Christina C. Pallitto, Ph.D.
Domestic violence is a serious public health problem that affects the health and well-being of millions of women and families throughout the world. While the risk factors vary across cultures, similar consequences have been observed globally, ranging from psychological sequelae to death. Furthermore, domestic violence during or around the time of pregnancy can lead to unique risks for maternal, perinatal and child health. A limited body of literature has documented some of the adverse outcomes associated with domestic violence, however, more research is needed to establish the validity of the relationship in a broader range of settings and to gain a better understanding of the mechanisms through which abuse adversely affects women’s and infants’ health.
The purpose of this document is to summarize the existing research exploring the association between domestic violence and health problems, focusing mainly on maternal and infant health, but also exploring the relationship between domestic violence and reproductive health, mental health, and physical health. First, the global magnitude of the problem of IPV will be discussed. Then, the existing research on the relationship between domestic violence and adverse maternal, perinatal, and infant health outcomes will be critically reviewed. Next, the mechanisms through which violence affects health will be described, and support for these hypothesized mechanisms will be provided from the literature. Finally, some conclusions and suggestions for future research will be outlined.
Magnitude of the problem
Globally, abuse by an intimate partner is the most common form of violence against women (WHO, 2002) with great human and economic costs. High rates of abuse have been found throughout the world, according to a review of prevalence studies from 50 countries (Heise, Ellsberg, Gottemoeller, 1999). Heise and colleagues (1999) found lifetime rates of physical abuse by a partner ranging from 10% to 52%, including results from a study from Chile in which researchers found that 26% of Chilean women experienced abuse in their current relationship and a study from Leon, Nicaragua in which 52% of women experienced abuse in their lifetime. In a population-based survey of reproductive age women currently or formerly in an intimate relationship in Colombia, 41% reported physical abuse by an intimate partner, 11% reported sexual abuse, and 26% reported emotional abuse (Ojeda, Ordoñez, Ochoa, 2000). Abuse during pregnancy poses additional risks to maternal and infant health, and the high prevalence rates found in a compilation of studies shows the need to focus attention on this aspect of violence against women. Prevalence rates of abuse during pregnancy ranged from .9% to 20.1%, with most studies ranging between 3.9% and 8.3% (Gazmararian et al., 1996), depending on the definitions and samples used.
ADVERSE OUTCOMES ASSOCIATED WITH DOMESTIC VIOLENCE
Prior to reviewing the research exploring the connection between domestic abuse and adverse outcomes to maternal, perinatal, infant, and reproductive health, a brief discussion will follow outlining a few methodological issues that are important to consider when critically assessing these studies.
An effort was made to review studies from a variety of settings, but the majority of the studies are from developed countries since they are more likely to appear in peer-reviewed and scientific journals and meet scientific standards. In reviewing studies on the association between abuse, especially pregnancy-related abuse, and maternal, perinatal, infant, and reproductive health, it is important to keep in mind a few definitional and methodological issues. For example, the reader must consider how the researchers define the type of abuse being assessed (emotional, physical, or sexual), what instruments are used to measure abuse, whether these instruments are validated, and whether the time period captures lifetime abuse, abuse during pregnancy, or abuse around the time of pregnancy (Ballard et al., 1998). It is also important to consider at what point in the pregnancy or postpartum period the respondent was interviewed, and whether the perpetrator is a partner, a family member, or a non-family member since different pathways and mechanisms would operate in each case. In addition, the authors should describe how they chose their sample (random or convenience), what type of study design they used, and finally what factors they controlled for in the analysis. Limitations in study design, analytical techniques, and varying definitions can bias the results and limit the conclusions that can be drawn from the studies therefore this review presents the findings of the studies in light of these methodological considerations (See Appendix for tables providing more details of studies reviewed here).
MATERNAL AND INFANT HEALTH
Low birthweight (LBW)
Numerous studies have explored the relationship between domestic violence and low birth weight, yet only a few have found a significant association after adjusting for confounding factors (Bullock & McFarlane, 1989; Nuñez-Rivas et al., 2003; Valladares et al., 2002). In the first study to find this association, the authors interviewed women in two settings in —private and public hospitals in the U.S.—and analyzed these two distinct groups separately (Bullock & McFarlane, 1989). They found that abused women in private hospitals had four times the odds of having a low birthweight infant than non-abused women, after controlling for the effects of race, smoking, alcohol use, and other factors related to gynecological and obstetric history. No significant association was found among the public hospital population. The authors hypothesize that the reason they found greater effects of abuse among the private patient population could be the fact that public hospital patients had more risk factors for low birthweight that were not controlled for in the analysis, therefore, it was easier to isolate the effect of abuse on birthweight in the higher socio-economic population. Despite the strengths of the study design, the authors failed to control for drug use and low maternal weight gain, and the private patients were not randomly selected.
Valladares and colleagues (2002) assessed birthweight and violence exposure among 101 Nicaraguan mothers of low birthweight infants and 202 control mothers with normal birthweight infants. They not only found significant associations after adjusting for a variety of socio-economic factors, parity, and smoking status, but they also found that when stratified by causes of low birthweight (preterm, growth restricted, acute growth restricted, and chronic growth restricted) that the relationship remained significant. Unfortunately, the authors did not adjust for maternal weight gain, which other researchers have found to be a confounding factor in the association between abuse and birthweight (Campbell et al., 1999).
In a smaller study of 118 women in Costa Rica, the researchers found that abused women had a birthweight that was, on average, 449 grams less than non-abused women (Nuñez-Rivas, 2003). The risk of low birthweight was three times greater among abused women as compared to non-abused women, even after controlling for a variety of risk factors, including maternal weight gain, interpregnancy interval, and alcohol and tobacco use. The researchers used a broad definition of violence to include violence by persons other than intimate partners, therefore the prevalence rates of abuse during pregnancy were higher than those found in other studies (29.7%), and the results cannot be directly compared to the results of other studies.
Another study finding a positive association between abuse and low birthweight used a meta-analysis technique in which the authors pooled data from 8 published studies and found that birthweights of infants born to abused women were slightly lower than those of non-abused women (Murphy et al., 2001). The difference, although small, was statistically significant.
