Article Text

Download PDFPDF
Childhood experiences of interparental violence as a risk factor for intimate partner violence: a population-based study from northern Vietnam
  1. N D Vung1,2,
  2. G Krantz2,3
  1. 1
    Department of Demography, Faculty of Public Health, Hanoi Medical University, Hanoi, Vietnam
  2. 2
    Division of Global Health (IHCAR), Department of Public Health Sciences, Karolinska Institutet, 17177 Stockholm, Sweden
  3. 3
    Department of Community Medicine and Public Health, the Sahlgrenska Academy at University of Gothenburg, Gothenburg, Sweden
  1. Correspondence to Dr N D Vung, Department of Demography, Faculty of Public Health, Hanoi Medical University, Hanoi, Vietnam; ndvung755{at}yahoo.com

Abstract

Background: This study investigated the act of witnessing interparental violence as a girl and its association with experience of intimate partner violence later in life. A higher level of acceptance of violence was tested as an explanatory factor, in the light of learnt behaviour and gender socialisation theories.

Methods: This cross-sectional study used structured interviews for data collection. In all, 730 married women, aged 17–60, in rural Vietnam participated, with a response rate of 83%. Bi- and multivariate analyses are presented.

Results: 16% reported witnessing interparental violence as a child. Of these, 40% had experienced physical/sexual violence in their intimate relationship over time and 16% in the past year. The risk of lifetime (OR 2.85, 95% CI 1.88 to 4.34) and past-year physical and sexual violence (OR 2.33, 95% CI 1.31 to 4.1) was significantly higher for those who had witnessed interparental violence during childhood than those with no such experience. Women with experience of interparental violence during childhood reported acceptance of violence within an intimate relationship to a higher extent than those with no such experience.

Conclusion: These findings should be understood within the frame of existing theories on social learning, learnt behaviour and the gender socialisation process. The findings further highlight the importance of the intergenerational effect and the need for preventative strategies focused on improving women’s autonomy, for gender equity and for effective legislation that stops all kinds of violence against women.

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

What is already known on this subject

  • Men’s violence against women in intimate relationships commonly occurs in Vietnam.

  • Men’s violence against women in intimate relationships is a serious threat to women’s and children’s health.

What this study adds

  • This study confirms earlier findings that for women who witnessed interparental violence during childhood there is an apparent risk of also being subjected to intimate partner physical/sexual violence as an adult.

  • Few studies in Vietnam touch on the issue of witnessing interparental violence as a child and its consequences, and there are no estimates of how common this problem is in the population.

  • This study is able to show that there is a more tolerant attitude towards violence among women who had to witness interparental violence during childhood than among women with no such experience of violence.

Policy implications

It is important to emphasise the two-generation effect of violence in the family when meeting and treating women exposed to intimate partner violence. Healthcare staff, women's groups, youth groups and others acting locally should be trained in how to support these families and convey life skills training to prevent violence between partners. Support and shelters for women and children who are victims of violence are also needed at local level. Policy-makers, the media and others able to influence public opinion should pay serious attention to family violence.

Intimate partner violence (IPV) is a common and serious public health problem in all parts of the world. As a consequence, many children witness interparental violence, and two studies from the USA indicate that 14–18% of adults report such experience.1

Within the family of origin, children may witness interparental violence, experience it themselves, or both. Studies of largely high-income countries performed in recent decades found strong associations between having witnessed interparental violence as a child, in most cases the father’s violence towards the mother, and a woman’s risk of intimate partner violence in adulthood, after controlling for important family risk factors.1234 These studies do, however, suffer from some methodological problems owing, on the one hand, to the use of cross-sectional designs and, on the other, to the retrospective nature of data and recall bias as pointed out in a systematic review, and actions commonly suggested are still not evidence based.5

A more severe trauma affecting young girls is being the victim of physical or sexual violence, and several studies report an increased risk of re-victimisation by a male partner in adulthood.6789

Among the several studies performed in recent years on associations between early experience of interparental violence and women’s later exposure to IPV, few studies within the public health field have made an attempt to explain why this relationship exists. It has been discussed as a possible learned behaviour, that is, an acquired tolerance towards violence learned through the gendered socialisation process.210 Owing to existing cross-cultural differences in expected gender roles, such behaviour might be more pronounced in some cultures and less in others. Other theoretical models to explain the transmission of IPV draw on the literature on family relations and basic developmental research.11 Children who witness such violence are affected in various domains, including their physical or biological functioning, behaviour, emotions, cognitive development and social adjustment,12 all of which can have profound and lasting emotional and behavioural effects that may contribute to other difficulties in life.

