Postpartum Depressive Symptoms as a Mediator Between Intimate Partner Violence During Pregnancy and Maternal-Infant Bonding in Japan

Abstract

Studies show that experiencing intimate partner violence (IPV) during pregnancy is related to poor maternal–infant bonding. However, the mechanisms underlying this relationship are unclear. This article aims to examine whether maternal postpartum depressive (PPD) symptoms mediate the association between pregnancy IPV and maternal–infant bonding, and whether the relationship differs by maternal–infant bonding subscales—lack of affection, anger/rejection. A survey was conducted among women who participated in a postpartum health check-up program in Aichi prefecture, Japan (N = 6,590) in 2012. We examined whether experiences of emotional and physical IPV were related to maternal–infant bonding and whether PPD symptoms mediated this relationship. Path analysis showed that emotional and physical IPV were associated with PPD symptoms, and PPD symptoms predicted poor bonding. The total effect of emotional IPV on poor bonding was significant, showing a marginally significant direct effect and statistically significant indirect effect. The total effect of physical IPV on poor bonding was not statistically significant. Emotional IPV was significantly associated with both lack of affection and anger/rejection bonding subscales, which were similarly mediated by PPD symptoms. Findings revealed a modest indirect association between IPV, emotional IPV in particular, and poor maternal–infant bonding, which was mediated by PPD symptoms. While prevention of IPV is the ultimate goal, the treatment of PPD symptoms among women who experience IPV during pregnancy may improve maternal–infant bonding and mitigate cross-generational effects of IPV. Identifying opportunities for detection of IPV and PPD symptoms, as well as prevention and early intervention, may improve maternal–infant bonding.

Keywords

intimate partner violence (IPV)pregnancy IPV emotional IPV maternal–infant bonding postpartum depressive (PPD) symptoms path analysis Japan

Introduction

Intimate partner violence (IPV), defined as physical violence, sexual violence, stalking or psychological aggression perpetrated by a current or former intimate partner (Breiding, Basile, Smith, Black, & Mahendra, 2015), is known to have detrimental impacts on victim’s physical, mental, and social health (J. C. Campbell, 2002; Dillon, Hussain, Loxton, & Rahman, 2013; Heise & Garcia-Moreno, 2002; Mechanic, Weaver, & Resick, 2008; Plichta, 2004). Both men and women can be victims of IPV, but for the purpose of the article, we define IPV as violence perpetrated against women by their male partners. Women can be the targets of violence at any time of their life but IPV during pregnancy is particularly common (Devries et al., 2010; Gazmararian et al., 2000). Numerous studies revealed that experiencing IPV during pregnancy negatively affects both mothers and infants during the postpartum period, causing preterm birth, low birth weight, and poor postpartum mental health (Alhusen, Ray, Sharps, & Bullock, 2015; J. C. Campbell, 2002; Huth-Bocks, Levendosky, & Bogat, 2002; Silverman, Decker, Reed, & Raj, 2006). In particular, some researchers have found that exposure to IPV during pregnancy is associated with poor maternal–infant bonding in the postpartum period (Kita, Haruna, Matsuzaki, & Kamibeppu, 2016; Levendosky, Lannert, & Yalch, 2012; Quinlivan & Evans, 2005; Zeitlin, Dhanjal, & Colmsee, 1999). Maternal–infant bonding is an emotional experience reflecting a tie between a mother and her child (Bicking Kinsey & Hupcey, 2013; Klaus & Kennell, 1976), and it is an essential factor for infant physical (Mäntymaa et al., 2003) and mental development (Greenberg, 1999; Kochanska, Aksan, Knaack, & Rhines, 2004; Sullivan, Perry, Sloan, Kleinhaus, & Burtchen, 2011). Considering the role of maternal–infant bonding in child development, identifying mechanisms through which IPV influences maternal–infant bonding can be important for identifying intervention strategies for mitigating cross-generational effects.

Postpartum Depressive (PPD) Symptoms: A Plausible Mediator

Impacts of IPV during pregnancy on maternal–infant bonding may be direct or indirect, and operate through different pathways. Some researchers have proposed that antenatal depression (Kita et al., 2016), maternal–fetal bonding (Zeitlin et al., 1999), reduced oxytocin levels (Samuel et al., 2015), or impairment in parenting capacity (Bogat, Levendosky, von Eye, & Davidson, 2011) could indirectly influence the impact of IPV during pregnancy on maternal–infant bonding during the postpartum period. Existing research on associations between IPV during pregnancy and maternal–infant bonding has been limited by relatively small sample sizes, failure to distinguish between different types of IPV (e.g., physical, emotional), and lack of clarity regarding mechanisms to explain this relationship (Kita, Yaeko, & Porter, 2014; Zeitlin et al., 1999). Based on previous research, PPD symptoms could be one factor that results from IPV victimization during pregnancy and then leads to poor maternal–infant bonding. Studies have consistently found a relationship between pregnancy IPV and PPD symptoms (Ahmad et al., 2018; Beydoun, Al-Sahab, Beydoun, & Tamim, 2010; Edhborg, Matthiesen, Lundh, & Widström, 2005; Islam, Broidy, Baird, & Mazerolle, 2017; Lau & Chan, 2007; Ludermir, Lewis, Valongueiro, de Araújo, & Araya, 2010; Miura & Fujiwara, 2017) as well as between PPD symptoms and poor maternal–infant bonding (Kumar, 1997; Martins & Gaffan, 2000; Moehler, Brunner, Wiebel, Reck, & Resch, 2006; O’Higgins, Roberts, Glover, & Taylor, 2013; Seng et al., 2013), suggesting that PPD symptoms may mediate the relationship between pregnancy IPV and bonding during the postpartum period. Japanese women who have experienced physical abuse during pregnancy had about 7 times the odds of PPD symptoms compared with those who did not experience physical abuse. Within the same population, women who were often verbally abused had about a fivefold increased odds of PPD symptoms (Miura & Fujiwara, 2017). Moreover, depressed mothers find it difficult to respond to infants appropriately (S. B. Campbell, Cohn, & Meyers, 1995; Murray, 1992; Weissman, Paykel, & Klerman, 1972), which potentially leads to poor maternal–infant bonding. Overall, there is evidence of an association between IPV during pregnancy and PPD as well as PPD and poor maternal–infant bonding. Nonetheless, research linking the two associations through mediation analyses is lacking, resulting in limited understanding of the direct and indirect effects of a mother’s exposure to IPV during pregnancy on bonding with her infant during the postpartum period.

IPV During Pregnancy and the Japanese Context

IPV consists of four types of behavior, including physical, emotional, sexual violence, and stalking (Breiding et al., 2015). Physical IPV is defined as the intentional use of physical force potentially resulting in death, disability, injury, or harm by hitting, kicking, pushing, or use of a weapon. Emotional IPV is the use of verbal or nonverbal communication with the intent to cause emotional harm and/or exert control over another person. It occurs when the partner’s words or actions cause someone to feel stupid or worthless (Breiding et al., 2015; Kumagai & Ishii-Kuntz, 2016). Sexual IPV is a sexual act that is attempted or committed by a partner without having received consent. Stalking is a pattern of repeated, unwanted attention and contact resulting in concern or fear (Breiding et al., 2015). The prevalence of IPV during pregnancy varies widely by setting and measurement tool. Large multicountry studies have estimated the ever-prevalence or physical IPV among women to range from 4% to 14% (Devries et al., 2010). Although there is inconsistent evidence as to whether pregnancy protects women against IPV (J. C. Campbell, García-Moreno, & Sharps, 2004; Koenig et al., 2006; McMahon & Armstrong, 2012; World Health Organization, 2005), some have reported that pregnancy decreases the severity and frequency of physical IPV (J. C. Campbell et al., 2004; McMahon & Armstrong, 2012), whereas emotional IPV appears to increase with the onset of pregnancy (Martin, Harris-Britt, Morraco, Kupper, & Campbell, 2004).