A few additional studies found a significant relationship between abuse and low birthweight, but these studies did not adjust for confounding factors (Curry et al., 1998; Dye et al., 1995), and one of the studies referred to history of having a low birthweight infant rather than assessing birthweight in the current pregnancy (Curry et al., 1998). Several other studies found either no significant association (Berenson et al., 1994; Grimstad et al., 1997; O’Campo et al., 1994) or an association that was significant prior to adjusting for confounding factors and which lost significance after controlling for them in the analysis (Campbell et al., 1999; McFarlane et al., 1996; Schei et al., 1991; Webster et al., 1996).
Although Campbell and colleagues (1999) failed to show a significant association in the multivariate analysis, their study did make several contributions to the literature on this topic. First, they demonstrated the importance of considering low maternal weight gain and other maternal health problems in the relationship between intimate partner violence and low birthweight. Also, their decision to conduct separate analyses for full term and preterm infants limited their statistical power to find significant associations because of the reduced sample sizes. The fact that the abuse was found to be a significant risk factor for low birthweight for full term infants but not preterm infants in the unadjusted analysis shows the importance of controlling for gestational age in these types of analysis.
Similarly, the study by McFarlane and colleagues (1996), while failing to show a statistical association between abuse and birthweight after adjusting, did contribute several other findings that help to advance this area of research. They found a modifying effect of smoking on the relationship between abuse and low birthweight, meaning that abuse was a risk factor for low birthweight among women who smoked. They also found that race was a modifier in the relationship between alcohol or drug use and low birthweight, and specifically African-American and Hispanic women who used alcohol or drugs had a higher rate of low birthweight than White women who used alcohol or drugs, suggesting uncontrolled confounding and/or distinct mechanisms among different racial groups.
Intrauterine Growth Restriction
In one of the only published studies exploring the connection between domestic violence and intrauterine growth restriction (IUGR), the researchers found that women abused during pregnancy were significantly more likely to experience intrauterine growth restriction, however, this relationship was no longer significant after adjusting for alcohol, drug, and tobacco use (Jannsen et al., 2003). These findings are based on a large population-based sample of 4750 postpartum women in Canada, but the researchers only questioned women about abuse during pregnancy and found a prevalence rate of abuse of 1.2%, which was well below that found in previous studies. According to some researchers, the risk of abuse may actually decrease during pregnancy (Castro et al., 2003; Saltzman et al., 2003), and by failing to measure abuse experience outside of the gestational period, the authors found lower prevalence rates and had less power to find a statistical association. The mechanism through which abuse could affect fetal growth rates may be affected by behavioral or physiological patterns that began prior to the pregnancy, as will be discussed shortly.
Another study conducted in Texas investigated the effect of abuse on fetal growth, but failed to find a significant association (Berenson et al., 1994). However, the study design was not optimal for finding an association. The researchers recruited study participants from a prenatal clinic that serves mainly poor women who are generally at low-risk for pregnancy complications, since the high-risk patients are referred to another facility, and compared outcomes among 32 abused women and 352 controls. Given the small number of abused women in the study and the low prevalence of complications among this population, the researchers had inadequate statistical power for finding an association.
In their study on the association between abuse during pregnancy and low birthweight, Valladares and colleagues (2002) showed that when results were stratified by types of growth restriction, abused women with all types of intrauterine growth restriction (acute, chronic, and general) had higher odds of low birthweight after adjusting for confounding factors.
A few studies have provided evidence of the connection between abuse and preterm labor (Cokkinides et al., 1999; Schei et al., 1991; Shumway et al., 1999). One of these studies, which admittedly had some methodological limitations, demonstrated a significant association (Shumway et al., 1999). While the preterm labor prevalence in prenatal populations is normally around 6.9% to 10.0%, Shumway and colleagues (1999) found prevalence rates of preterm labor of 15.4% among moderately abused women and 17.2% among severely abused women. The odds of having preterm labor were 4.1 times higher among abused women as compared to non-abused women, but the authors failed to adjust for confounding factors.
Using a large sample of 6143 postpartum women in South Carolina, Cokkinides and colleagues (1999) found that abused women had an 80% increased risk of hospitalization for preterm labor compared to non-abused women, after adjusting for age, SES, prenatal care, and smoking during pregnancy. However, they failed to adjust for maternal weight gain and drug or alcohol use during pregnancy, and they did not find an association between abuse and preterm birth. They also found that abused women were almost twice as likely to have preterm births of normal birthweight than non-abused women, while no association was found between abuse and preterm births of low birthweight.
While the study by Campbell and colleagues (1999) focused on the relationship between abuse and birthweight, their study sheds some light on the effect of abuse on preterm vs. full term births. They conducted a stratified analysis, in which they found that full term infants of physically or emotionally abused women had a significantly higher risk of low birthweight in the unadjusted analysis, while preterm infants of abused women did not have a significantly higher risk of low birthweight, thus showing the importance of considering gestational age when studying risk factors for low birthweight and that different risk factors and pathways might operate among preterm and full term low birthweight outcomes in relation to abuse.
Fetal Or Infant Death
A few studies have explored the effect of abuse on the risk of fetal or infant death, and significant associations were found in a study from Canada (Jannsen et al., 2003) as well as a study from India (Jejeebhoy, 1998). The Canadian study, which was described earlier in this document, found that women abused during pregnancy had more than seven times the risk of having a perinatal death than non-abused women after controlling for SES, race, and substance use, although they failed to control for gynecological and obstetric history (Jannsen et al., 2003).
In the Indian study, researchers analyzed two culturally distinct geographic regions of the country and conducted separate analyses for each of the two outcomes—perinatal death and infant death—as well as a joint analysis of fetal or infant death (Jejeebhoy, 1998). While the effect of physical violence on infant death is significant in both of the regions of the country, even after adjusting for confounding factors, the effect of violence on fetal death is significant only in the more conservative region of the country. The findings are limited by some methodological shortcomings, however, including the decision to measure lifetime abuse rather than abuse during pregnancy. In addition, since the study was not designed to measure these associations, there is not adequate data available to control adequately for factors, such as maternal health and weight gain, which might be confounding the observed relationship.
