Introduction

In 2006, the World Health Organization estimated the global lifetime prevalence of partner violence against women at 30% [1]. United States population-based estimates indicate that between 4 and 6 million women are physically or sexually assaulted annually by either a current or former intimate partner [2].

Effects of intimate partner violence (IPV) on women include serious injury and death as well as a wide range of physical and mental health outcomes, including sexually transmitted disease, chronic pelvic and abdominal pain, sexual dysfunction, substance abuse, suicide attempts, depression, and post-traumatic stress disorder [3–5]. Some studies report that abused pregnant women are at increased risk of adverse pregnancy outcomes such as miscarriage, premature birth, and low infant birth weight [2,6].

A plethora of negative health outcomes are associated with IPV against pregnant women. Amidst the many physical and psychological complexities include concern for the future safety of the unborn child and of the safety of the mother post-birth, which may lead a victim to consider elective pregnancy termination. A recent meta-analysis indicated that nearly a quarter of women seeking elective pregnancy termination were at one time victims of intimate partner violence [7]. Elective pregnancy termination among women with IPV was associated with sexual assault, lack of control over contraceptive choices, and coercive decision-making, and women who had not informed their partner about the termination were 3 times more likely to be victims of IPV than women who disclosed to their partner [7]. Studies that have examined factors influencing a woman’s decision to terminate a pregnancy find that relationship problems and concern for the effects on existing children are common themes [8]. Furthermore, this research shows that most women’s decisions involve complex and multifactorial issues, indicating that the co-occurrence of these concerns is common.

The National Survey of Children’s Health conducted in 2012 by the National Center for Health Statistics estimated that 7% of US children witness domestic violence in the home each year [9]. A growing body of research ties adverse child experiences, including witnessing violence in the home, to lifelong adverse health consequences [10]. Children exposed to IPV in the home are at risk for behavioral and emotional problems, poor academic performance, and adverse health outcomes [11–13].

Most studies of IPV among women seeking pregnancy termination have relatively small sample sizes, and none have examined the prevalence of current children in the home. The objective of this study, conducted in a large clinical population of women seeking pregnancy termination, was to examine the prevalence of children living in the home in relation to reporting of intimate partner violence among women seeking elective abortions.

Methods

Results

Of 1,415 pregnancy termination clients seen in the clinic over the study period: 1,193 were eligible, 1,108 were invited to participate, 990 consented, and 986 completed the questionnaire. Participation was high: 82.6% (986/1,193) of women eligible for the study filled out the survey. For this analysis, we examined data from the 957 participants with non-missing responses to physical and sexual abuse items and number of children in the home.

Characteristics of the overall sample and prevalence rates have been reported previously [17]. To summarize, 9.9% reported physical abuse and 2.3% sexual abuse by a current or former partner. The combined prevalence of any physical and/or sexual abuse perpetrated by a current or former partner was 10.8%. The 26% of respondents not in a current intimate relationship reported the highest prevalence of physical and/or sexual IPV (16.0%), which by definition was perpetrated by a former partner.

Of the 957 participants included in this analysis, 603 (63.0%) reported that they lived with children under the age of 18. Of those 603, 491 (81.4%) households had either one or two children, and an additional 112 (18.6%) lived with three or more children. Of the 1,064 children identified, almost half were under the age of 4 (512; 48.1%), with a mean age of 5.8 years (SD: 4.5).

Having children was associated with being older, being more likely to be non-white, and having less educated than women reporting no children (Table 1). Almost three-quarters of women with no children (262; 74.0%) reported either full- or part-time employment as compared to 69.3% (418) of women with children. Not having children was associated with being on public insurance or having no insurance coverage. Child status was not significantly associated with relationship status; however, women with children were significantly more likely to report cohabitating with a partner (233; 70.4%) as compared to women without children (98; 29.6%).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Demographic characteristics of participants seeking induced abortion at a women’s health clinic, by presence of children in the household, November 2007-July 2008.

