the samples needed clari cation of the questions, the researcher provided su cient explanations. Two Afghan women uent in Persian were used for communicating with Afghan women due to di culty in verbal communication.
Statistical Analysis SPSS version 18 was used. The demographic and background characteristics of the samples were presented using descriptive statistics. The results of the Kolmogorov-Smirnov test showed that only the total maternal-fetal attachment score in Iranian pregnant women followed a normal distribution and other variables did not have normal distribution. Therefore, Mann-Whitney U test was used to compare maternal-fetal attachment and domestic violence between Iranian and Afghan pregnant women. Spearman correlation coe cient was also used to investigate the correlation between attachment and domestic violence against pregnant women. Independent t-and ANOVA tests and Mann-Whitney, Kruskal-Wallis and Spearman correlation coe cient tests were used to investigate maternal-fetal attachment and domestic violence variables in both groups in terms of background-demographic variables. Signi cance level was considered to be 0.05.

Results
The mean ages of Iranian and Afghan pregnant women were 28.3 ± 5.83 and 21.84 ± 5.92, respectively. 66.2% of Iranian pregnant women and 97.9% of Afghan women were housewives. The majority of Iranian pregnant women were academically educated, while the majority of Afghan pregnant women were illiterate. The majority of husbands of Iranian pregnant women were academically educated, while most husbands of Afghan pregnant women had middle/high school degrees ( Table 1). The majority of pregnant women in both groups had done ultrasonography for their current pregnancy. The gender of the fetus was female in the majority of Iranian samples, and it was male in the majority of the afghan samples. The majority of samples had healthy embryos. The majority of samples in both groups had no history of abortion and infertility. The majority of Iranian samples had wanted pregnancy, but the majority of Afghan samples had unwanted pregnancy. The majority of the samples and their husbands in both groups had no history of physical and mental illness (Table 1). The mean scores of maternal-fetal attachment in Iranian and Afghan pregnant women were 92.87 ± 12.55 and 103.6 ± 19.65, respectively. According to the median score of the questionnaire, which is 72, the maternal-fetal attachment rate was higher than the average in both groups. Also, the total scores of mother-fetal attachment and all its dimensions in Afghan pregnant women were higher than that in Iranian pregnant women (Table 2). Among the underlying variables of Iranian pregnant women, the total maternal-fetal attachment score was signi cantly different only in husband nationality. In other words, pregnant women with non-Iranian husbands had higher maternal-fetal attachment score than those with Iranian husbands (P < 0.05). The older the age of Afghan pregnant women, and their husbands, as well as the higher their marital duration, the lower the maternal-fetal attachment score ,and the severity of these relationships, was moderate (r = 0.2-0.4) (P < 0.05). Also, the maternal-fetal attachment score was higher in Afghan women with unwanted pregnancy than those with wanted pregnancy (P < 0.05).
The mean scores of domestic violence in Iranian pregnant women and Afghan pregnant women were 9.63 ± 8.13 and 10.51 ± 9.1, respectively. As the median score of the questionnaire was 67.5, the level of domestic violence was low in both groups. Also, there was no signi cant difference between Iranian and Afghan pregnant women in the total score of domestic violence. However, among the dimensions of domestic violence, the mean dimension of verbal and emotional-psychological violence in Iranian pregnant women was signi cantly higher than that of Afghan pregnant women, while the mean dimension of sexual violence in Afghan pregnant women was signi cantly higher than that of Iranian pregnant women. There were also no differences between the two groups in physical and nancial violence (Table 3). There was no signi cant association between background variables and total score of domestic violence in Iranian pregnant women (P > 0.05). The higher the age of afghan pregnant women, their husbands, and marital duration, the higher the score of domestic violence and the severity of these relationships was moderate (r = 0.2-0.4) (P < 0.05). Pregnant women with a personal house also reported more violence than other Afghan women. Women whose fetal gender was unknown reported more violence. Women who had unwanted pregnancies reported higher rates of violence than women who had wanted pregnancy. Women with more than two children reported more violence than other Afghan pregnant women. Afghan pregnant women with physical illness reported more violence than healthy women (P < 0.05).
There was a signi cant inverse relationship between the mean maternal-fetal attachment and the mean domestic violence in Iranian pregnant women. In other words, the higher the score of domestic violence, the lower the maternal-fetal attachment score, and vice versa. Also, the maternal-fetal attachment score had a poor and inverse relationship only with the dimensions of verbal, emotional-psychological, and nancial violence (Table 4). There was no signi cant association between the mean total score of maternal-fetal attachment and the mean total score of domestic violence in Afghan pregnant women. However, there was a signi cant inverse association amongst verbal, physical, emotional-psychological and nancial violence and score of maternal-fetal attachment and all its dimensions. In other words, the higher the scores of these dimensions, the lower the scores of maternal-fetal attachment and its dimensions, and vice versa. Also, there was a direct and moderate association between the dimension of sexual violence and the scores of maternal-fetal attachment and all its dimensions. In other words, the higher the sexual violence, the higher the maternal-fetal attachment rate in Afghan pregnant women (Table 5).