Other Pregnancy Complications
Researchers analyzing data from South Carolina found an increased risk of hospitalization for several pregnancy-related complications among abused women as compared to non-abused women (Cokkinides et al., 1999). Specifically, abused women were significantly more likely to be hospitalized for kidney infection, preterm labor, or trauma due to a fall or blow to the stomach area after controlling for age poverty, prenatal care, and smoking during pregnancy. However, the researchers did not control for obstetric history or maternal health status, which could affect the likelihood of being hospitalized for the first two conditions, nor did they differentiate between intentional and unintentional injuries to the abdomen. The authors also found that abused women were more likely to have cesarean delivery than non-abused women, although they did not control for previous c-section. They also found a borderline significant association between abuse and fetal distress.
Janssen and colleagues (2003), in their study of Canadian women, found that women abused during pregnancy had 3.51 higher odds of antepartum hemorrhage than non-abused women, after controlling for substance use, SES, and race/ethnicity. In a stratified analysis, they also showed that the relationship between abuse and antepartum hemorrhage was stronger among non-users of alcohol, drugs, and tobacco.
Reproductive And Physical Health
In addition to the research on pregnancy-related outcomes associated with domestic violence, additional research has explored some of the negative reproductive and physical health effects associated with domestic violence.
Gynecological and sexual health
In a recent study, Campbell and colleagues (2002) expanded the literature on this topic by showing that abused women in the general population of women enrolled in health maintenance organizations in multiple cities in United States reported significantly more gynecological, central nervous system, and chronic health problems. Specifically, significantly more abused women reported pain, headaches, loss of appetite, abdominal pain, digestive problems as compared to non-abused women. Likewise, in terms of gynecological problems, abused women were significantly more likely to report STDs, vaginal bleeding, vaginal infection, pelvic pain, painful intercourse, and urinary tract infections (UTIs), and sexually abused women experienced more gynecological problems than physically abused women or never abused women. Although this study was not related to pregnant women, it shows the lifetime patterns of abuse and reproductive and physical health problems.
In another study among the general population, researchers studied the effects of lifetime trauma, including childhood sexual abuse, criminal victimization, and spouse abuse, on women’s gynecological health among 1599 women in the United States (Plichta & Abraham, 1996). Women abused by their spouses were 3.47 times more likely to have had a gynecological problem, including severe menstrual problems, endometriosis, HIV/AIDS, STD, or urinary tract infection, in the five years prior to the survey as compared to non-abused women. Childhood sexual abuse increased a woman’s odds of having a gynecological problem by two times after controlling for confounding factors.
Two additional studies with smaller sample sizes (159 and 111, respectively) found associations between sexual abuse by a partner and gynecological symptoms among women from the United States (Campbell & Soeken, 1999) and physical or sexual abuse by a spouse and sexual problems among a sample of Norwegian women (Schei & Bakkteig, 1989). Since the former study recruited a voluntary sample of women having relationship problems and therefore at higher risk for sexual abuse, these findings while informative are not generalizable to the general population nor the population of abused women. The latter study was limited by its small sample size and the failure to adjust for confounding factors. Another study among US women also found that physically or sexually abused women had higher rates of gynecological problems and sexual dysfunction than matched controls of non-abused women, although the author did not adjust for confounding factors (Chapman, 1989).
Additional studies also showed the increased risk of pelvic pain (Schei, 1990) and pelvic inflammatory disease (Schei, 1991) among physically abused women in Norway. Although the authors adjusted for some confounding factors, the small sample size and sampling strategy limit the generalizability of these findings.
HIV/AIDS and STDs
The risk of HIV/AIDS may also be greater among abused women, according to studies from Africa (Maman et al., 2002; van der Straten et al., 1998) and the United States (Zierler, 1996). One study showed that the odds of partner violence was 10 times higher among young HIV positive women as compared to young HIV negative women, after adjusting for confounding factors (Maman et al., 2002).
Among the studies on gynecological and sexual health already described, a few specifically found a positive association between abuse and STDs (Campbell et al., 2002; Plichta & Abraham, 1996). A few additional studies have also found that physically or sexually abused women were more likely to report STDs than non-abused women (Amaro et al., 1990; Martin et al., 1999a), but these studies report the relationship as correlates. Additional evidence of the association can be found in the study on pelvic inflammatory disease described above, since PID often results from STDs (Schei, 1991). On the other hand, in a study from India by Martin and colleagues (1999b), the authors directly studied the relationship between male STDs and male reports of their own physical and sexual abuse against their partners. The authors found that men who had sexually abused their partners were significantly more likely to have had an STD as compared to non-abusive men, and men who physically abused their partners were more likely to have had an STD, although the relationship was not statistically significant. They were unable to show that the STD was a direct result of the abuse, however. Despite this limitation, these findings are important since they use a large population sample of men. In addition, the authors explored the distinct relationships between physical and sexual abuse on STD risk since the pathways between different abuse types and STDs vary. This issue will be discussed in more detail in a subsequent section on the mechanisms of abuse and adverse outcomes.
Mechanisms Through Which Abuse Leads To Adverse Maternal And Infant Health Outcomes
After reviewing the literature demonstrating evidence of the association between abuse and adverse outcomes, this section will explore the mechanisms through which domestic violence can affect maternal, perinatal, and infant health. These mechanisms include direct trauma, and especially direct trauma to the abdominal area among pregnant women, as well as homicide of a pregnant or non-pregnant woman by her partner or other family member; stress-related behavioral and physiological responses resulting from domestic violence; and male control over female autonomy, sexual decision-making, and fertility control leading directly and indirectly to adverse maternal and infant health consequences.
Trauma Resulting In Injury Or Homicide
Trauma resulting in injury
According to one study, 6-7% of pregnancies were complicated by trauma due to intentional and accidental injuries (Connolly et al., 1997). When a woman is physically abused during pregnancy, injuries or pregnancy complications may directly result from the trauma, especially in cases of trauma to the abdomen, which can lead to rupture of the membranes or damage to the placenta (Connolly et al., 1997; Rogers et al., 1999; Ribe, Teggatz, & Harvey,1993). Some researchers have found that the most common sites of injury among abused women receiving medical care were the face, neck, upper torso, breast, or abdomen (Helton et al., 1987).
Researchers in North Carolina reviewed six years of records of pregnancies complicated by trauma and found that domestic abuse accounted for 22.3% of the traumas and that trauma was associated with uterine contractions in 40% of all patients and with preterm labor in 11%, while placental abruption occurred in 7 patients (less than 2%), five of whom had experienced domestic violence (Connolly et al., 1997). Maternal and fetal deaths are two additional outcomes associated with trauma during pregnancy, according to other researchers, although not all of the trauma in these studies was a result of domestic abuse (Ribe et al., 1993; Rogers et al., 1999).