This study was performed in rural northern Vietnam, where gender relations today are influenced by older patriarchal norms concerning gender roles, referring to “tradition” and “customs”, but coexisting with increased opportunities for women to participate alongside men in the economy and in society at large.1314 Men, however, continue to be seen as the primary breadwinners while rural women are primarily responsible for housework and childcare and are expected to maintain “family harmony and happiness”.15 One study investigated attitudes to gender equality and found that the majority of the participating men and women supported statements reflecting inequitable gender norms that upheld traditional gender roles.16

Physical and psychological IPV tends to be prevalent in rural parts,161718 and is considered a sensitive and private issue19 that women avoid discussing with outsiders.162021 No study has investigated the witnessing of interparental violence as a child and its long-term consequences from a public health perspective, and there are no estimates of how common this problem is in the population.

This study aimed to investigate, in a general population sample in rural Vietnam, whether women who witnessed their father’s use of physical violence towards their mother during their childhood were at increased risk of intimate partner physical or sexual violence in adulthood. An attempt was further made to investigate whether re-victimisation was associated with more liberal attitudes towards the use of violence, this hypothesis being built on the gender socialisation process and learnt behaviour theories.

Methodology

Design and sample

This cross-sectional study was conducted within the framework of a demographic surveillance site, in FilaBavi, Hatay Province, in northern Vietnam. FilaBavi consists of a cohort of approximately 50 000 individuals (69 clusters) selected by means of a stratified cluster-sampling procedure from the 240 000 individuals living in the district.22 A number of households were randomly selected from each cluster, proportional to the total number of households in each cluster. Married or partnered women aged 17–60 years old were eligible for the study. The random selection of the clusters was done in proportion to cluster population size, providing a self-weighted sample.

Face-to-face interviews were used to collect the data. The 39 female interviewers and the six field supervisors engaged in the field site were trained by the principal investigator in how to manage the specific circumstances and ethical implications of studies of violence. A pilot study was performed and the interviewers were encouraged to withdraw from participation if they did not feel comfortable, but no one did. Because the interviewers were engaged in regular data collections at the field site, they were recognised by some of the participants, but the interviewers were not living in the same area and had no other connection with the participants.23

Sample size calculations were based on a hypothesised risk of exposure to physical and/or sexual violence. In order to detect a twofold increase in the risk of physical/sexual violence with 80% probability and a prevalence of this exposure of 20% in the study sample, a sample size of 850 individuals was needed. In all, 884 households containing a married or partnered woman aged 17–60 were approached. Only one woman declined to participate, owing to psychiatric illness. Of these women, 867 were currently married and 16 were in a stable sexual relationship with a man (hereinafter referred to as married women). Of the 883 interviewed women, 730 women (82%) responded that they had witnessed or had not witnessed their father (or the mother’s husband/boyfriend) use violence towards their mother and these were included in the study, whereas the 153 who did not remember or could not answer with accuracy were excluded.

The Women’s Health and Life Experiences Questionnaire, developed by the World Health Organization (WHO) for violence research in different cultures, was used for structured interviews.24 The abuse questions were developed on the basis of other abuse assessment scales (Index of Spouse Abuse and the Conflict Tactics Scales) and have high reliability and construct validity.2526 The questionnaire was validated in pilot interviews after translation into Vietnamese and a review panel considered each item for appropriateness. Selected items that were not considered appropriate in a Vietnamese context were removed.

Only women participated directly in this study, and data relating to husbands/partners were obtained from the participating women.

Independent variables

Sociodemographic and psychosocial factors

Sociodemographic and psychosocial variables were tested as independent risk factors. Age was divided into two groups: 17–29 and 30–60. Educational attainment was grouped into primary schooling (⩽5 years) and secondary and higher education (>5 years). Annual household income was divided into quintiles whereby the two lowest income categories formed the exposure category (⩽US$425), with higher incomes as the reference. The respondents’ occupations were divided into farmers and hired labourers. Husbands’ working specifics were grouped into professionals versus semiskilled and unskilled combined. Husbands with more than one recognised wife/partner were divided into having and not having more than one wife/partner.