Although the reported prevalence of physical IPV during pregnancy in Japan is among the lowest in the world (at about 1%; Kumagai & Ishii-Kuntz, 2016; World Health Organization, 2005), the prevalence of IPV sharply increases when expanding the definition to include physical, emotional, and/or sexual harm. The prevalence of this expanded definition of IPV during pregnancy was estimated to be between 16% and 31% (Inami, Kataoka, Eto, & Horiuchi, 2010; Kataoka, Imazeki, & Shinohara, 2016; Kita et al., 2014), much less than the prevalence of IPV prior to pregnancy (34.9%; Kataoka et al., 2016). The heterogeneity in these estimates appeared to depend on participant demographics, measurement tools, and questions (Inami et al., 2010; Kataoka et al., 2016; Kita et al., 2014). Correlates of IPV include maternal age, paternal age younger than 30 years, multiparity, or previous experience of physical violence (Inami et al., 2010; Kita et al., 2014). Although emotional IPV has been less studied than physical IPV (Bonomi, Anderson, Rivara, & Thompson, 2007; Koenig et al., 2006; World Health Organization, 2005), it can have a substantial impact on maternal mood (e.g., depressive symptoms; Yoshihama, Horrocks, & Kamano, 2009), and researchers consider emotional IPV to be as damaging as physical IPV (Pico-Alfonso et al., 2006). The impact of different forms of IPV needs to be examined separately to identify whether the various types of IPV have distinct mechanisms that could highlight specific intervention strategies and recommendations.

The relatively low reported prevalence of IPV in Japan may derive from either low occurrence or low reporting, or a combination of both. While some research indicates that unequal partner relationships may lead to increased risk of IPV (Morash, Bui, Zhang, & Holtfreter, 2007; Xu et al., 2005), other research suggests that firm patriarchal and hierarchal family and social structures in Japan may contribute to the low reported prevalence of IPV (Kumagai & Ishii-Kuntz, 2016). Several cultural factors such as (a) women’s economic dependence on men, (b) social stigma associated with divorce, and (c) the pressure on mothers to create an ideal environment for their children likely play a role in underreporting.

In 2018, Japan ranked 110th out of 149 countries in the World Economic Forum’s global gender equality ranking (World Economic Forum, 2018), reflecting low political and economic status and participation rates of women compared with men. While more women than men currently enroll in higher education, economic dependence on men continues to influence marriage rates and dynamics (Raymo & Iwasawa, 2005). Many women quit their jobs upon marriage or once they get pregnant, choosing to focus on their families (Borovoy, 2005), contributing not only to decreased lifetime earnings but also to decreased workplace opportunities, and leading to increased dependence on their husbands. These factors may contribute to reluctance to report IPV in the context of marriage due to economic and social fears of what might happen if the marriage ended.

There is significant stigma associated with divorce and single parenthood in Japan. Women contemplating divorce are often concerned about not only their own potential to get remarried but also about their own children’s potential to get married, as well as get into certain schools or companies. Thus, many women wait until their children are “safely” married before initiating their own divorces. Japan has one of the lowest “out of wedlock” births in the industrialized world (Hertog & Iwasawa, 2011), and women will often pressure men into a dekichatta kekkon (shotgun wedding), even if the men are abusive or absent (Hertog, 2009), because socially having a father is critical to children’s success in Japan.

Maternal–fetal bonding is considered to be critical for babies’ development in Japan by health practitioners and mothers, with strong emphasis on the ecosystemic environment of the fetus in Japanese prenatal care—as opposed to genetic determinism (Ivry, 2010). Most women faithfully record data from all prenatal visits in their government-issued “mother child health handbooks,” demonstrating the importance of maternal efforts to ensure optimal fetal development (Takeuchi, Sakagami, & Perez, 2016). Although the books contain questions about depression during the postpartum period, they do not have such specific questions during the prenatal period, and none that directly address IPV.

The Present Study

Based on previous studies, we hypothesized that emotional and physical IPV during pregnancy impacts maternal–infant bonding, which may be explained, in part, by PPD symptoms. The primary aim of this study was to examine the association between pregnancy IPV (emotional and physical) and maternal–infant bonding and investigate whether this relationship is mediated by PPD symptoms in Japan. For the study’s secondary aim, we used two subscales of maternal–infant bonding—(a) lack of affection and (b) anger and rejection—and assessed whether the mediating relationship remains the same for each subscale.

Method

Participants

In Japan, municipalities are responsible for promoting maternal and child health based on Maternal and Child Health Act. For a research project initiated by the National Center for Child Health and Development, municipalities in Aichi Prefecture in central Japan were approached to disseminate questionnaires to women enrolled in a 3- or 4-month postpartum health check-up program conducted by the municipalities. The purpose of the survey was to improve government services by understanding the daily lives of families with newborns, including how they manage child care. Among 54 municipalities in Aichi Prefecture, 45 took part in the study, including the prefecture capital, Nagoya city. Although monetary incentives were not provided to the participating municipalities, the research team compiled and provided results such as the prevalence of child abuse, PPD symptoms, and IPV to the government of the Aichi prefecture and each municipality. The provision of results could be considered the incentives for taking part in this research. The survey was conducted between October and November 2012. Almost all women (97.9%, N = 9,707) enrolled in the postpartum health check-up program in the participating municipalities were given a four-page questionnaire for the current study. Participants were informed that the survey was anonymous. For 11 municipalities, the questionnaire was distributed at the check-up centers and the participants returned the completed questionnaires to the center by mail. For the remaining 34 municipalities, the questionnaire was mailed to each participant before the check-up and the completed form was collected at the check-up center. Overall, 6,590 women (response rate 68% of women across all the municipalities) chose to participate, with response rates ranging from 24.2% to 81.0% depending on the municipality. Although the average response rate of municipalities overall was high (approximately 70%), the municipalities that received responses by mail showed relatively lower response rates. Our study was approved by the Ethics Committee of the National Center for Child Health and Development in Tokyo, Japan.

Assessment of Emotional and Physical IPV During Pregnancy

IPV during pregnancy was assessed using two questions from the Japanese version of revised Conflict Tactics Scales Short Form (J-CTS2SF) that focused on emotional and physical IPV (Straus, Hamby, Boney-McCoy, & Sggarman, 1996; Umeda & Kawakami, 2014). The original J-CTS2SF has been validated in Japan (Umeda & Kawakami, 2014). It includes five subscales with 20 total items. Because of limited time during health check-up visits and to minimize the respondent burden, only two questions from the full scale were included. Prior studies showed that these two questions are predictors of PPD and maternal abusive behavior toward infants (Amemiya & Fujiwara, 2016; Miura & Fujiwara, 2017). Emotional IPV was assessed with the following question: “While you were pregnant, were you ever insulted or verbally abused (cursed at, shouted at, spoken ill of) by your partner?” Physical IPV was assessed with the following question: “While you were pregnant, did a fight with your partner ever lead to him hitting or striking you, causing an injury?” The four response categories were never, rarely, sometimes, and often.

Assessment of PPD Symptoms

PPD symptoms were assessed using the Japanese version of the Edinburgh Postnatal Depression Scale (EPDS), a 10-item self-report questionnaire (Cox, Holden, & Sagovsky, 1987). The EPDS is the most widely used PPD measurement tool based on self-report, with total scores ranging from 0 to 30 (Boyd, Le, & Somberg, 2005; Cox et al., 1987). Higher scores indicate more severe PPD symptoms. The Japanese version of EPDS showed fairly high internal consistency (Cronbach’s α = .78) and good test–retest reliability (Spearman correlation = .92; Okano et al., 1996).

Assessment of Bonding

Maternal–infant bonding was evaluated using the Japanese version of the Maternal-Infant Bonding Scale (MIBS; Yoshida, Yamashita, Conroy, Marks, & Kumar, 2012). The original version of the MIBS was created by Kumar (1997), with responses recorded on a self-reported 10-item scale (Kumar, 1997). The modified Japanese version includes eight items, after exclusion of two items with low factor loadings. Each item has a 4-point Likert-type scale (from 0 = not at all to 3 = very much), with total scores ranging from 0 to 24. Higher scores indicate poorer maternal–infant bonding. Confirmatory factor analysis conducted in Japan demonstrated a two-factor structure among the eight items: one factor indicates lack of positive intimacy toward the baby (lack of affection, MIBS-LA) and the other factor reflects a mother’s anger and rejection toward her baby (anger and rejection, MIBS-AR). Both showed fairly high internal consistency (Cronbach’s α = .71 for MIBS-LA, and Cronbach’s α = .59 for MIBS-AR; Yoshida et al., 2012).