Trauma resulting in homicide
The most severe consequence of intimate partner violence among pregnant and non-pregnant women is homicide. Researchers in Maryland reviewed 247 maternal deaths over a five-year period and found that homicide was the leading cause of death among this population (Horon & Cheng, 2001). Researchers in Chicago conducted a retrospective analysis of maternal deaths in Chicago and found trauma to be the leading cause of maternal death, accounting for 46.3% of all maternal deaths over a four year period (Fildes, Reed, Jones, Martin, & Barrett, 1992). Although the study included all forms of trauma and was not limited to trauma associated with domestic violence, it established evidence of the mechanisms through which trauma leads to adverse maternal outcomes, in this case, death.
In another study presenting data on homicides and attempted femicides in ten cities in the United States, researchers found that certain abuse patterns were more likely to be associated with homicide or attempted homicide. Specifically, they found that abuse during pregnancy was a risk factor for subsequent homicide or attempted homicide by a partner even after controlling for demographic characteristics (McFarlane, Campbell, Sharps, & Watson, 2002). Their findings show that women who are abused during pregnancies are not only at risk for pregnancy complications and adverse outcomes, but also they face an additional risk in the postpartum period.
The findings related to maternal mortality gathered in these studies are from a developed country, specifically the United States. Developed countries tend to have lower rates of maternal mortality, therefore the risk of murder by an intimate partner is easier to detect and may be more explanatory in these settings than in developing countries, where rates of maternal mortality are greater in general and where causes other than homicide are more common.
Stress-Related Behavioral And Physiological Responses
In the previous section, the mechanism through which direct trauma can lead to injuries and pregnancy complications was discussed, and now the discussion turns to a less direct yet equally compelling mechanism in which the environment of stress and fear that can permeate an abusive household can also lead to adverse outcomes directly through physiological responses triggered by stress or through more indirect mechanisms in which stress provokes certain behaviors that help abused persons cope with the stress. Additionally, stress due to abuse can lead to depression, poor weight gain, or somatic complaints.
One study providing important insight into the mediating effect of stress on the relationship between domestic violence and adverse outcomes was conducted among 808 low income women delivering infants in Baltimore (Altarac & Strobino, 2002). The researchers found that after controlling for possible confounding factors, including smoking, drug and alcohol use, and medical conditions, abused women reporting stress had twice the odds of having a low birthweight infant than women not reporting stress. Interestingly, they did not find an association between abuse alone and birthweight, which implies that the mechanism through which abuse leads to low birthweight among this low income population was stress-related.
Physiological hypothalamic-pituitary response to stress
There are several biologically plausible explanations to explain the pathway through which abuse can lead to adverse pregnancy outcomes. First, stress from abuse can affect hypothalamic-pituitary-adrenal production, which can constrict blood vessels and affect fetal growth and development, as well as affect the motility of smooth muscles of the uterus, initiating premature labor (Austin & Leader, 2000). Several researchers have cited this mechanism to explain findings of stress-induced or abuse-related adverse pregnancy outcomes (Newberger et al., 1992; Newton & Hunt, 1984; Valladares et al., 2002).
Increased alcohol, tobacco, or drug use
Other researchers have focused on the increased use of tobacco, alcohol, and drug use observed among abused women as compared to non-abused women (Amaro et al., 1990; Bullock & McFarlane, 1989; Cokkinides et al., 1999; Grimstad et al., 1997; Janssen et al., 2003; Stewart & Cecutti, 1993). The mediating effect of substance use on the relationship between abuse and adverse pregnancy outcomes, specifically low birthweight, growth restriction, and possibly preterm labor, represents another possible pathway that is supported in the literature but that warrants further investigation.
Several studies have found a significant association between abuse and mental health problems, such as suicidality, post traumatic stress disorder, and depression (Golding, 1999; Ellsberg, 1999; Fischbach & Herbert, 1997; Koss, 1990), and other studies have found that mental health problems can affect antenatal outcomes. For example, researchers in Tennessee found that women who were depressed or suffered from low self-esteem were significantly more likely to deliver a preterm baby after controlling for other factors (Jesse et al., 2003), and researchers in Baltimore found similar risk of preterm birth among depressed African American women (Orr et al., 2002). Another study from North Carolina found women with high levels of anxiety during pregnancy had twice the odds of having a preterm birth, while reports of negative life events and racial discrimination were also significantly associated with preterm birth (Dole et al., 2003).
In addition to the risk of adverse pregnancy outcomes associated with depression in the prenatal period, depression in the postpartum period is also a risk factor for the health of newborns (Murray et al., 2003). Researchers have also found that postpartum depression can actually reduce the duration of breastfeeding, thereby adversely affecting the health of infants (Henderson et al., 2003).
There is also evidence that abused women may experience physiological symptoms due to depression, such as loss of appetite (Campbell et al., 2002; McCauley et al., 1995) or an unhealthy diet (Stewart & Cecutti, 1993), both of which can cause low maternal weight gain (Parker et al., 1994), which in turn can lead to adverse pregnancy outcomes. Campbell and colleagues (2002) also hypothesize that emotional abuse about an abused woman’s weight can lead to low maternal weight gain as well.
Male Control Affecting Sexual And Reproductive Decision-Making
A less explored mechanism through which intimate partner violence can affect maternal and infant health is the environment of fear and threat of abuse by an intimate partner. This can lead to lack of control over health-seeking behavior, sexual decision-making and contraceptive use, which in turn can lead to adverse pregnancy outcomes or risks to women’s sexual and reproductive health.
Lack of medical care and prenatal care
Although they failed to adequately adjust for confounding factors, several researchers have demonstrated the association between domestic violence and delayed entry into prenatal care (Dietz et al., 1997; McFarlane et al., 1992; Taggart & Mattson, 1998). Since lack of adequate prenatal care is associated with adverse maternal and perinatal outcomes (Lazariu-Bauer et al., 2004; Moos, 1989), these findings represent yet another mechanism through which abuse leads to risks in maternal and infant health. This relationship was especially strong among women over 25 years old and women of higher socioeconomic status according to a national sample from the United States (Dietz et al., 1997). Participants from public prenatal clinics in Houston, Texas and Baltimore, Maryland were twice as likely to begin preanatal care during the third trimester, and the authors hypothesize that the power and control dynamic within abusive relationships can at least partially explain the “forced avoidance” (p. 3177) of prenatal care (McFarlane et al., 1992). Similarly, a study from Los Angeles, California and Seattle, Washington showed a 6.5 week delay in prenatal care among abused women as compared to non-abused women (Taggart & Mattson, 1998).
