Interpersonal violence is responsible for more ill-health and premature death in women under the age of 45 years than any other preventable health conditions such as hypertension, obesity and diabetes [1]. When interpersonal violence is experienced during pregnancy, it not only affects the health and well-being of the mother but is also associated with adverse health outcomes for the foetus [2]. This has significant ramifications because negative birth outcomes represent a significant cost to society. For example, low birth weight and pre-term infants contribute disproportionately to neonatal morbidity and health care costs, as well as leading to a multitude of short and long term health problems [3].

The term 'interpersonal violence' refers to "the intentional use of physical force, or power, threatened or actual, against oneself, another person, or against a group or community, that either results in, or has a likelihood of resulting in injury, death, psychological harm, mal-development or deprivation" [4]. This definition includes victimisation perpetrated against intimate partners, parents, siblings, children, other relatives, friends, acquaintances, colleagues and strangers [4–6]. Interpersonal violence during pregnancy may be perpetrated by current or previous intimate partners, family members, and strangers or may occur as a result of fighting. Previous research, however, has indicated that violence during pregnancy is more commonly perpetrated by an intimate partner [7–9].

The effects of violence during pregnancy on both maternal and foetal health have been extensively investigated in the literature. However, conflicting results have been reported, possibly due to limitations in sample sizes, study methodology and operational definitions [10, 11]. Recent studies conducted in the USA used population-based data to examine the adverse effects of violence during pregnancy [12, 13]. In Western Australia (WA), a population-based study reported a rising incidence of interpersonal violence hospitalisations particularly among women of child bearing age, highlighting this serious public health issue [14]. However, there has been minimal research specifically targeting this vulnerable group in Australia.

This population-based, retrospective cohort study utilised the Western Australian Data Linkage System to determine the association between exposure to interpersonal violence during pregnancy and adverse maternal and foetal health outcomes from 2002 to 2008.

A large majority of pregnant women exposed to violence tended to be 25 years of age and under (61.6%), Indigenous (66.5%), multiparous (65.2%), smoked during their pregnancy (59.0%), and were not in a married or de-facto relationship (59.1%). The most common type of violence was inflicted by bodily force (65.8% of cases), followed by rape (15.4%), assault with a blunt/sharp object (11.1%) and other types of assault (11.1%). Of those cases where a code existed describing the relationship between the perpetrator and the victim (72.0%), the majority of pregnant women were assaulted by either their spouse or partner (69.5%), followed by a person where the relationship was not specified (20.8%), or another family member (5.6%) (not shown in the table).

The non-exposed group of women reported a significantly higher percentage of foetal distress compared to women who had been exposed to violence during their pregnancy. Adverse foetal outcomes such as low birth weight and foetal deaths, however, occurred significantly more often among women exposed to violence during their pregnancy, compared to the non-exposed group.

The study has highlighted that pregnant women exposed to violence may sustain poor health outcomes for themselves and their baby. In this whole population study, pregnant women hospitalised for an incident of interpersonal violence were at almost double the risk of experiencing one or more adverse maternal complication than the non-exposed group. The findings provide further evidence of an association between antepartum hemorrhage and exposure to violence [17]. Violence is often directed towards the pregnant women's abdomen and the high prevalence of injury due to blunt force may explain these results [17] Consistent with previous research,[7–9] the majority of violent events were perpetrated by the pregnant women's partner or spouse. The overall risk of one or more adverse foetal complications among pregnant women exposed to violence was also increased 2-fold.

It is important to note the multifactorial relationship between violence and its impact on maternal and foetal outcomes during pregnancy.[9, 12, 13, 18–27] Similar to the literature, [12, 28–30] certain demographic characteristics were found to be associated with the risk of violence, such as younger age, marital status and parity. Population subgroups should be targeted for violence screening and counselling during pregnancy.

While mixed findings have been reported on the association between ethnicity and rates of interpersonal violence in the US and South America,[25, 31, 32] it is evident that significant differences exist between Indigenous and non-Indigenous pregnant women. Pregnant Indigenous women accounted for 67% of all hospitalisations due to violence despite representing approximately 4% of the WA population. This is consistent with previous research which identified that hospital admissions by Indigenous females due to violence were consistently higher than Indigenous males, non-Indigenous males and females [14]. High rates of established behavioural health factors such as being disadvantaged economically, broken family ties, smoking, alcohol and substance abuse increased the risk of experiencing interpersonal violence among Indigenous families,[33, 34] which may explain the higher occurrence of interpersonal violence related hospitalisations among our cohort of Indigenous women. In addition, the risk of experiencing adverse maternal and foetal outcomes for this group was significantly increased. These findings have important implications for the planning of health services and distribution of resources to effectively reduce the burden of interpersonal violence in the Indigenous community.

This study has addressed a number of shortcomings of past research. The use of the WA Data Linkage System assisted in reducing selection bias, minimised loss to follow up and enabled the assessment of the association between pregnant women exposed to violence and adverse maternal and foetal outcomes at the population level. It also had the advantage of detecting small differences by inclusion of a large number of cases. Another strength was the use of high quality, objective data which should be more accurate than information obtained via participant self-report measures [35, 36]. Moreover, the results obtained from this population-based study may be generalisable to the rest of Australia, because WA is considered to be representative of the broader Australian population in terms of key socio-demographic and health economic indicators [37].

Several limitations should be taken into account. Firstly, our findings reflect only the severe cases of interpersonal violence, namely those which led to hospitalisation. Therefore, findings cannot be generalised to pregnant women exposed to less severe violence not requiring hospitalisation. Also, many violent events in Australia are never reported,[38] particularly among Indigenous people and particularly incidents involving domestic violence. Limited access to hospitals in rural and remote areas may also result in underreporting [39]. We also acknowledge that women in the comparison group might have been exposed to violence during pregnancy that did not require hospitalisation or was not reported, consequently underestimating the risk of adverse outcomes associated with violence. Despite this, it is known that pregnant women are more likely to seek medical attention due to concern for their unborn baby [40]. This study investigated maternal and foetal outcomes for women who were victims of an acute episode of violence during pregnancy. Exposure to chronic violence, on the other hand, may affect pregnancy outcomes differently and warrant further studies.

Finally, a large number of factors other than interpersonal violence are known to contribute to poor maternal and foetal outcomes. These include alcohol and drug usage, living conditions, health conditions, nutrition, level of prenatal care and socioeconomic status. Adverse outcomes can occur as a result of interpersonal violence, biological, behavioural and socioeconomic factors [10]. However, information on such potential confounding factors was not captured in the available databases.

In conclusion, pregnant women hospitalised for an incident of interpersonal violence are at significantly increased risk of adverse maternal and foetal outcomes. Since poor maternal and foetal outcomes can have long term health consequences, greater priority needs to be given to the primary prevention of violence against pregnant women and the care given to these women following an incident of violence. In developing a response to violence, the findings indicate that prevention programs should target younger pregnant women and focus on preventing intimate partner violence. It is imperative that culturally appropriate intervention programs are implemented to reduce the high rate of violence against Indigenous pregnant women and also that they receive appropriate care and monitoring after exposure to violence.