Interparental violence

The items relating to childhood experience of witnessing interparental violence were phrased as: “When you were a child, was your mother hit by your father (or her husband or boyfriend)?” The next question was: “As a child, did you see or hear this violence?” The witnessing interparental violence variable was dichotomised into “witnessing”, if the answer was “yes”, and “not witnessing”, if the answer was “no” to both questions. Only male (husband/partner) IPV against the mother was investigated and this will subsequently be referred to as interparental violence. Owing to the gender-specific nature of IPV in the Vietnamese context, no items relating to the mother’s use of violence towards the father were included.

Attitudes to violence

Six questions on the use of violence within an intimate relationship were put to the participants. These were: “In your opinion, does a man have a good reason to hit his wife if: (a) She does not complete her housework to his satisfaction (yes/no), (b) She disobeys him (yes/no), (c) She refuses to have sexual relations with him (yes/no), (d) She asks him whether he has other girlfriends (yes/no), (e) He suspects that she is unfaithful (yes/no) and (f) He finds out that she has been unfaithful (yes/no)”. The answer “yes” was interpreted as accepting violence in certain situations and the answer “no” was considered equivalent to not accepting violence. No distinction between possible levels of acceptance was made.

Dependent variables

Intimate partner violence

Physical and sexual violence exposure over the respondent’s life time and in the past year formed the dependent variables. Physical abuse was assessed by six items: slapping or throwing things; pushing or shoving, hitting; kicking, dragging, or beating; choking or burning; and threatening or using a weapon (gun, knife or other object). Sexual abuse was assessed by three items: having sexual intercourse against the respondent’s will; using physical force to have sexual intercourse; and making the respondent take part in a sexual act that she found unnatural or degrading. Physical and sexual violence were combined owing to low reporting of sexual violence (referred to as physical/sexual violence). Those with only one single experience of violence over a lifetime (n = 88) were categorised as not being exposed in order to strengthen the outcome variable as it should reflect a long-term effect of a childhood experience of violence.

Statistical analysis

SPSS V.10.0 was used for all statistical purposes.27 Crude and adjusted odds ratios (ORs) and 95% confidence intervals (CIs) were estimated.

Logistic regression analyses were conducted and two different outcome variables were used: physical/sexual violence ever or in the past year. The main independent variable was the respondents’ exposure to interparental violence during childhood and the sociodemographic and psychosocial factors that proved statistical significance in the bivariate analysis were tried as possible confounding factors.

Ethical considerations

The women were interviewed in their own home or at a nearby health centre if privacy could not be guaranteed in the home. All women were informed about their right to withdraw at any stage of the project. The WHO has issued guidelines for violence research28 and these were strictly followed. The ethics committees at the University of Gothenburg, Sweden, Hanoi Medical University and the Ministry of Health, Vietnam, gave their approval for this research project.

Results

More men than women had completed secondary school and higher education: 83.7% and 78.2%, respectively. About 85% of the respondents were farmers. The majority of the husbands were unskilled workers (73.5%) and 15.5% (n = 130) of the men had more than one wife (table 1).

Table 1

Distribution of sociodemographic and psychosocial factors in the sample of women, their partners and in the household, N = 730

Prevalence factors

Of the 730 married or partnered women, 16.3% (n = 119) had witnessed interparental physical violence as a child. Of these, 40.3% (n = 48) had experienced physical/sexual violence in their lifetime and 16.0% (n = 19) had experienced it in the past year, whereas for the total population the corresponding figures were considerably lower, that is, 22.6% (n = 165) and 8.9% (n = 65), respectively (table 2).

Table 2

Prevalence of physical and sexual violence in the female sample, N = 730

Associations between witnessing interparental violence and IPV

Women who had witnessed interparental violence as a child were almost three times as likely to have experienced physical/sexual violence when grown-up compared with women with no such experience (OR 2.85, 95% CI 1.88 to 4.34 for lifetime exposure; OR 2.33, 95% CI 1.31 to 4.15 for past-year exposure) (table 3). These associations remained of the same strength for lifetime physical/sexual violence but lost statistical significance for past-year violence when sociodemographic factors (husband’s education, family income and husband with more than one wife) were controlled for (OR 2.50, 95% CI 1.57 to 3.97 for lifetime exposure; OR 1.81, 95% CI 0.95 to 3.43 for past-year exposure) (table 4).