The association between domestic abuse and sexually transmitted diseases or HIV/AIDS, which was briefly discussed in an earlier section of this document, can be explained through either direct or indirect mechanisms. The direct mechanism would be a case of STD or HIV infection resulting from the act of sexual abuse. While these cases are more difficult to document in larger population studies, the biological plausibility of this mechanism, as well as the evidence from the study by Martin and colleagues (1999b) in which the authors demonstrated a strong association between sexual abuse and STDs, suggests this direct pathway accounts for at least some of the increased STD risk observed among abused women.
Additional evidence can be found for the less direct pathway, in which physical abuse leads to increased STD infection, because of women’s lack of power and control over sexual decision making and condom use. Some of the evidence of this mechanism can be observed in the literature on condom use negotiation, in which researchers have demonstrated that violence can accompany women’s attempts at condom use and fear of violence can prevent women from initiating condom use (Eby et al., 1995; Kalichman et al., 1998; Wingood & DiClimente, 1997; Worth, 1989). In addition, the study by Martin and colleagues (1999b) shows a positive, marginally significant, relationship between physical abuse and STDs.
Lack of fertility control
Only a few studies have directly explored the relationship between abuse by an intimate partner and reports of unintended pregnancy and found a significant association (Goodwin et al., 2000; Gazmararian et al., 1995; Pallitto & O’Campo, 2004). Additional studies have indirectly explored the relationship and found that abused women had higher rates of abortion history or unintended pregnancies than non-abused women, but these studies did not control for confounding factors (Amaro et al., 1990; Hillard, 1985; Jacoby et al., 1999; Martin et al., 1999; Stewart & Cecutti, 1993; Yoshihama & Sorenson, 1994). Only one of these studies provided data from a Latin American country and adjusted for confounders, while also providing an in-depth discussion about the mechanism of male control through which abuse could lead to lack of female fertility control and unintended pregnancy (Pallitto & O’Campo, 2004). A related study by the same authors found that not only was one’s individual experience of violence associated with a significantly higher risk of having an unintended pregnancy, but also that living in a community categorized as highly patriarchal increased one’s odds of having an unintended pregnancy by almost four times, after adjusting for individual experience of abuse as well as several other factors (Pallitto & O’Campo, 2005).
As mentioned earlier, violence or fear of violence has been found to be associated with condom use negotiation. In addition, researchers have shown that women’s use of female-controlled contraceptive methods might be limited by fear of discovery (Bawah et al., 1999; Biddlecom & Fapohunda, 1998) or violent accusations of infidelity (Rao, 1997). These findings provide greater evidence of the mechanism through which women’s fertility control is limited by an environment of control present in abusive relationships.
Other researchers have also indirectly provided support for this hypothesized mechanism of male control being associated with violence and lack of fertility control in a variety of international settings by arguing that when women lack autonomy or status, they are either more likely to be abused by their partners (Counts, Brown, Campbell, 1999; Heise et al., 1999; Hindin & Adair, 2002; Koenig, 2003; Schuler, 1996; Smith, 1990) or have limited contraceptive use and lack of fertility control (Govindasamy & Malhotra,1996; Hindin, 2000). It follows that when violence is viewed as a culturally acceptable expression of male control over women, public policies will be lacking or not adequately enforced, and violence and its adverse outcomes will be tolerated as a result.
Recommendations For Future Research
Several recommendations can be made based on the shortcomings of the studies reviewed here and from insights provided by authors of these studies.
1. Type of abuse measured: It is important to consider which mechanism is being hypothesized in a given study since some mechanisms, such as STD or HIV transmission, would require measuring levels of sexual abuse, while others mechanisms would look for an association between physical or emotional abuse and adverse outcomes.
2. Current or lifetime abuse: Researchers must consider whether they should assess abuse by a current partner or lifetime experience of abuse, depending on the outcome of interest and the proposed mechanism to explain it. Clearly, prevalence rates of lifetime abuse would be expected to be higher than prevalence rates of current abuse, which would possibly make it easier to detect a connection between abuse and a given outcome, regardless of whether there is a plausible explanation for finding such a connection.
3. Defining abuse during pregnancy: Varied definitions related to abuse during pregnancy make it difficult to compare findings across studies. Ballard and colleagues (1998) stress the importance of differentiating between ongoing and pregnancy specific patterns of abuse. They suggest grouping abuse into mutually exclusive categories—violence starts during pregnancy, violence continues, or violence ceases during pregnancy (Ballard et al., 1998). Others have argued for expanding these categories to consider post partum abuse as well (Campbell et al., In Press; Harrykissoon, Rickert, & Weimann, 2002).
4. Defining the perpetrator: Researchers should also consider whether abuse by someone other than the intimate partner, such as a family member or non-family would be counted as abuse in their study. Studies strictly focusing on intimate partner violence would not qualify abuse by a parent or sibling as abuse. These differences clearly lead to different prevalence rates and can bias findings. In the case of abuse during pregnancy, it is also important to consider whether or not the abuser was aware of a woman’s pregnancy status in order to understand the motives for abuse.
5. Timing of interview: Researchers have found that when women are asked multiple times during their pregnancy about their experience with abuse, they are more likely to report higher, presumably more accurate, prevalence rates than those women who were only asked once during the first prenatal visit (Campbell et al., In Press). In addition, asking women about abuse only in the postpartum period might bias the results. For example, women who experience adverse pregnancy outcomes might be more likely to report abuse than those who do not experience adverse outcomes as they search for explanations for their adverse outcomes. This type of bias is known as recall bias and is classic in retrospective studies when exposure and outcomes are assessed simultaneously.