Covariates

Among the potential covariates, maternal age, paternal age, and household financial status were considered important theoretical confounders. Subjective household financial status was measured with four categories: stable, able to manage, difficult to manage, and unstable. This variable was dichotomized into manageable (stable and able to manage) versus not manageable (difficult to manage and unstable) to improve the parsimony of the final model. Maternal age was reported in years as a continuous variable. Paternal age was also reported in years and dichotomized into age equal to or older than 30 (reference group), and younger than 30 years on the basis of prior studies about factors associated with IPV (Inami et al., 2010; Kita et al., 2014).

Statistical Analysis

Of the women who completed the questionnaire, only those who completed the MIBS were included in our analysis (N = 6,459; 98% of the full sample). Women with a history of depressive symptoms during the year prior to pregnancy were excluded from the analysis to differentiate incident from chronic depressive symptoms (N = 935; remaining sample N = 5,524). The history of depressive symptoms was measured by the following question: “In the year prior to pregnancy, have you ever experienced periods of sleeplessness, agitation, weepiness, or lack of motivation that lasted for 2 weeks or longer?” We examined the distribution and frequency of demographic characteristics, IPV, PPD symptoms, and maternal–infant bonding in the sample. Due to the small number of positive responses for emotional and physical IPV, we dichotomized each of these variables to indicate the presence of any emotional or physical IPV during pregnancy. We also compared the difference between mean score of the EPDS and the MIBS for women exposed and unexposed to IPV. The effect sizes of these two groups were estimated using Cohen’s d (Cohen, 1998). PPD symptoms, bonding total scores, and bonding subscale scores (MIBS-LA and MIBS-AR) were included as continuous variables, all of which were positively skewed. Examined multiple data transformations to improve normality including a logarithm, square root, cube root, and reciprocal transformation. A logarithmic transformation was chosen because it was the best option for changing the skewed data into an approximately normal distribution.

We then constructed a path model, a type of structural equation model, to examine whether emotional and physical IPV were related to maternal–infant bonding and, furthermore, whether these relationships were mediated by PPD symptoms. Compared with traditional analytic strategies such as ordinary least squares (OLS) regression, structural equation models allow the simultaneous examination of various paths through a mediator to multiple outcomes (Bollen, 1989; Kline, 2005). In further analyses, we examined the two MIBS subscales, MIBS-LA and MIBS-AR, using these in place of the full maternal–infant bonding scale score. We allowed the two forms of IPV and two subscales of MIBS to be correlated. Maternal age and household financial status were included as potential confounders of the relationship between the two forms of IPV and PPD symptoms. To distinguish older from younger fathers, we included a dichotomized covariate of paternal age younger than 30 and paternal age equal to or older than 30.

We then calculated the direct, indirect, and total effect of each form of IPV on bonding as well as the direct effect of PPD symptoms on bonding. We evaluated the fit of the model using the following fit indices and cutoffs indicating good fit: root mean square error approximation (RMSEA < .06), standardized root mean square residual (SRMR < .08), comparative fit index (CFI > .95), and Tucker–Lewis Index (TLI > .95; Bollen, 1989; Bryan, Schmiege, & Broaddus, 2007). These models reflect a complete case analysis, which was selected based on low levels of missingness (i.e., 3.8% of data were missing on all of the variables included in the path analysis) after excluding 2.0% missing data on maternal–infant bonding and dropping 14% of the participants with a history of depressive symptoms during the year prior to pregnancy.

To test the robustness of the results, we conducted a sensitivity analysis by including participants with a history of depressive symptoms in our depressed group. All statistical analyses were performed with Stata 15.1 (StataCorp, College Station, TX, USA).

Results

Table 1 displays the participant descriptive statistics (N = 5,524). The mean age of women was 31.56 years (SD = 4.72), and most women were married or living with a partner (98%). Seventy-seven percent were unemployed (or employed women who were taking parental leave), and 10% reported that their financial status was difficult to manage or unstable. The mean scores for the Japanese version of EPDS and MIBS were 7.19 (SD = 1.85) and 1.38 (SD = 1.77), respectively.

Table 1.

Characteristics of the Study Participants.

	Total (N = 5,524)
	n or M	% or SD
Maternal age (years)	31.56	4.72
Paternal age
<30	1,314	23.79
30+	4,210	76.21
Marital status
Married/living with partner	5,438	98.44
Single/divorced/widowed	74	1.34
Missing	12	0.22
Maternal working status
Not working	4,271	77.32
Full-time	934	16.91
Part-time	256	4.63
Missing	63	1.14
Subjective economic status
Stable	2,551	46.18
Able to manage	2,225	40.28
Difficult to manage or unstable	542	9.81
Missing	206	3.73
Sex of the child
Male	2,791	50.52
Female	2,685	48.61
Missing	48	0.87
Birth weight of the child
2,500 g+	5,054	91.49
<2,500	449	8.13
Missing	21	0.38
Multiple birth
Single	5,431	98.32
Twin	93	1.68
Pregnant with first child
Yes	2,768	50.11
No	2,741	49.62
Missing	15	0.27
EPDS score
Mean	7.19	1.85
MIBS score
Mean	1.38	1.77

Note. EPDS = Edinburgh Postnatal Depression Scale; MIBS = Maternal-Infant Bonding Scale.

Table 2 reports the co-occurrence of emotional and physical IPV (dichotomized). More participants experienced emotional IPV (8.7%) than physical IPV (0.9%). Among the 47 participants who experienced physical IPV, 34 also experienced emotional IPV. Most of the participants who were exposed to emotional IPV did not experience physical IPV. The odds of being exposed to the emotional IPV was 29.6 times higher in those exposed to physical IPV, compared with those not exposed to physical IPV (95% confidence interval = [15.06, 61.44]).

Table 2.

Associations Between Emotional IPV and Physical IPV During Pregnancy (N = 5,524).

	Emotional IPV
	Unexposed (%)	Exposed (%)	No Response (%)	Total (%)
Physical IPV
Unexposed (%)	5,010 (90.70)	445 (8.06)	15 (0.27)	5,470 (99.02)
Exposed (%)	12 (0.22)	34 (0.62)	1 (0.02)	47 (0.85)
No response (%)	2 (0.04)	0 (0.00)	5 (0.09)	7 (0.13)
Total	5,024 (90.95)	479 (8.67)	21 (0.38)	5,524 (100.00)

Note. IPV = intimate partner violence.

Table 3 shows the mean scores of the EPDS, full scores of the MIBS, subscales of bonding (MIBS-LA and MIBS-AR), and the effect size estimates stratified by the emotional and physical IPV exposures. Participants who were exposed to emotional IPV had statistically significant higher scores for all scales compared with participants unexposed to emotional IPV. All scores were higher for women who were exposed to physical IPV than women who were not, and EPDS scores of women who experienced physical IPV were 2.19 point higher than the unexposed (p < .001). For all other scores, no differences were found between exposed and unexposed (p = .62, p = .56, p = .81). The mean difference of EPDS scores between women exposed and unexposed to physical IPV indicated a significantly large effect size with a standard deviation of 1.18, whereas differences in EPDS scores between women exposed and unexposed to emotional IPV showed a medium effect size estimate.

Table 3.

Mean Scores and Effect Size Estimates of EPDS, MIBS, and MIBS-LA and MIBS-AR Stratified by the Exposure of Emotional and Physical IPV.

	Emotional IPV				Physical IPV
	Exposed, M (SD)	Unexposed, M (SD)	p Value	Effect Size Estimate	Exposed, M (SD)	Unexposed, M (SD)	p Value	Effect Size Estimate
EPDS	8.23 (2.59)	7.09 (1.74)	<.001	.62*	9.36 (3.80)	7.17 (1.82)	<.001	1.18*
MIBS	1.65 (1.88)	1.35 (1.75)	<.001	.17*	1.51 (1.82)	1.38 (1.77)	.618	0.07
MIBS-LA	0.72 (1.17)	0.53 (1.03)	<.001	.18*	0.64 (1.17)	0.55 (1.05)	.557	0.09
MIBS-AR	0.94 (1.14)	0.82 (1.13)	.033	.10*	0.87 (1.10)	0.83 (1.13)	.813	0.03

Note. EPDS = Edinburgh Postnatal Depression Scale; MIBS = Maternal-Infant Bonding Scale; MIBS-LA = Maternal-Infant Bonding Scale–Lack of Affection; MIBS-AR = Maternal-Infant Bonding Scale–Anger and Rejection; IPV = intimate partner violence.