(N = 957).

https://doi.org/10.1371/journal.pone.0186389.t001

Overall, having children was associated with drinking less alcohol, with just over one-quarter of women with children (27.7% vs. 15.8% for women with no children) reporting no alcohol use (Table 1). Not having children was associated with consuming a greater quantity of drinks per week and significantly higher proportion of binge drinking (i.e., consumption of ≥4 drinks in a two-hour period) per month (27.7% for women with no children vs. 16.9% for women with children).

After adjusting for participant age (Table 2), the prevalence of physical and/or sexual IPV for women with children living at home was more than double that of women with no children in their household (AOR: 2.33; 1.41–3.85). We examined the odds of physical and sexual abuse for women with children living at home compared to those without for a variety of characteristics. The odds of abuse for women ages 18–20 years old with children was 4.69 times higher (95% CI 1.83–12.03) than for women without children; the older age groups did not show a significantly increased odds for abuse by child status. Women with the following characteristics had a significantly higher odds of abuse if they had children living at home than women who did not: white race (AOR 2.18; 1.27–3.74), non-Latina ethnicity (AOR: 2.76; 1.62–4.68), at least a high school diploma (AOR 2.62; 1.32–5.18), unemployed (AOR 3.48; 1.24–9.66), private health insurance (AOR 2.85; 1.14–7.16), have a current partner (3.09; 1.59–6.03); high levels of depressive symptomology (AOR 2.72; 1.13–6.55); had a medical termination (AOR 3.21; 1.44–7.18); light (AOR 3.08; 1.45–6.53) moderate (AOR 4.02; 1.35–11.97); and non-binge drinking (AOR 4.15; 1.73–9.96).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Prevalence rate and odds ratios (crude and age-adjusted) of physical/sexual IPV associated with having children <18 years in the household among participants seeking induced abortion at a women’s health clinic, November 2007-July 2008.

https://doi.org/10.1371/journal.pone.0186389.t002

In addition, we found that women with a current partner had three times the odds of abuse if they had children in the home than if they did not, while no increase was found among women without a current partner (violence perpetrated by a former partner). We previously reported on the high risk to women of physical and sexual abuse by former partners among women in this sample (7.3%) [17]. In this analysis we found that women not living with their current partner had a non-significant but greater increased odds for abuse with children in the home (4.31; 0.53–4.02) than women living with their current partner (2.07; 0.77–5.57).

We explored the possibility of interaction between relationship status (i.e., current partner, former partner, unknown), cohabitation (for women with a current partner only), and risk of abuse by presence of children <18 in the household. We conducted a logistic regression analysis for the subsample of women with a current partner where physical/sexual abuse was the outcome, and controlled for participant age (3 categories) and presence of children in the household (yes/no). Women living with children and in a current relationship but not cohabitating with that person were at nearly 11 times the risk of physical or sexual abuse perpetrated by a former partner (Table 3). This risk estimate was not altered when controlling for length of the current relationship.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Numbers and age-adjusted odds ratios of physical/sexual IPV by partner status and the interaction variables of cohabitation status and children less than 18 years old living in the home, among participants seeking induced abortion at a women’s health clinic, November 2007 –July 2008.

https://doi.org/10.1371/journal.pone.0186389.t003

Discussion

Nearly one of every 14 children reported by participants in this study lived in a home in which the woman reported IPV. Furthermore, homes with IPV were associated with younger children (0 to four years of age) when compared with homes in which no IPV was reported. Given these high rates, healthcare visits for family planning, including contraception and pregnancy termination, may provide an important opportunity to intervene on family violence.