6. Instruments used: Several validated instruments have been developed for assessing abuse. Many of the studies described in this document used the Abuse Assessment Screen, which is a four question clinical screen instrument, while other studies used the Conflict Tactics Scale, or other validated instruments. Assessing the validity of the instrument is important for developing credible findings and for comparing across studies that used the same validated instrument. Researchers with the World Health Organization have also developed instruments for assessing violence against women in a variety of settings using standardized instruments in their multi-country study (WHO, 2001).
7. How data is collected: Researchers in the U.S. found that higher rates of abuse were reported to healthcare providers than on self-report questionnaire (McFarlane et al., 1991), and other researchers found that routine screening for abuse was acceptable to most women (Campbell et al., 2002). World Health Organization researchers have developed an elaborate safety and ethical protocol for keeping research participants, interviewers, and researchers safe while conducting research on domestic violence (WHO, 2001).
8. Controlling for confounding factors: In social science and medical research, a confounding factor is a factor that is associated with both the exposure (in this case, domestic violence) and the outcome (in this case, one of the adverse maternal, infant, or reproductive health outcomes considered). When researchers fail to control for confounding factors, they might find spurious associations when they do not exist or they might fail to find an association that does in fact exist. Adequate conceptualization of the risk factors for the outcome of interest must be developed prior to designing the study and analysis plan. Once the data is collected, the analysis should determine whether the factors are in fact associated with both the exposure and outcome, and proper adjustment should be made.
9. Consider moderating variables: In addition to adjusting for confounding factors, other factors may operate as moderators or “effect modifiers” on the relationship between abuse and adverse outcomes. These variables cannot be “controlled” in the same sense that confounding factors are controlled, but they can add important explanatory power to the analysis. Several researchers have found evidence of moderating variables in their analysis of domestic violence and adverse outcomes. For example, socio-economic class (Parker et al., 1994), smoking (McFarlane et al., 1996), prematurity (Bullock & McFarlane, 1989; Campbell et al., 1999), and growth restriction status (Valladares et al., 2002) were all found to modify the effect of abuse on outcomes of birthweight.
This review of the literature has provided an in-depth analysis of the research assessing the relationship between intimate partner violence and maternal, perinatal, infant, and reproductive health in a variety of settings throughout the world. Evidence of an association between abuse and low birthweight, intrauterine growth restriction, perinatal or infant death, and antepartum hemorrhage was found, despite the fact that some of these studies had methodological limitations and that other studies failed to find statistically significant associations. The evidence of association between abuse and other reproductive health consequences was mixed, but several methodologically rigorous studies demonstrated the connection between abuse and gynecological and sexual problems, an increased risk of STD and HIV-infection, delayed prenatal care, reduced condom use, and greater likelihood of having an unintended pregnancy, as well as higher levels of stress, depression, and physical health problems.
This literature review not only reviewed the evidence demonstrating an association between violence and adverse outcomes, but also provided a thorough discussion of the mechanisms through which the connection might operate. Several pathways were suggested, including a direct mechanism in which trauma leads to physical injuries or death among pregnant or non-pregnant women. Alternatively, increased rates of stress could trigger physiological or behavioral responses in response to the stress. These responses could have deleterious effects on the woman herself or on fetal development if she is pregnant. An additional pathway was proposed in which abused women may lack adequate control over their own health-seeking behavior, sexual decision-making, or fertility control, resulting in adverse pregnancy outcomes, STD infection, or unintended pregnancy.
The findings of this review demonstrate the importance of protecting women from abuse, improving screening to identify abused women in prenatal and reproductive health settings, promoting prenatal care among high-risk women, providing appropriate intervention in cases of abuse among pregnant and non-pregnant women, prioritizing women’s reproductive health and maternal and infant health nationally and internationally, and improving women’s status at the societal level.
Altarac M. & Strobino D. 2002. Abuse during pregnancy and stress because of abuse during pregnancy and birthweight. Journal of American Medical Women’s Association 57(4):208-214.
Amaro H, Fried LE, Cabral H, & Zuckerman B. 1990. Violence during pregnancy and substance use. American Journal of Public Health 80(5):575-579.
Austin MP & Leader L. 2000. Maternal stress and obstetric and infant outcomes: epidemiological findings and neuroendocrine mechanisms. Australian New Zealand Journal of Obstetrics and Gynecology 40(3):331-7.
Ballard TJ, Saltzman LE, Gazmararian JA, Spitz AM, Lazorick S, & Marks JS. 1998. Violence during pregnancy: measurement issues. American Journal of Public Health 88(2):274-276.
Bawah AA, Akweongo P, Simmons R, & Phillips JF. 1999. Women’s fears and men’s anxieties: the impact of family planning on gender relations in northern Ghana. Studies in Family Planning 30(1):54-66.
Berenson AB, Wiemann CM, Wilinson GS, Jones WA, Anderson GD. 1994. Perinatal morbidity associated with violence experienced by pregnant women. American Journal of Obstetrics and Gynecology 170:1760-9.
Biddlecom AE & Fapohunda BM. 1998. Covert contraceptive use: prevalence, motivations, and consequences. Studies in Family Planning 29(4):360-372.
Bullock LF & McFarlane J. 1989. The birth-weight/battering connection. American Journal of Nursing 89:1153-1155.
Campbell J, Garcia-Moreno C, & Sharps P. (In Press). Abuse during pregnancy in industrialized and developing countries. Violence and Victims.
Campbell J, Jones AS, Dienemann J et al., 2002. Intimate partner violence and physical health consequences. Archives of Internal Medicine 162:1157-1163.
Campbell JC, Torres S, Ryan J, King C, Campbell DW, Stallings RY, & Fuchs SC. 1999. Physical and nonphysical partner abuse and other risk factors for low birth weight among full term and preterm babies: A multiethnic case-control study. American Journal of Epidemiology 150(7):714-26.
Campbell JC & Soeken KL. 1999. Forced sex and intimate partner violence: effects on women’s risk and women’s health. Violence Against Women 5(9):1017-1035.
Chapman JD. A longitudinal study of sexuality and gynecological health in abused women. Journal of Osteopathic Medicine 89(5):619-24.
Cokkinides VE, Coker AL, Sanderson M, Addy C, Bethea L. 1999. Physical violence during pregnancy: maternal complications and birth outcomes. Obstetrics and Gynecology 93:661-6.
Connolly A, Katz VL, Bash KL, MacMahon MJ, & Hansen WF. 1997. Trauma and pregnancy. American Journal of Perinatology 14(6):331-6.