The effect size estimate is statistically significantly different at the .05 level.

Our first path analysis model (Model 1) was used to assess the relationship between emotional IPV, physical IPV, and poor bonding as well as the role of PPD symptoms as a mediator (Figure 1). We included the logarithmically transformed PPD symptom and bonding variables in these models to fulfill assumptions associated with the use of a normal distribution. The model showed good fit (RMSEA = 0.030, SRMR = 0.013, CFI = 0.978, TLI = 0.903). Both emotional and physical IPV were directly related to PPD symptoms (β = .13, p < .001 for emotional IPV; β = .05, p < .001 for physical IPV), but not to poor maternal–infant bonding. PPD symptoms were directly related to poor maternal–infant bonding (β = .24, p < .001). Furthermore, we investigated the direct, indirect, and total effects of emotional and physical IPV on the poor maternal–infant bonding through PPD symptoms. As reported in Table 4(a), emotional IPV was significantly associated with poor bonding (β = .06, p < .001). This relationship had a nonstatistically significant direct effect (β = .02, p = .08) and a statistically significant indirect effect (β = .03, p < .001), suggesting that the relationship between emotional IPV and poor bonding is largely explained by PPD symptoms. We adopted a calculation of effect sizes for mediation models and presented the ratio of indirect to total effects (Wen & Fan, 2015). For poor maternal–infant bonding, the indirect effect size of emotional IPV through PPD symptoms was about 0.5 (0.03 [indirect effect]/0.06 [total effect]). Physical IPV had neither a statistically significant direct effect nor a total effect on poor bonding.

Figure 1.

Final mediation model including full scores of MIBS (N = 5,438).

Table 4.

Associations Between IPV and Total Scores, and Subscales of MIBS.

Outcome	Exposure	Direct Effect	Indirect Effect	Total Effect
(a) Total score of MIBS	Emotional IPV	0.02 (p = .077)	0.03 (p < .001)	0.06 (p < .001)
(a) Total score of MIBS	Physical IPV	−0.02 (p = .120)	0.01 (p < .001)	−0.01 (p = .530)
(b) Lack of affection	Emotional IPV	0.04 (p = .006)	0.02 (p < .001)	0.06 (p < .001)
(b) Lack of affection	Physical IPV	−0.02 (p = .217)	0.01 (p < .001)	−0.01 (p = .541)
(c) Anger and rejection	Emotional IPV	0.001 (p = .913)	0.03 (p < .001)	0.03 (p = .018)
(c) Anger and rejection	Physical IPV	−0.02 (p = .181)	0.01 (p < .001)	−0.01 (p = .676)

Note. IPV = intimate partner violence; MIBS = Maternal-Infant Bonding Scale.

Contrary to our hypotheses, we did not detect a significant total effect of physical IPV on poor bonding (β = –.01, p = .53). Surprisingly, we found that the indirect (via PPD symptoms) and direct pathways have opposing effects on bonding, which negate one another when examined together (i.e., total effect). When the total effect is deconstructed, the direct effect was not statistically significant (β = –.02, p < .12), whereas the indirect effect through PPD symptoms was statistically significant (β = .01, p < .001). The final model is presented in Figure 1 with standardized parameters.

To understand the effect of emotional and physical IPV on the MIBS-LA and MIBS-AR subscales via PPD symptoms, we included two subscales of MIBS in place of the full scores of MIBS (Model 2). MIBS-AR had a higher mean score for the subscale (M = 0.83, SD = 1.13) than that of the MIBS-LA (M = 0.55, SD = 1.05). Presented in Figure 2, the path analysis model showed good fit (RMSEA = 0.036, SRMR = 0.019, CFI = 0.970, TLI = 0.894). In Figure 2, the direct effects of emotional/physical IPV on PPD symptoms were the same as those in Figure 1. The direct effect of PPD symptoms on both MIBS-LA and MIBS-AR were statistically significant (β = .16, p < .001 for MIBS-LA; and β = .24, p < .001 for MIBS-AR). However, all direct effects from both forms of IPV on both subscales of MIBS were not statistically significant except the effect of emotional IPV on MIBS-LA (β = .04, p = .006). The direct, indirect, and total effects of emotional/physical IPV on the subscales of MIBS-LA and MIBS-AR are shown in Table 4(b) and (c). Regarding the total effects, only emotional IPV on MIBS-LA and MIBS-AR was statistically significant (β = .06, p < .001 for LA; and β = .03, p = .02 for AR). All indirect effects via PPD symptoms were statistically significant, suggesting that the relationship between emotional and physical IPV on the two forms of poor maternal–infant bonding, lack of affection and anger and rejection, were mediated by PPD symptoms.

Figure 2.

Final mediation model including subscales of MIBS-LA, and MIBS-AR (N = 5,438).

We then calculated the effect sizes of the mediation model, the ratio of indirect to total effects (Wen & Fan, 2015). For MIBS-LA and MIBS-AR, the indirect effect sizes of emotional IPV through PPD symptoms were 0.33 (0.02 [indirect effect]/0.06 [total effect]), and 1 (0.03 [indirect effect]/0.03 [total effect]), respectively. The effect sizes indicated that the effect of emotional IPV on MIBS-LA was partially mediated by PPD symptoms, whereas the impact of emotional IPV on MIBS-AR was fully mediated by PPD symptoms. For physical IPV on MIBS-LA and MIBS-AR, all direct and total effects were not statistically significant. Similar to what was observed in the primary analyses, the null total effect of physical IPV on bonding is comprised of a negative direct effect and a positive indirect effect. The final models for both MIBS-LA and MIBS-AR are presented in Figure 2 with standardized parameters.

The results of the sensitivity analysis, which included participants with a history of depressive symptoms in the depressed group, were not substantially different from our original results that excluded those participants (data not shown).

Discussion

Our study found a modest indirect association between emotional and physical IPV and poor maternal–infant bonding among Japanese postnatal women, which was mediated by PPD symptoms. The direct effects between both forms of IPV (emotional and physical) and bonding were not statistically significant, suggesting that the primary pathway by which IPV is associated with poor bonding is through increased PPD symptoms. Further examination of these relationships in our subanalyses of bonding subscales showed that emotional IPV was statistically significantly associated with MIBS-LA and MIBS-AR and they were mediated by PPD symptoms.

These findings are consistent with previous literature reporting associations between IPV during pregnancy and PPD symptoms (Kabir, Nasreen, & Edhborg, 2014; Kita et al., 2016; Miura & Fujiwara, 2017; Tsai, Tomlinson, Comulada, & Rotheram-Borus, 2016) as well as an association between PPD symptoms and maternal–infant bonding (Baines, Wittkowski, & Wieck, 2013; Edhborg et al., 2005; O’Higgins et al., 2013). Our research extends the findings to integrate these associations into a single framework that further clarifies the relative roles of IPV and PPD symptoms in relation to maternal–infant bonding. Although prevention of IPV is the ultimate solution for improving maternal–infant bonding among women who are exposed to IPV during pregnancy, secondary strategies such as treating PPD symptoms and counseling victims could improve maternal-infant bonding, if complete prevention of IPV is not possible. Given the findings that PPD symptoms largely explained the association between emotional IPV and maternal–infant bonding, the screening, prevention and the treatment of PPD symptoms during pregnancy may be critical to prevent the effects of IPV on poor maternal–infant bonding. Depression can be successfully treated in the postpartum period (Hantsoo et al., 2014). Increasing evidence surrounding the effective integration of depression treatment modalities into postpartum care is emerging (Dennis & Dowswell, 2013) and may serve as effective models for enhancing maternal–infant bonding postpartum in Japan. In addition to improving IPV victims’ coping skills, referring them to counseling or providing them resources and information (Bailey, 2010; Mitchell et al., 2006) may facilitate recovery, potentially decrease PPD symptoms, and improve maternal–infant bonding. A qualitative study from Japan found that counseling services and group meetings at mental health clinics helped victims recover from IPV (Nemoto, Rodriguez, & Mkandawire-Valhmu, 2008). Intervention programs for male perpetrators would be another route to reduce the impact of IPV on PPD and maternal–infant bonding (Eckhardt, Murphy, Black, & Suhr, 2006). Furthermore, it may be important that providers focus on promoting positive intimacy as well as reducing anger and rejection toward infants among mothers exposed to emotional IPV as indicated by results of our subanalyses, which found a strong association between emotional IPV and MIBS-LA and MIBS-AR.