Discussion
The results of this study showed that maternal-fetal attachment in pregnant Iranian and Afghan pregnant women was more than average. Also, maternal-fetal attachment and all its dimensions were signi cantly higher in Afghan pregnant women than in Iranian pregnant women. Results of several studies on Iranian pregnant women showed that maternal-fetal attachment in pregnant women was more than average [1][2][3]26]. Also, the results of various studies on pregnant women in Korea, Africa, Turkey, and the USA showed that the maternal-fetal attachment was also above average, all of which were in line with the present study [4,7,8,27]. But O'Malley et al. (2018) in Ireland showed that the rate of maternal-fetal attachment in Irish pregnant women was moderate [28]. This inconsistency may be due to the low sample size, differences in research methodology, and cultural differences. Based on the above studies, it can be concluded that maternal-fetal attachment rate among pregnant women (either in Iran or in different countries) is favorable. Maternal attachment to the fetus during pregnancy affects the formation and development of a safe and healthy child attachment [9]. With increased maternal attachment to the fetus, mothers are more likely to engage in health behaviors during pregnancy such as stopping smoking and alcohol, proper nutrition, exercise, continued antenatal care, tendency to recognize the fetus, and attending birth preparation classes, as well as increasing breastfeeding rates. All of these behaviors lead to a satisfactory pregnancy outcome and improve maternal-fetal health [10]. Therefore, it is necessary to consider various factors that may affect maternal-fetal attachment when caring for pregnant women. For example, in the present study, pregnant women with non-Iranian spouses had higher maternal-fetal attachment scores than those with Iranian spouses, and Afghan women's age, spouse age, marital length, and unwanted pregnancy in uenced maternal-fetal attachment rate. In other words, the higher the age of pregnant women, their spouses and duration of marriage, the lower the maternal-fetal attachment score and the severity of these relationships was moderate. Also, the maternalfetal attachment score of Afghan women with unwanted pregnancy was higher than that of Afghan pregnant women with wanted pregnancy. In the studies, older age [1,3,28], low gestational age [3,10], low education [3,10,28], low education-related jobs such as agriculture [1,3,10], dissatisfaction with fetus gender [3], complications of current pregnancy [3], and unwanted pregnancy [10] reduce maternalfetal attachment and antenatal education [4], number of pregnancies [1,3,10], and breastfeeding experience [4] increases the maternal-fetal attachment. These studies were inconsistent with the present study. But among the above variables, age was in line with Afghan women. No study was found to investigate maternal-fetal attachment in Afghan women.
The results of this study showed that the level of domestic violence was low in Iranian and Afghan pregnant women. Besides, verbal and emotional-psychological abuses in Iranian pregnant women were signi cantly higher than that of Afghan pregnant women, while sexual abuse in Afghan pregnant women was signi cantly higher than that in Iranian pregnant women. There was also no difference between the two groups in physical abuse, nancial abuse, and total score of domestic violence. The Cengiz et al. found that the rate of domestic violence against pregnant women was high in these regions [17,30].
These studies were inconsistent with the present study. Differences in sample size, cultural differences in different countries and communities, differences in education level, inaccessibility to accurate information about marital relationship, and differences in data collection questionnaires can lead to inconsistencies amongst studies. Studies conducted on Iranian pregnant women showed that domestic violence was high in these women [13,14,31]. These studies were also inconsistent with the present study. Reasons for such inconsistency can be differences in sample size, research method, differences in questionnaires used to measure the domestic violence rate, and cultural differences in respondents' willingness to disclose their marital experiences. In this study, despite the trust-building, it should be noted that since domestic violence is related to the privacy of one's life, the women under investigation might either have denied the violence or have not stated it correctly for some reasons. The results of the studies showed that the most common type of violence was sexual one and then psychological and emotional violence [14,17,20,29] which were in line with the present study. But Yohannes et al.'s (2019) study in southeastern Ethiopia was not consistent with the present study which showed that physical violence had the highest rate of violence