Counts D., Brown, J, & Campbell, J (Eds.). 1999. To have and to hit: Anthropological perspectives on wife beating. Urbana: Univ. of IL Press.
Curry MA, Perrin N, Wall E. 1998. Effects of abuse on maternal complications and birth weight in adult and adolescent women. Obstetrics and Gynecology 92(4 Part 1):530-4.
Dietz PM, Gazmararian JA, Goodwin MM et al. 1997. Delayed entry into prenatal care: effect of physical violence. Obstetrics and Gynecology 90(2):221-4.
Dole N, Savitz DA, Hertz-Picciotto I, Siega-Riz AM et al. 2003. Maternal stress and preterm birth. American Journal of Epidemiology 157(1):14-24.
Dye TD, Tolliver NJ, Lee RV, Kenney CJ. 1995. Violence, pregnancy and birth outcome in Appalachia. Paediatric and Perinatal Epidemiology 9(1):35-47.
Eby KK, Campbell JC, & Sullivan CM. 1995. Health effects of experiences of sexual violence for women with abusive partners. Health Care for Women International 16:563-576.
Ellsberg M. 1999. Domestic violence and emotional distress among Nicaraguan women: results from a population-based study. American Psychologist 54(1):30-36.
Fildes J, Reed L, Jones N, Martin M, & Barrett J. 1992. Trauma: the leading cause of maternal death. Journal of Trauma 32(5):643-5.
Fischbach RL & Herbert B. 1997. Domestic violence and mental health: correlates and conundrums within and across cultures. Social Science and Medicine 45(8):1161-1176.
Gazmararian JA, Lazorick S, Spitz AM, Ballard TJ, Saltzman LE, & Marks JS. 1996. Prevalence of violence against pregnant women. JAMA 275(24):1915-20.
Golding JM. 1999. Intimate partner violence as a risk factor for mental disorders: a meta-analysis. Journal of Family Violence 14(2):99-132.
Govindasamy P & Malhotra A. 1996. Women’s position and family planning in Egypt. Studies in Family Planning 27(6):328-340.
Grimstad H, Schei B, Backe B, Jacobsen G. 1997. Physical abuse and low birthweight: a case-control study. British Journal of Obstetrics and Gynecology 104:1281-1287.
Harrykissoon SD, Rickert VI, & Wiemann CM. 2002. Prevalence and patterns of intimate partner violence among adolescent mothers during the postpartum period. Archives of Pediatrics and Adolescent Medicine 156(4):325-30.
Heise L, Ellsberg M, & Gottemoeller M. 1999. Ending violence against women. Population Reports Series L, No. 11. Baltimore, Johns Hopkins University School of Public Health, Population Information Program.
Helton AS, McFarlane J, & Anderson ET. 1987. Battered and Pregnant: a prevalence study. American Journal of Public Health 77(10):1337-1339.
Henderson JJ, Evans SF, Straton JAY, Priest SR, Hagan R. 2003. Impact of postnatal depression on breastfeeding duration. Birth 30(3):175-180.
Hindin MJ. 2000. Women’s autonomy, women’s status and fertility-related behavior in Zimbabwe. Population Research and Policy Review 19:255-282.
Hindin MJ & Adair LS. 2002. Who’s at risk? Factors associated with intimate partner violence in the Phillipines. Social Science and Medicine 55:1385-1399.
Horon IL & Cheng D. 2001. Enhanced surveillance for pregnancy-associated mortality—Maryland, 1993-1998. JAMA 285(11):1455-1459.
Janssen PA, Holt VL, Sugg NK, Emanuel I, Critchlow CM, & Henderson AD. 2003. Intimate partner violence and adverse pregnancy outcomes: a population-based study. American Journal of Obstetrics and Gynecology 188(5):1341-7.
Jejeebhoy SJ. 1998. Associations between wife-beating and fetal and infant death: impressions from a survey in rural India. Studies in Family Planning 29(3):300-308.
Jesse DE, Seaver W, Wallace DC. 2003. Maternal psychosocial risks predict preterm birth in a group of women from Appalachia. Midwifery 19:191-202.
Kalichman SC, Williams EA, Cherry C, Belcher L, Nachimson D. 1998. Sexual coercion, domestic violence, and negotiating condom use among low-income African American women.Journal of Women’s Health 7(3):371-8.
Koenig M, Ahmed S, Hossain MB, & Mozumber ABM. 2003. Women’s status and domestic violence in rural Bangladesh: individual- and community-level effects. Demography 40(2):269-288.
Koss MP. 1990. The women’s mental health research agenda: Violence against women. American Psychology 45(3):374-380.
Lazariu-Bauer V, Stratton H, Pruzek R, Woelfel ML. 2004. A comparative analysis of effects of early versus late prenatal WIC participation on birth weight: NYS, 1995. Maternal and Child Health Journal 8(2):77-86.
Maman S, Mbwambo JK, Hogan NM et al. 2003. HIV-positive women report more lifetime partner violence: findings from a voluntary counseling and testing clinic in Dar es Salaam, Tanzania. American Journal of Public Health 92:1331-7.
Martin SL, Matza LS, Kupper LL, Thomas JC, Daly M, & Cloutier S. 1999a. Domestic violence and sexually transmitted diseases: the experience of prenatal care patients. Public Health Reports 114(3):262-268.
Martin SL, Kilgallen B, Tsui AO, Maitra K, Singh KK, & Kupper LL. 1999b. Sexual behaviors and reproductive health outcomes: associations with wife abuse in India. JAMA 282(20):1967-1972.
McCauley J, Dern DE, Kolodner K et al. 1995. The “Battering síndrome”: prevalence and clinical characteristics of domestic violence in primary care internal medicine practices. Annals of Internal Medicine 123:737-746.
McFarlane J, Campbell JC, Sharps P, & Watson K. 2002. Abuse during pregnancy and femicide: urgent implications for women’s health. Obstetrics and Gynecology 100(1):27-36.
McFarlane J, Parker B, Soeken K. 1996. Abuse during pregnancy: associations with maternal health and infant birth weight. Nursing Research 45(1):37-42.
McFarlane J, Parker B, Soeken K, & Bullock L. 1992. Assessing for abuse during pregnancy: Severity and frequency of injuries and associated entry into prenatal care. JAMA 267(23):3176-8.