We also identified qualitative differences in the way in which different forms of IPV may relate to depressive symptoms and, subsequently, maternal–infant bonding. Most studies have investigated the aggregate impact of IPV, rather than distinguishing between different types of IPV. Kita and colleagues collected two waves of survey data, first in the third trimester and then at 1-month postpartum among 562 Japanese women living in a suburb of Tokyo. The authors used a structural equation model to show that IPV during pregnancy was associated with poor maternal–infant bonding at 1-month postpartum. However, exposure to IPV was not directly associated with PPD; rather, the relationship had an indirect effect through antenatal depression (Kita et al., 2016). Although we did not include participants who reported antenatal depression, our results clearly suggest that emotional and physical IPV during pregnancy are both directly associated with PPD symptoms and, subsequently, may be related to poorer bonding via increased PPD symptoms. Another small study in the United Kingdom (N = 38) compared maternal–fetus bonding between mothers who were exposed to IPV during pregnancy (N = 11) and those who were not (N = 27). The authors found that mothers who were abused during pregnancy are less likely to bond well with their fetuses and subsequently their infants, and the severity of abuse was correlated with the extent of poor bonding (Zeitlin et al., 1999). These results are consistent with our findings regarding the total effect of emotional IPV on poorer bonding. In contrast, in a small study of Latinas living in the United States who experienced traumatic events in adulthood, such as IPV, women who had experienced traumatic events had better postpartum bonding with infants than women who did not experience traumatic events (Lara-Cinisomo et al., 2018).

The inconsistencies observed across studies (Kita et al., 2016; Lara-Cinisomo et al., 2018; Zeitlin et al., 1999) may be a result of the differential recall periods (e.g., IPV in the past month vs. in the past year) or differences in sample characteristics. These inconsistencies may also reflect differences in the measurement of IPV. While many studies combined types of IPV to measure its impact on women’s health, our analysis differentiated emotional and physical IPV, and dichotomized each variable into exposed versus unexposed (due to data skewness and the low prevalence of any form of IPV in our sample). Respective patterns of each form of IPV were distinct; while women who were exposed to physical IPV had 30 times higher odds of being exposed to emotional IPV, women who were exposed to emotional IPV did not necessarily experience physical IPV. Commonly, IPV begins with sporadic emotional abuse that evolves into a more severe form of violence (Feld & Straus, 1989; Reichenheim, Moraes, Lopes, & Lobato, 2014). In our study, the high overlap of physical and emotional IPV, and the relatively large number of participants who experienced emotional IPV raise the possibility of IPV progressing into a more severe form. In addition, despite a common assumption that emotional abuse results in less of a psychological toll than physical abuse (O’Leary, 1999), our findings extend previous research on the importance of emotional IPV in the absence of physical and/or sexual abuse. Emotional IPV can be as detrimental as physical or sexual IPV, and is considered more acceptable particularly in patriarchal societies where men tend to be more controlling (Keashly, 1998; Lammers, Ritchie, & Robertson, 2005; Tiwari et al., 2008). Our findings suggest that covert emotional IPV may be important.

The findings of this study should be interpreted in light of several limitations. First, despite having sufficient power for the emotional IPV outcome, statistical power for the physical IPV outcome was low because only 47 (0.86%) women reported physical IPV. Specifically, we found the prevalence of any IPV to be lower than that reported in prior research from hospital-based studies in Japan (Inami et al., 2010; Kataoka et al., 2016; Kita et al., 2014), which might be due to the population-based nature of our study. According to the World Health Organization, being slapped, pushed, or shoved are considered moderate physical violence, whereas being hit with a fist, kicked, dragged, threatened with a weapon, or having a weapon used against a woman are defined as severe physical violence (World Health Organization, 1996). Our question, asking whether the participants were hit or punched resulting in injury, may have influenced participants to use a more severe definition of IPV and therefore be less likely to disclose mild or moderate levels of physical IPV. The assessment of emotional IPV was limited to being humiliated or yelled at, which may similarly result in underestimation of the prevalence. Second, this study was cross-sectional so the findings cannot be interpreted as causal. Future longitudinal studies should evaluate the mediating role of PPD in the relationship between IPV and bonding to specifically validate the temporal relationships between these related constructs. Also, because we lacked data on exposure to IPV before pregnancy, we may not have captured women who were victimized before pregnancy. Third, all measures included in this analysis relied on self-report and asked about culturally sensitive topics such as history of depressive symptoms, IPV, PPD symptoms, and maternal–infant relationships. Although our measure of history of depressive symptoms in the past year may lack precision (i.e., it was assessed with one question asking about feelings of depression, irritation, or sadness), sensitivity analyses showed that including participants who indicated a history of symptoms in our depressed group did not change our results substantially, relative to the analyses excluding these participants. This suggests that the association between IPV and bonding via postpartum depression can be observed regardless of depressive symptoms during pregnancy. Fourth, to minimize respondent burden, though using items from a standard questionnaire with high face-validity, our IPV measures only included two single questions. Also, sexual IPV was not assessed because of its low prevalence/reporting in Japan (between 0.5% and 1.5%; Fujiwara & Kawakami, 2011). The use of these measures may yield an underestimation of IPV. To minimize potential underreporting in Japanese culture, future studies need to consider improving measurement methods and evaluating potential biases associated with social desirability and stigma of reporting.

Despite the limitations, our study has strengths. First, this study is one of the first to explore mediators of the relationship between IPV and maternal–infant bonding. Mediators that have been previously examined include reduced oxytocin levels, which were found to be on the pathway between adult trauma, including exposure to IPV, and weakened bonding (Samuel et al., 2015). Psychological mediators examined previously include retrospective measures, such as perinatal depression (Kita et al., 2016), poor maternal–fetus bonding (Zeitlin et al., 1999), and impairment in parenting capacity (Bogat et al., 2011). Although the effect size estimates are relatively small except for the estimates between PPD symptoms and poor maternal–infant bonding, our study sheds light on PPD symptoms as a mediator, which can be treated concurrently in the postpartum period with other strategies to improve poor maternal–infant bonding. Using path analysis that allowed for simultaneous exploration of multiple forms of IPV, PPD symptoms, and poor bonding provides a comprehensive picture of the relationship. Furthermore, the goodness-of-fit indices of our model indicate that an a priori model fit our data well (Hooper, Coughlan, & Mullen, 2008). The fit of our path analysis for the total score (Model 1) and subscale scores of MIBS-LA and MIBS-AR (Model 2) was good. Second, most studies evaluating the association between IPV and maternal–infant bonding have had relatively small sample sizes and have produced inconsistent results. Despite the low prevalence of IPV, our study was large (N = 5,524) and was representative of the communities that participated in a 3- or 4-month postpartum health check-up program, for which the enrollment proportion was 97.9%. Third, our study is the first to examine these relationships using maternal–infant bonding subscales, which revealed different relationships between IPV and the two subscales. Most prior studies have focused on the impacts of physical and/or sexual violence (World Health Organization, 2011). Our analysis is one of the first to reveal the role of emotional IPV on PPD symptoms and bonding among women during the postpartum period. This study builds upon prior studies of Japanese women by including a larger, possibly more representative sample of women with greater economic and geographic variability compared with previous studies (Kita et al., 2016).

