Associations between bride price stress and intimate partner violence amongst pregnant women in Timor-Leste

Although the custom of bride price varies in its detail and implementation across diverse cultures, the core universal element involves the transfer of offerings, goods or funds principally from the groom and his family, to the bride’s family [7]. Bride price obligations occur both at the time of marriage, and in many societies including Timor-Leste, for an indefinite period thereafter [8, 9]. Commonly, increased expectations of payments and offerings occur during ceremonial events including the birth of children, the prospect resulting in the perinatal period being a period of high stress for the couple [9, 10]. It is important to distinguish bride price from the practice of dowry in which the transfer is from the bride’s family to the new couple. It is notable however, that even where the practice involves the reverse exchange of goods and funds, such as in India and Pakistan, the practice is associated with violence against women [11‐13].

The bride price custom is widespread, being practiced in countries throughout Asia, Africa, the Middle East and the Pacific, and extending to women from these regions who are immigrant and refugees resettled or seeking asylum in high income countries. Although in aggregate, the at risk population exceeds two billion women, the exact number affected by stress related to bride price remains to be determined (see Table 1). There are reasons to hypothesize that the bride price custom may put women at risk in relation to their overall gender status and, more specifically, to IPV. Qualitative data offer support for these associations [9, 14‐17]. In a cross sectional study of pregnant women in Timor-Leste, we found an associations of bride price stress with impairment in women’s mental health, and more generally, with increased family and spousal conflict. The design of the study did not allow, however, for a direct examination of a possible relationship between bride price and IPV [8]. Notably, a cross-sectional multi-country study found inconclusive results when examining for associations between the customs of bride price and dowry and risk of IPV. Although the custom of dowry was documented in six sites only four had higher than average levels of IPV and one had lower IPV rates. Bride price was associated with decreased IPV in four sites and increased IPV in two sites [18].

Our study is grounded in both Feminist and Marxist theory. The assumption is that assigning a cost to a woman inevitably ‘commodifies’ her as an object of transaction in all cultures. This financial value placed on women has manifold effects, for example, increasing the risk that daughters will be offered for early marriage, a practice that undermines the potential for a gender equal marital relationship, and which likely increases the risk of violence against the women enacted by the husband or his family [17, 19, 20]. Allocating a price to a woman also legitimizes the right and power of men to continue to treat their female partner as an acquired object. As such, the custom of bride price may maintain and strengthen other known risk factors for IPV, including community attitudes and perceptions that support gender inequality and male entitlement to treat women violently [21].

Timor-Leste has experienced a prolonged history of conflict, the massive disruption to the society impacting adversely on pervasive conditions of poverty and underdevelopment [1]. The country was occupied by Indonesia in 1975, provoking a protracted resistance war which culminated in a humanitarian emergency in 1999. Approximately 600,000 persons died. There was loss of property and livelihoods [22], and the country is ranked as one of the poorest in the world [23].

As in other low-income, post-conflict countries, gender inequality is pervasive in Timor-Leste, with women being disproportionately disadvantaged in all walks of life [2, 20‐25]. Close to half of women interviewed in a national survey in the capital city (Dili) and surrounding district reported exposure to IPV, a figure that is comparable to other low-income, conflict-affected countries [24, 25]. High fertility rates take a major toll on women, the maternal death rate remaining high (215 per 100,000 live births in 2015) [26].

It has been argued that, within a social system characterised by gender inequality, the stress of poverty may trigger male perpetrated IPV [4, 27]. Research also indicates that economic inequality between partners significantly heightens the risk of IPV [27]. It seems plausible therefore, that the risk of IPV may be heightened by pressures to remit the bride price. Hence, it is vital to consider the role of poverty when examining the relationship between bride price and IPV in a low-income society.

In the present analysis, we tested the following hypotheses [1]: That the stress of poverty reported by pregnant women would be associated with IPV [2]; That the stress of bride price would be associated with IPV [3]; That, taking into account sociodemographic, pregnancy-related factors and poverty, the stress of bride price would make an independent and strong contribution to IPV.

The DILI birth cohort study is an ongoing multi-wave inquiry conducted in the geographical district including and surrounding the capital city of Timor-Leste as described previously [28]. The baseline data reported here were obtained between June, 2013 and June, 2014 from pregnant women attending the four largest community health centres, situated in the capital city suburbs and providing antenatal care for the entire district including urban and rural areas. In the latest census, Dili district was home to 21.9% of the Timorese population of 1.1 million persons. Government statistics indicated that 96% of pregnant women in the Dili District attended antenatal clinics [24]. Only a small number of women (estimated to be less than 5%) attend smaller satellite antenatal clinics or private facilities.

The second trimester was selected because most women initiate antenatal clinic attendance during that period of pregnancy. All consecutive women were approached while registering their visits. We excluded women with psychosis, severe intellectual disability or severe medical conditions requiring referral to the national hospital in Dili. Informed consent procedures were undertaken in private. Once consent was obtained, women were asked about their preferences for an interview at the health centre or at a subsequent home visit (4.1% of interviews).