McFarlane, J, Christoffel, K, Bateman, L, Miller, V,& Bullock, L. 1991. Assessing for abuse: Self-report versus nurse interview. Public Health Nursing 8: 245-250.
Moos MK. 1989. Preconceptional health promotion: a health education opportunity for all women. Women Health 15(3):55-68.
Murphy CC, Schei B, Myhr TL, & Du Mont J. 2001. Abuse a risk factor for low birth weight? A systematic review and meta-analysis. Canadian Medical Association Journal 164(11):1567-72.
Murray L, Woolgar M, Murray J, Cooper P. 2003. Self-exclusion from health care in women at high risk for postpartum depression. Journal of Public Health Medicine 25(2):131-7.
Newberger EH, Barkan SE, Lieberman ES. 1992. Abuse of pregnant women and adverse birth outcome. JAMA 267(17):2370-2.
Newton RW & Hunt LP. 1984. Psychosocial stress in pregnancy and its relation to low birth weight. British Medical Journal 288:1191-1194.
Nuñez-Rivas H, Monge-Rojas R, Grios-Davila C, Elizondo-Ureña AM, Rojas-Chavarria A. 2003. La violencia física, psicológica, emocional y sexual durante el embarazo: riesgo reproductivo predictor de bajo peso al nacer en Costa Rica. Pan American Journal of Public Health 14(2):75-83.
O’Campo P, Gielen AC, Faden RR, Kass N. 1994. Verbal abuse and physical violence among a cohort of low-income pregnant women. Women’s Health Issues 4(1):29-37.
Ojeda G, Ordoñez M, & Ochoa LH. 2000. Salud Sexual y Reproductiva en Colombia: Encuesta nacional de demografía y salud. Resultados. Profamilia. Calverton, Maryland: Macro International, Inc.
Orr ST, James SA, Prince CB. 2002. Maternal prenatal depressive symptoms and spontaneous preterm births among African-American women in Baltimore. American Journal of Epidemiology 156(9):797-802.
Pallitto CC & O’Campo P. 2004. The relationship between intimate partner violence and unintended pregnancy: analysis of a national sample from Colombia. International Family Planning Perspectives 30(4):165-173.
Pallitto CC & O’Campo P. 2005. Community level effects of gender inequality on intimate partner violence and unintended pregnancy in Colombia: testing the feminist perspective. Social Science and Medicine 60(10):2205-2216.
Parker B, McFarlane J, & Soeken K. 1994. Abuse during pregnancy: effects on maternal complications and birth weight in adult and teenage women. Obstetrics and Gynecology 84(3):323-8.
Plichta SB & Abraham C. 1996. Violence and gynaecological health in women < 50 years old. American Journal of Obstetrics and Gynaecology 174:903-907.
Rao V. 1997. Wife beating in rural south India: a qualitative and econometric analysis. Social Science and Medicine 44(8):1169-1180.
Ribe JK, Teggatz JR, & Harvey CM. 1993. Blows to the maternal abdomen causing fetal demise: report of three cases and a review of the literature. Journal of Forensic Science 38(5):1092-6.
Rogers FB, Rozycki GS, Osler TM, Shackford SR, Jalbert J, Kirton O et al., 1999. A multi-institutional study of factors associated with fetal death in injured pregnant patients. Archives of Surgery 134(11):1274-7.
Schei B. 1991. Physically abusive spouse—a risk factor of pelvic inflammatory disease? Scandanavian Journal of Primary Health Care 9(1):41-5.
Schei B, Samuelsen SO, & Bakketeig LS. 1991. Does spousal physical abuse affect the outcome of pregnancy? Scandanavian Journal of Social Medicine 19(1):26-31.
Schei B. 1990. Psycho-social factors in pelvic pain. A controlled study of women living in physically abusive relationships. Acta Obstet Gynecol Scand 69(1):67-71.
Schei B & Bakketeig LS. 1989. Gynaecological impact of sexual and physical abuse by spouse. A study of a random sample of Norwegian women. British Journal of Obstetrics and Gynaecology 96:1379-1383.
Schuler SR, Hashemi SM, Riley AP, & Akhter S. 1996. Credit programs, patriarchy and men’s violence against women in rural Bangladesh. Social Science and Medicine 43(12):1729-1742.
Shumway J, O’Campo P, Gielen A, Witter FR, Khouzami AN, Blakemore KJ. 1999, Preterm labor, placental abruption, and premature rupture of membranes in relation to maternal violence or verbal abuse. Journal of Maternal and Fetal Medicine 8:76-80.
Smith MD. 1990. Patriarchal ideology and wife beating: a test of a feminist hypothesis. Violence and Victims 5(4):257-273.
Stewart DE & Cecutti A. 1993. Physical abuse in pregnancy. Canadian Medical Association Journal 149(9):1257-1263.
Taggart L, Mattson S. 1996. Delay in prenatal care as a result of battering in pregnancy: cross cultural implications. Health Care Women International 17:25-34.
Valladares E, Ellsberg M, Pena R, Hogberg U, & Persson LA. 2002. Obstetrics and Gynecology 100(4):700-5.
16. Van der Straten A, King R, Grinstead O, Vitinghoff E, Serufilira A, Allen S. 1998. Sexual coercion, physical violence and HIV infection among women in steady relationships in Kigali, Rwanda. AIDS Behavior 2:61–73.
Webster J, Chandler J, & Battistutta D. 1996. Pregnancy outcomes and health care use: effects of abuse. American Journal of Obstetrics and Gynecology 174(2):760-767.
Wingood GM, DiClemente RJ. 1997. The effects of an abusive primary partner on the condom use and sexual negotiation practices of African-American women. American Journal of Public Health 87:1016–1018.
World Health Organization. Putting Women First: ethical and safety recommendations for research on domestic violence against women. World Health Organization, Geneva, Switzerland: 2001.
World Health Organization (WHO). World Report on Violence and Health. Geneva, Switzerland, World Health Organization: 2002.
Worth D. Sexual decision-making and AIDS: why condom promotion among vulnerable women is likely to fail. 1989. Studies in Family Planning 20:297–307.
Zierler S, Witbeck B, Mayer K. 1996. Sexual violence against women living with or at risk for HIV infection. American Journal of Preventive Medicine 12:304–309.