Conclusions

IPV, PPD symptoms, and poor maternal–infant bonding are critical correlates of poor maternal and child outcomes. In our analysis, we found that the relationship between IPV and poor bonding was largely mediated by PPD symptoms. While prevention of IPV is the ultimate goal, our results suggest that screening for PPD and interventions that address PPD might mitigate multigenerational effects of IPV during pregnancy. Identifying opportunities for detection of IPV and PPD symptoms, as well as prevention and early intervention, may improve maternal–infant bonding; however, further longitudinal studies are needed to clarify the temporality of these relationships.

Footnotes

Declaration of Conflicting Interests

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

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was partially supported by the following grants for T.F.: Grant-in-Aid for Young Scientists (B) Scientific Research, from the Ministry of Education, Culture, Sports, Science and Technology (KAKENHI21790593) and the Ministry of Health, Labor, and Welfare (H23-Seisaku-Ippan-005) T.F. is also supported by JST RISTEX Grant Number JPMJRX16G5, Japan. S.P. is supported by the International Fellowship from American Associations for University Women (AAUW). M.C.G. is supported by the National Institute on Drug Abuse (T32DA007292) and the National Institute of Mental Health (T32MH096724).

ORCID iDs

Soim Park

Takeo Fujiwara

Author Biographies

Soim Park is a doctoral student in the Social and Behavioral Interventions Program at the Johns Hopkins Bloomberg School of Public Health. Her work focuses on social determinants of health especially in marginalized populations.

M. Claire Greene is a postdoctoral research fellow in global mental health implementation science at Columbia University and New York State Psychiatric Institute. Her research focuses on developing, implementing, and evaluating interventions for co-occurring mental health problems, substance use, and violence in humanitarian settings.

Melissa K. Melby is a medical anthropologist and professor in the Department of Anthropology, College of Arts and Sciences, at the University of Delaware. She is also a fellow in Canadian Institute for Advanced Research’s (CIFAR) Humans & the Microbiome program. Her research examines biocultural factors influencing health across the life course.

Takeo Fujiwara is a social epidemiologist and professor in the Department of Global Health Promotion at Tokyo Medical and Dental University. His research includes prevention of child abuse, mental health, and life course epidemiology.

Pamela J. Surkan is a social epidemiologist and associate professor in the Social and Behavioral Interventions Program at Johns Hopkins Bloomberg School of Public Health. Her research largely centers on maternal child health, mental health, and child development.

References

Ahmad

N. A.

Silim

U. A.

Rosman

Mohamed

Chan

Y. Y.

Mohd Kasim

. . . Aris

(2018). Postnatal depression and intimate partner violence: A nationwide clinic-based cross-sectional study in Malaysia. BMJ Open, 8(5), e020649. doi:10.1136/bmjopen-2017-020649

Alhusen

J. L.

Ray

Sharps

Bullock

(2015). Intimate partner violence during pregnancy: Maternal and neonatal outcomes. Journal of Women’s Health, 24, 100-106. doi:10.1089/jwh.2014.4872

Amemiya

Fujiwara

(2016). Association between maternal intimate partner violence victimization during pregnancy and maternal abusive behavior towards infants at 4 months of age in Japan. Child Abuse & Neglect, 55, 32-39. doi:10.1016/j.chiabu.2016.03.008

Bailey

B. A.

(2010). Partner violence during pregnancy: Prevalence, effects, screening, and management. International Journal of Women’s Health, 2, 183-197.

Baines

Wittkowski

Wieck

(2013). Illness perceptions in mothers with postpartum depression. Midwifery, 29, 779-786. doi:10.1016/j.midw.2012.06.020

Beydoun

H. A.

Al-Sahab

Beydoun

M. A.

Tamim

(2010). Intimate partner violence as a risk factor for postpartum depression among Canadian women in the maternity experience survey. Annals of Epidemiology, 20, 575-583. doi:10.1016/j.annepidem.2010.05.011

Bicking Kinsey

Hupcey

J. E

. (2013). State of the science of maternal-infant bonding: A principle-based concept analysis. Midwifery, 29, 1314-1320. doi:10.1016/j.midw.2012.12.019

Bogat

G. A.

Levendosky

A. A.

von Eye

Davidson

W. S. I.

(2011). Effects of intimate partner violence on the attachment relationship between mother and child: Data from a longitudinal study beginning during pregnancy. In Graham-Bermann

S. A.

Levendosky

A. A.

(Eds.), How intimate partner violence affects children: Developmental research, case studies, and evidence-based intervention (pp. 19-46). Washington, DC: American Psychological Association.

Bollen

K. A.

(1989). Structural equations with latent variables. New York, NY: John Wiley.

10.

Bonomi

A. E.

Anderson

M. L.

Rivara

F. P.

Thompson

R. S.

(2007). Health outcomes in women with physical and sexual intimate partner violence exposure. Journal of Women’s Health, 16, 987-997. doi:10.1089/jwh.2006.0239

11.

Borovoy

(2005). The too-good wife: Alcohol, codependency, and the politics of nurturance in postwar Japan. Berkeley: University of California Press.

12.

Boyd

R. C.

H. N.

Somberg

(2005). Review of screening instruments for postpartum depression. Archives of Women’s Mental Health, 8, 141-153. doi:10.1007/s00737-005-0096-6

13.

Breiding

M. J.

Basile

K. C.

Smith

S. G.

Black

M. C.

Mahendra

(2015). Intimate partner violence surveillance: Uniform definitions and recommended data elements. Atlanta, GA: National Center for Injury Prevention and Control, Centers for Disease Control and Prevention.

14.

Bryan

Schmiege

S. J.

Broaddus

M. R.

(2007). Mediational analysis in HIV/AIDS research: Estimating multivariate path analytic models in a structural equation modeling framework. AIDS and Behavior, 11, 365-383. doi:10.1007/s10461-006-9150-2

15.

Campbell

J. C.

(2002). Health consequences of intimate partner violence. The Lancet, 359, 1331-1336. doi:10.1016/S0140-6736(02)08336-8

16.

Campbell

J. C.

García-Moreno

Sharps

(2004). Abuse during pregnancy in industrialized and developing countries. Violence Against Women, 10, 770-789. doi:10.1177/1077801204265551

17.

Campbell

S. B.

Cohn

J. F.

Meyers

(1995). Depression in first-time mothers: Mother-infant interaction and depression chronicity. Developmental Psychology, 31, 349-357. doi:10.1037/0012-1649.31.3.349

18.

Cohen

(1998). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, NJ: Lawrence Erlbaum.

19.

Cox

J. L.

Holden

J. M.

Sagovsky

(1987). Detection of postnatal depression: Development of the 10-item Edinburgh postnatal depression scale. British Journal of Psychiatry, 150, 782-786. doi:10.1192/bjp.150.6.782

20.

Dennis

C.-L.

Dowswell

(2013). Psychosocial and psychological interventions for preventing postpartum depression. Cochrane Database of Systematic Reviews, 2, CD001134. doi:10.1002/14651858.CD001134.pub3

21.

Devries

K. M.

Kishor

Johnson

Stöckl

Bacchus

L. J.

Garcia-Moreno

Watts

(2010). Intimate partner violence during pregnancy: Analysis of prevalence data from 19 countries. Reproductive Health Matters, 18, 158-170. doi:10.1016/S0968-8080(10)36533-5

22.

Dillon

Hussain

Loxton

Rahman

(2013). Mental and physical health and intimate partner violence against women: A review of the literature. International Journal of Family Medicine, 2013, Article 313909. doi:10.1155/2013/313909

23.

Eckhardt

C. I.

Murphy

Black

Suhr

(2006). Intervention programs for perpetrators of intimate partner violence: Conclusions from a clinical research perspective. Public Health Reports, 121, 369-381. doi:10.1177/003335490612100405

24.

Edhborg

Matthiesen

A. S.

Lundh

Widström

A. M.

(2005). Some early indicators for depressive symptoms and bonding 2 months postpartum—A study of new mothers and fathers. Archives of Women’s Mental Health, 8, 221-231. doi:10.1007/s00737-005-0097-5

25.

Feld

S. L.

Straus

M. A.

(1989). Escalation and desistance of wife assault in marriage. Criminology, 27, 141-162. doi:10.1111/j.1745-9125.1989.tb00866.x

26.