among pregnant women [30]. Also, the results of Hassan et al. (2013) study showed that physical violence and psychological violence were the most common types of violence in this study. This study was inconsistent with the present study regarding physical violence [13]. Kamali et al. (2015) study found that the most common type of violence was related to the physical, psychological, and sexual violence, respectively. This study investigated violence in urban and rural pregnant women. Violence was high in both communities, but physical, psychological, and sexual violence on rural pregnant women was higher than that on urban pregnant women [30]. Domestic violence during pregnancy increases the risk of miscarriage, vaginal bleeding, cesarean section, low birth weight and delayed growth in the uterus [13][14][15]. Therefore, it is necessary to take into account various factors that may affect domestic violence when taking care of pregnant women. For example, in the present study, variables such as age, spouse's age, marital duration, housing status, fetus gender, wanted pregnancy, number of children, and history of physical illness in the afghan women in uenced the rate of domestic violence. In other words, the higher the age of pregnant women, their spouses and the duration of marriage, the higher the score of domestic violence. Pregnant women with a personal house, women whose fetus gender was unknown, unwanted pregnancy, women with more than two children, and women with physical illness reported more violence than other Afghan women. In other studies, age [14,31], spouse's age [31], education [13,16,31], economic status [13,31], spouse's addiction (smoking and alcohol) [13,14,30,31], number of pregnancies and children [13,31], duration of marriage [14,31], experience of previous marriage [14], unwanted pregnancy [16] and spouse's occupation [13,30] in uenced the rate of violence against pregnant women. These studies were inconsistent with the  [21,23]. This study was not consistent with Iranian pregnant women, but it was consistent with Afghan women. Differences in research method, cultural differences, socioeconomic characteristics and method of data collection, different individual and social characteristics in two populations in two different cities, low number of samples under study and impact of mental states of samples on responses can cause such differences in the results of studies. Ghodrati et al. (2017) in their study concluded that there was a signi cant relationship between maternal-fetal attachment and social violence, but there was no relationship between maternal-fetal attachment and psychological, physical, nancial, and sexual violence [22]. This study was inconsistent with the present study. The reasons for inconsistency include differences in sample size, cultural differences between individuals in two different cities, differences in data collection questionnaires, and selection of 24-to 35-week gestational age for samples. From the above studies, it can be concluded that violence against pregnant women reduces maternal-fetal attachment. Women who are victims of domestic violence have a negative attitude toward pregnancy and their fetus. It is recommended to reduce and control violence, conduct family and marital counseling to increase maternal attachment to the fetus and improve motherhood process in these women [21,23]. The results of this study showed that the higher the sexual violence, the higher the maternal attachment to fetus in afghan pregnant women. To justify this result, one might point to a systematic study conducted by Vogel in 2013, in which 75% of women believed that sex, whether with consent or not, was a woman's duty [33]. However, further research is needed on the Afghan pregnant women population and the cultures dominating this community to justify the results of this study properly.

Limitations
Among the limitations of this study were "over-reporting" and "under-reporting" by volunteers. "Underreporting" is more common than "over-reporting" in socially unacceptable phenomena such as violence.
Another limitation of the study was the inclusion of all pregnant mothers regardless of gestational age, which may affect the results of the study.

Conclusion
In general, this study showed that the level of violence in the studied populations was low, and the maternal-fetal attachment rate was appropriate. However, violence had a negative role in maternal-fetal interaction in Iranian pregnant women, while there was a more complicated relationship between different dimensions of violence and maternal-fetal attachment in Afghan women, and there was a positive relationship in some dimensions while there was a negative relationship in some other dimensions. Therefore, further studies are needed to investigate the factors affecting maternal-fetal attachment, especially in Afghan women.

Maternal-fetal attachment scale
Declarations