Data from a pilot study (n = 427) indicated that 400 participants were required to generate stable prevalence estimates for each clinic [29].

We conducted interviews with 1672 pregnant women, with a response rate of 96%.

Surveys were checked for omissions or out of range values immediately following interview, and again at the research office prior to being entered in de-identified form by a trained data assistant into a password protected computer. Further systematic checking was undertaken of electronic files. Two records were excluded from the analysis because of missing data (>1 item). Surveys were locked in the research office prior to being transferred to the University of New South Wales in Sydney for secure storage.

We selected measures that were theoretically and contextually relevant, had utility in a resource-poor research setting, and showed good psychometric properties based on earlier qualitative and (longitudinal) psychiatric epidemiological research with women in the Dili district [8, 29].

Bride price is known as ‘lia’ or ‘barlake’ in Timor-Leste [9]. Lia involves an obligation on the husband and his wife, once married, to provide for all traditional obligations including funerals and weddings as well as ongoing support to the female partner’s family as a form of marriage expense. In this study we included lia as an item: “not enough money for traditional obligations (lia).” The item was one of a number in a measure of self-reported daily stressors that had occurred in the last 12 months. The rating options were: no problem; some problem; moderately serious problem and serious problem (rated 1–4).

Poverty was measured by three items that were also included in the daily stressors inventory, relating to the past 12 months. Poverty items included: not enough food; poor shelter; and not enough money for school fees or clothes. Each poverty item was scored on the same scale as bride price stress. The total poverty score reflected the sum of all three item scores (ranging from 3 to 12) and following examination of the distribution of scores, the poverty index was divided into three categories: no problem, some problem, and moderately/serious problem.

IPV was assessed using the World Health Organization Violence Against Women Instrument (VAWI) [25, 30]. The timeframe was 12-month prevalence enacted by the current or most recent partner. We excluded the sexual abuse items because the topic was highly sensitive for women in a traditional and deeply religious society. Based on considerations of the distribution and overlap of the psychological and physical domains of IPV, we assigned women to five hierarchically ordered categories. The categories were identified and tested in this setting and category 5 which includes the combination of severe psychological and physical abuse was the most strongly associated with symptoms of depression and mental distress [31, 32]. The categories are [1]: No abuse (no IPV items endorsed) [2]; Low respect/regard (whether spouse spends free time with the respondent, consults her on household matters; respects her and her wishes; trusts her with money) [3]; Severe psychological abuse (jealous or angry if she talks to other men; accusations of being unfaithful; does not permit meetings with friends; limits contact with family; insists on knowing woman’s whereabouts; humiliates her; threatens harm to her or someone close; lack of affection) [4]; Physical abuse alone (pushing, shaking or throwing items; slapping, twisting arm; punching; kicking/dragging; strangling/burning; threats with knife, gun or other weapon; attacks with a knife, gun or weapon; other forms of physical abuse); and [5] Combination of physical abuse and severe psychological abuse. The typology incorporates knowledge that forms of IPV often co-occur. The typology does not assume that allocation to one category will prevent future transference to another category. Assignment of women to the higher level categories [3‐5] was not dependent on whether or not they also endorsed items from category 1 [7].

Funding was from the National Health and Medical Research Council of Australia.

The Human Research Ethics Committee of the University of New South Wales (UNSW) and the National Health Research Cabinet of Timor-Leste approved the study (HC13049). Women were informed verbally about the study because of low literacy levels. They signed translated consent forms in the presence of a witness.

We applied bivariate and multivariate analyses to examine the relationship between poverty, stress of bride price, and IPV. Descriptive statistics for all variables are presented followed by bivariate analyses examining associations with IPV for each of the explanatory variables including relevant sociodemographic factors, stress of bride price, stress of poverty, and pregnancy-related factors. Preliminary bivariate analysis found a statistically significant association between poverty and bride price. This finding provided the rationale for a multivariate analysis to be undertaken to assess the unique contributions of each of the relevant explanatory variables, taking into consideration the degree of co-linearity between these variables. Hence, multiple logistic regression analysis was used to explore the independent contributions of each potential explanatory variable on IPV, adjusting for other variables. Given that in preliminary analyses significant associations were found between categories 4 and 5 IPV categories, that variable was dichotomised as follows: scores of 1 to 3 = no IPV (score = 0); and 4 or 5 = physical and severe psychological abuse =1. In addition, in preliminary analyses, an association was found between parity (whether first or subsequent births) and whether pregnancies were planned or not, so that only the parity variable was included as a proxy index of both indicators in the multiple logistic regression model.

The results of bivariate analyses are expressed as percentages and chi-square tests. The results of the multiple logistic regression analysis are reported as adjusted odds ratios (OR) with 95% confidence intervals (95% CI), statistical differences being set at p < 0.01 and p < 0.05. The analyses were performed using IBM SPSS version 22.