Fujiwara

Kawakami

(2011). Association of childhood adversities with the first onset of mental disorders in Japan: Results from the World Mental Health Japan, 2002-2004. Journal of Psychiatric Research, 45, 481-487.

27.

Gazmararian

J. A.

Peterson

Spitz

A. M.

Goodwin

M. M.

Saltzman

L. E.

Marks

J. S.

(2000). Violence and reproductive health: Current knowledge and future research directions. Maternal and Child Health Journal, 4, 79-84.

28.

Greenberg

M. T.

(1999). Attachment and psychopathology in childhood. In Cassidy

Shaver

P. R.

(Eds.), Handbook of attachment (pp. 469-498). New York, NY: Guilford Press.

29.

Hantsoo

Ward-O’Brien

Czarkowski

K. A.

Gueorguieva

Price

L. H.

Epperson

C. N.

(2014). A randomized, placebo-controlled, double-blind trial of sertraline for postpartum depression. Psychopharmacology, 231, 939-948. doi:10.1007/s00213-013-3316-1

30.

Heise

Garcia-Moreno

(2002). Violence by intimate partners. In Krug

E. G.

Dahlberg

L. L.

Mercy

J. A.

Zwi

A. B.

Lozano

(Eds.), World report on violence and health (pp. 87-121). Geneva, Switzerland: World Health Organization. Retrieved from http://apps.who.int/iris/bitstream/handle/10665/42495/9241545615_eng.pdf?sequence=1

31.

Hertog

(2009). Tough choices: Bearing an illegitimate child in Japan. Stanford, CA: Stanford University Press.

32.

Hertog

Iwasawa

(2011). Marriage, abortion, or unwed motherhood? How women evaluate alternative solutions to premarital pregnancies in Japan and the United States. Journal of Family Issues, 32, 1674-1699. doi:10.1177/0192513X11409333

33.

Hooper

Coughlan

Mullen

M. R.

(2008). Structural equation modelling: Guidelines for determining model fit. The Electronic Journal of Business Research Methods, 6, 53-60. doi:10.1037/1082-989X.12.1.58

34.

Huth-Bocks

A. C.

Levendosky

A. A.

Bogat

G. A.

(2002). The effects of domestic violence during pregnancy on maternal and infant health. Violence and Victims, 17, 169-185. doi:10.1891/vivi.17.2.169.33647

35.

Inami

Kataoka

Eto

Horiuchi

(2010). Intimate partner violence against Japanese and non-Japanese women in Japan: A cross-sectional study in the perinatal setting. Japan Journal of Nursing Science, 7, 84-95. doi:10.1111/j.1742-7924.2010.00140.x

36.

Islam

M. J.

Broidy

Baird

Mazerolle

(2017). Intimate partner violence around the time of pregnancy and postpartum depression: The experience of women of Bangladesh. PLoS ONE, 12, Article e0176211. doi:10.1371/journal.pone.0176211

37.

Ivry

(2010). Embodying culture: Pregnancy in Japan and Israel. New Brunswick, NJ: Rutgers University Press.

38.

Kabir

Z. N.

Nasreen

H.-E.

Edhborg

(2014). Intimate partner violence and its association with maternal depressive symptoms 6–8 months after childbirth in rural Bangladesh. Global Health Action, 7, 24725. doi:10.3402/gha.v7.24725

39.

Kataoka

Imazeki

Shinohara

(2016). Survey of intimate partner violence before and during pregnancy among Japanese women. Japan Journal of Nursing Science, 13, 189-195. doi:10.1111/jjns.12093

40.

Keashly

(1998). Emotional abuse in the workplace: Conceptual and empirical issues. Journal of Emotional Abuse, 1, 85-117. doi:10.1300/J135v01n01

41.

Kita

Haruna

Matsuzaki

Kamibeppu

(2016). Associations between intimate partner violence (IPV) during pregnancy, mother-to-infant bonding failure, and postnatal depressive symptoms. Archives of Women’s Mental Health, 19, 623-634. doi:10.1007/s00737-016-0603-y

42.

Kita

Yaeko

Porter

S. E.

(2014). Prevalence and risk factors of intimate partner violence among pregnant women in Japan. Health Care for Women International, 35, 442-457. doi:10.1080/07399332.2013.857320

43.

Klaus

M. H.

Kennell

J. H.

(1976). Maternal-infant bonding. Saint Louis, MO: C.V. Mosby.

44.

Kline

R. E.

(2005). Principles and practice of structural equation modeling (2nd ed.). New York, NY: Guilford Press.

45.

Kochanska

Aksan

Knaack

Rhines

H. M.

(2004). Maternal parenting and children’s conscience: Early security as moderator. Child Development, 75, 1229-1242. doi:10.1111/j.1467-8624.2004.00735.x

46.

Koenig

L. J.

Whitaker

D. J.

Royce

R. A.

Wilson

T. E.

Ethier

Fernandez

M. I.

(2006). Physical and sexual violence during pregnancy and after delivery: A prospective multistate study of women with or at risk for HIV infection. American Journal of Public Health, 96, 1052-1059. doi:10.2105/AJPH.2005.067744

47.

Kumagai

Ishii-Kuntz

(2016). Family violence in Japan: A life course perspective. Singapore: Springer.

48.

Kumar

R. C.

(1997). “Anybody’s child”: Severe disorders of mother-to-infant bonding. British Journal of Psychiatry, 171, 175-181. doi:10.1192/bjp.171.2.175

49.

Lammers

Ritchie

Robertson

(2005). Women’s experience of emotional abuse in intimate relationships. Journal of Emotional Abuse, 5, 29-64. doi:10.1300/j135v05n01_02

50.

Lara-Cinisomo

Zhu

Fei

Weston

A. P.

Ravat

(2018). Traumatic events: Exploring associations with maternal depression, infant bonding, and oxytocin in Latina mothers. BMC Women’s Health, 18, 1-9. doi:10.1186/s12905-018-0520-5

51.

Lau

Chan

K. S.

(2007). Influence of intimate partner violence during pregnancy and early postpartum depressive symptoms on breastfeeding among Chinese women in Hong Kong. Journal of Midwifery & Women’s Health, 52(2), e15-e20. doi:10.1016/j.jmwh.2006.09.001

52.

Levendosky

A. A.

Lannert

Yalch

(2012). The effects of intimate partner violence on women and child survivors: An attachment perspective. Psychodynamic Psychiatry, 40, 397-433. doi:10.1521/pdps.2012.40.3.397

53.

Ludermir

A. B.

Lewis

Valongueiro

S. A.

de Araújo

T. V. B.

Araya

(2010). Violence against women by their intimate partner during pregnancy and postnatal depression: A prospective cohort study. The Lancet, 376, 903-910. doi:10.1016/S0140-6736(10)60887-2

54.

Mäntymaa

Puura

Luoma

Salmelin

Davis

Tsiantis

. . . Tamminen

(2003). Infant-mother interaction as a predictor of child’s chronic health problems. Child: Care, Health and Development, 29, 181-191. doi:10.1046/j.1365-2214.2003.00330.x

55.

Martin

S. L.

Harris-Britt

Morraco

K. E.

Kupper

L. L.

Campbell

J. C.

(2004). Changes in intimate partner violence during pregnancy. Journal of Family Violence, 19, 201-210.

56.

Martins

Gaffan

E. A.

(2000). Effects of early maternal depression on patterns of infant-mother attachment: A meta-analytic investigation. Journal of Child Psychology and Psychiatry, 41, 737-746. doi:10.1111/1469-7610.00661

57.

McMahon

Armstrong

D. Y.

(2012). Intimate partner violence during pregnancy: Best practices for social workers. Health & Social Work, 37, 9-17. doi:10.1093/hsw/hls004

58.

Mechanic

M. B.

Weaver

T. L.

Resick

P. A.

(2008). Mental health consequences of intimate partner abuse: A multidimensional assessment of four different forms of abuse. Violence Against Women, 14, 634-654. doi:10.1177/1077801208319283

59.

Mitchell

M. D.

Hargrove

G. L.

Collins

M. H.

Thompson

M. P.

Reddick

T. L.

Kaslow

N. J.

(2006). Coping variables that mediate the relation between intimate partner violence and mental health outcomes among low-income, African American women. Journal of Clinical Psychology, 62, 1503-1520. doi:10.1002/jclp.20305

60.