Studies from the US, including both urban and rural populations, as well from countries including Haiti, Nepal, India, Turkey and Spain have found that presence of children in the home is associated with increased prevalence of IPV [12, 18–23]. The highest odds of abuse was among the subgroup of women living with children in the home but not cohabitating with their current partner; such abuse was largely perpetrated by a former partner. This group of women seeking elective pregnancy termination had ten times the odds of abuse compared to similar women who were not living with children in the home. We could find no published studies that examined these characteristics, although previous literature has found a strong relationship between abuse by a former partner and homicide [24] as well as IPV leading to injury [25]. One study found that more than half of women killed during pregnancy or the post-partum period were killed by a former partner [24]. For violent relationships in which the partners share a child, the presence of the child may necessitate ongoing partner contact, which may increase exposure to abuse. Other research has indicated that the risk for IPV occurring with children in the home is particularly high when the male partner is not the biological father of the children [25–27]; however, we did not have information about the type of parental relationship to compare to these findings.

We found that women who opted for a medical termination had an increased prevalence of abuse when children were in the home, but no increase among women who opted for surgical termination. Medical termination is required occur early in the pregnancy, as current evidence shows that it is effective through the 9^th week of pregnancy [28]. This may indicate that victims of IPV who have children in the home are more decisive or proactive in wanting to end their pregnancy, perhaps because they want to hide the pregnancy from their abusive partner. A meta-analysis of women seeking pregnancy termination found that IPV was associated with non-disclosure to the partner, which is consistent with our interpretation [7]. Although we did not collect data to substantiate this finding, it generates the hypothesis that women who have children and who experience IPV are reluctant to bring more children into the violent situation.

Women who drank less, measured both as drinks per week and binge drinking, and who lived with children had an elevated odds for IPV than women not living with children who drank similar amounts. Women who drank more than 14 drinks per week or who reported binge drinking did not have an elevated odds for abuse based on the presence of children. Although alcohol use has been consistently tied to abuse as both a risk factor and a consequence [29], women with children who are victims of abuse may drink less due to their caregiver role and potentially to be better able to protect both their children and themselves from the violence.

Our study did not measure violence reported by the father, violence against the children, or whether or not the children witnessed the violence. Regardless, the violence experienced by the female head of household will likely have deleterious effects on the children. Few studies have examined the influence of violence victimization on the caregiving role, although existing research theorizes that violence victimization by the mother can negatively influence attachment [30] and that mothers experiencing IPV have needs for support that are not currently being met [31]. The increased prevalence of IPV in homes with children, especially young children, suggest this as an important avenue for future research. Similarly, IPV interventions have focused almost exclusively on the victims or perpetrators, and rarely on the family as a whole. Interventions that incorporate family as well as individual priorities are likely to be most successful and are badly needed.

This study has several limitations. This study was cross sectional so directionality cannot be determined. The sample of women included only the high-risk population of women seeking elective termination of their pregnancies and thus may not generalize to a broader population. We did not ask women about their relationship to the child in the home, nor did we ask about the perpetrators’ relationship to the child. Our survey did not ask specifically about children’s exposure to the violence. We report physical and sexual violence only; battering and emotional abuse could have equally deleterious consequences. Data collection focused on women’s experiences as the target of violence; no questions about the context of the abuse or possibility of bi-directional violence were sought. Finally, with nearly 1,000 participants, this is one of the largest surveys of its kind; however, after stratification by selected covariables, the numbers of subjects were sometimes small.

Conclusion

This study found that the patients seeking elective pregnancy termination who have children in the home may be at elevated risk for partner abuse, by either current or former partners. An increasing number of healthcare practices are implementing elements of trauma informed care, which helps identify the impact of trauma on health and healthcare delivery [32]. Awareness of family violence can help practitioners better understand the circumstances of their patients, and the implementation of trauma informed care into healthcare practices can help providers tailor care and ensure that appropriate referrals are made.

Acknowledgments

The authors would like to acknowledge Hind Beydoun for contributions to the survey instrument, Fiona Tubmen-Scovack for facilitating data collection and Sherry Sperlich for support of the field operations. The findings and conclusions in this article are those of the authors and do not necessarily represent the views of Planned Parenthood Federation of America, Inc.