Miura

Fujiwara

(2017). Intimate partner violence during pregnancy and postpartum depression in Japan: A cross-sectional study. Frontiers in Public Health, 5, Article 81. doi:10.3389/fpubh.2017.00081

61.

Moehler

Brunner

Wiebel

Reck

Resch

(2006). Maternal depressive symptoms in the postnatal period are associated with long-term impairment of mother-child bonding. Archives of Women’s Mental Health, 9, 273-278. doi:10.1007/s00737-006-0149-5

62.

Morash

Bui

Zhang

Holtfreter

(2007). Risk factors for abusive relationships: A study of Vietnamese American immigrant women. Violence Against Women, 13, 653-675. doi:10.1177/1077801207302044

63.

Murray

(1992). The impact of postnatal depression on infant development. Journal of Child Psychology and Psychiatry, 33, 543-561. doi:10.1111/j.1469-7610.1992.tb00890.x

64.

Nemoto

Rodriguez

Mkandawire-Valhmu

(2008). Battered Japanese women’s perceptions and experiences of beneficial health care. Japan Journal of Nursing Science, 5, 41-49.

65.

O’Higgins

Roberts

I. S. J.

Glover

Taylor

(2013). Mother-child bonding at 1 year; Associations with symptoms of postnatal depression and bonding in the first few weeks. Archives of Women’s Mental Health, 16, 381-389. doi:10.1007/s00737-013-0354-y

66.

Okano

Murata

Masuji

Tamaki

Nomura

Miyaoka

(1996). Validation and reliability of Japanese version of the EPDS. Archives of Psychiatric Diagnostics and Clinical Evaluation, 7, 525-533.

67.

O’Leary

K. D.

(1999). Psychological abuse: A variable deserving critical attention in domestic violence. Violence and Victims, 14, 3-23.

68.

Pico-Alfonso

M. A.

Garcia-Linares

M. I.

Celda-Navarro

Blasco-Ros

Echeburúa

Martinez

(2006). The impact of physical, psychological, and sexual intimate male partner violence on women’s mental health: Depressive symptoms, posttraumatic stress disorder, state anxiety, and suicide. Journal of Women’s Health, 15, 599-611. doi:10.1089/jwh.2006.15.599

69.

Plichta

S. B.

(2004). Intimate partner violence and physical health consequences: Policy and practice implications. Journal of Interpersonal Violence, 19, 1296-1323. doi:10.1177/0886260504269685

70.

Quinlivan

J. A.

Evans

S. F.

(2005). Impact of domestic violence and drug abuse in pregnancy on maternal attachment and infant temperament in teenage mothers in the setting of best clinical practice. Archives of Women’s Mental Health, 8, 191-199. doi:10.1007/s00737-005-0079-7

71.

Raymo

J. M.

Iwasawa

(2005). Marriage market mismatches in Japan: An alternative view of the relationship between women’s education and marriage. American Sociological Review, 70, 810-822. doi:10.1177/000312240507000504

72.

Reichenheim

M. E.

Moraes

C. L.

Lopes

C. S.

Lobato

(2014). The role of intimate partner violence and other health-related social factors on postpartum common mental disorders: A survey-based structural equation modeling analysis. BMC Public Health, 14, Article 427. doi:10.1186/1471-2458-14-427

73.

Samuel

Hayton

Gold

Feeley

Carter

C. S.

Zelkowitz

(2015). Attachment security and recent stressful life events predict oxytocin levels: A pilot study of pregnant women with high levels of cumulative psychosocial adversity. Attachment & Human Development, 17, 272-287. doi:10.1080/14616734.2015.1029951

74.

Seng

J. S.

Sperlich

Low

L. K.

Ronis

D. L.

Muzik

Liberzon

(2013). Childhood abuse history, posttraumatic stress disorder, postpartum mental health, and bonding: A prospective cohort study. Journal of Midwifery & Women’s Health, 58, 57-68. doi:10.1111/j.1542-2011.2012.00237.x

75.

Silverman

J. G.

Decker

M. R.

Reed

Raj

(2006). Intimate partner violence victimization prior to and during pregnancy among women residing in 26 U.S. states: Associations with maternal and neonatal health. American Journal of Obstetrics & Gynecology, 195, 140-148. doi:10.1016/j.ajog.2005.12.052

76.

Straus

M. A.

Hamby

S. L.

Boney-McCoy

Sggarman

D. B.

(1996). The Revised Conflict Tactics Scales (CTS2). Journal of Family Issues, 17, 283-316. doi:10.1177/019251396017003001

77.

Sullivan

Perry

Sloan

Kleinhaus

Burtchen

(2011). Infant bonding and attachment to the caregiver: Insights from basic and clinical science. Clinics in Perinatology, 38, 643-655. doi:10.1016/j.clp.2011.08.011

78.

Takeuchi

Sakagami

Perez

R. C.

(2016). The mother and child health handbook in Japan as a health promotion tool: An overview of its history, contents, use, benefits, and global influence. Global Pediatric Health, 3, 1-9. doi:10.1177/2333794X16649884

79.

Tiwari

Chan

K. L.

Fong

Leung

W. C.

Brownridge

D. A.

Lam

P. C.

(2008). The impact of psychological abuse by an intimate partner on the mental health of pregnant women. BJOG: An International Journal of Obstetrics & Gynaecology, 115, 377-384. doi:10.1111/j.1471-0528.2007.01593.x

80.

Tsai

A. C.

Tomlinson

Comulada

W. S.

Rotheram-Borus

M. J.

(2016). Intimate partner violence and depression symptom severity among South African women during pregnancy and postpartum: Population-based prospective cohort study. PLoS Medicine, 13, Article e1001943. doi:10.1371/journal.pmed.1001943

81.

Umeda

Kawakami

(2014). Cross-cultural measurement equivalence of the Japanese version of Revised Conflict Tactics Scales Short Form among Japanese men and women. Psychiatry and Clinical Neurosciences, 68, 804-811. doi:10.1111/pcn.12194

82.

Weissman

M. M.

Paykel

E. S.

Klerman

G. L.

(1972). The depressed woman as a mother. Social Psychiatry, 7, 98-108. doi:10.1007/BF00583985

83.

Wen

Fan

(2015). Monotonicity of effect sizes: Questioning kappa-squared as mediation effect size measure. Psychological Methods, 20, 193-203. doi:10.1037/met0000029

84.

World Economic Forum. (2018). The global gender gap report 2018. Geneva, Switzerland: Author. doi:10.1177/0192513X04267098

85.

World Health Organization. (1996). Violence against women: WHO consultation, Geneva, 5-7 February 1996. Geneva, Switzerland: Author. Retrieved from http://apps.who.int/iris/handle/10665/63277

86.

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

87.

World Health Organization. (2011). Information sheet Intimate partner violence during pregnancy. Geneva, Switzerland: Author. Retrieved from http://apps.who.int/iris/bitstream/10665/70764/1/WHO_RHR_11.35_eng.pdf

88.

Zhu

O’Campo

Koenig

M. A.

Mock

Campbell

J. C.

(2005). Prevalence of and risk factors for intimate partner violence in China. American Journal of Public Health, 95, 78-85. doi:10.2105/AJPH.2003.023978

89.

Yoshida

Yamashita

Conroy

Marks

Kumar

(2012). A Japanese version of mother-to-infant bonding scale: Factor structure, longitudinal changes and links with maternal mood during the early postnatal period in Japanese mothers. Archives of Women’s Mental Health, 15, 343-352. doi:10.1007/s00737-012-0291-1

90.

Yoshihama

Horrocks

Kamano

(2009). The role of emotional abuse in intimate partner violence and health among women in Yokohama, Japan. American Journal of Public Health, 99, 647-653. doi:10.2105/AJPH.2007.118976

91.

Zeitlin

Dhanjal

Colmsee

(1999). Maternal-foetal bonding: The impact of domestic violence on the bonding process between a mother and child. Archives of Women’s Mental Health, 2, 183-189. doi:10.1007/s007370050047