Table 3

Prevalence of violence and associations between sociodemographic and psychosocial factors and forms of violence, N = 730

Table 4

Multivariate association between witnessed interparental violence and lifetime and past-year physical and sexual violence, final models, adjusted for age, N = 730

The association found between witnessing interparental violence as a child and exposure to IPV led us to investigate possible differences with regard to sociodemographic and psychosocial factors between the “witnessing violence” group and those who had not witnessed any violence between their parents. It was found that women who had witnessed interparental violence were to a higher extent poorly educated, as were their husbands, who were also more likely to have multiple wives/official partners. Furthermore, the household income was on average lower than in households where the woman had not been exposed to interparental violence (data not shown).

Childhood history and attitudes to violence

In a final step, an attempt was made to test the hypothesis that women who had witnessed interparental violence as a child might have acquired a higher degree of acceptance of violence and abuse than those with no such experience. A “yes” response to any of the six attitudinal items was considered to imply acceptance. We found that women who had witnessed violence to a greater extent gave a “yes” response to all items but one (suspecting unfaithfulness) (table 5). However, only the items “not completing housework satisfactorily” and “being disobedient” displayed statistically significant differences in opinion between the groups of women with and without a history of interparental violence (p⩽0.011). This is further discussed below.

Table 5

Attitudes to violence among women with and without a history of witnessing interparental violence during childhood, N = 730

Discussion

This study confirms earlier findings of a considerable risk increase for women who witnessed violence between their parents of their own exposure to intimate partner violence later in life. Women with a history of interparental violence as a child seemingly displayed a higher degree of acceptance towards violence, a finding that should be regarded with caution but could possibly be understood in the light of the gender socialisation process and learnt behaviour theories and their interpretation in the Vietnamese context.

Strengths and weaknesses

This study relies on respondents’ self-reports and does not take account of the husbands’ experience of interparental violence. It is restricted to witnessing physical violence between parents although it might be that some of the respondents were themselves also victimised during childhood, which in turn might contribute to less recall bias and the rather strong association found with IPV.

The data were collected through personal interviews in a secure atmosphere and the response rate was extremely high, with just one woman refusing to participate. Retrospective reporting can produce false-positive as well as false-negative memories and this type of error increases as time between recall and witnessing increases. However, owing to cultural bias—that is, reluctance to reveal what is happening within the family or a belief that violence in the family is “normal” and also a wish to describe the childhood in positive and sometimes idealised terms—the retrospective approach is more likely to result in an underestimation.29 The fact that the field workers were recognised by some of the respondents could have caused some women to refrain from revealing their experiences of violence. Conversely, it might have contributed to feelings of trust and confidence and made disclosure rates higher.

The field workers were selected because of their experience in interviewing, and they were supervised by six field supervisors with backgrounds as assistant doctors or nurses; the supervisors checked all forms and re-interviewed 5% of the study population. The data collected were judged to be reliable in terms of representing minimum estimates of the true prevalence and to be applicable to a wider population of women in rural Vietnam.

Repeated violence victimisation was set as the exposure category; in this way, those reporting a single experience of violence were categorised as non-exposed. To capture a long-term effect of any childhood experience, we believe the outcome variable (violence victimisation) should reflect an effect extended over time before it can be justifiably linked to the exposure.

The items which mirror a possible acceptance of violent behaviour are included in the widely used WHO questionnaire on violence in intimate relationships,30 but so far no study has presented these data. Whether such reporting is synonymous with also holding such values cannot be judged from this study and it cannot be ruled out that these values signify cultural correctness rather than own values and beliefs. These items are used in their simplest form in this study and serve only as an indication of something that needs further investigation.

Finally, this is a cross-sectional study in which all data were obtained at one point in time and causality cannot be proved.

Witnessing interparental violence and own victimisation

Studies report diverging results when it comes to an association between a history of witnessing interparental violence as a girl and subsequent exposure to intimate partner violence, even though the majority support the intergenerational effect. In a population-based telephone survey in Washington State, USA, Bensley and colleagues1 found that among women who had witnessed interparental violence there was an almost fourfold increase in the risk of physical IPV. Also, in a well-controlled study that included 1443 women seeking medical care, Coker and colleagues31 found that women with a father who physically or emotionally abused their mother were at significantly increased risk of intimate partner violence, and concluded that there is a multigenerational effect of violence. A cross-sectional household survey from India similarly found an almost twofold increase of risk for physical IPV4 among women with experience of interparental violence and a similar risk was also found in a population sample from California.3

Ernst and colleagues,32 however, found no correlation of adults who witnessed IPV as children with the presence of adult IPV in a patient-focused study performed at an emergency unit with an admittedly small sample size (140 women). A recent study from the USA interestingly found that women who witnessed interparental physical violence during childhood were at an increased risk of using violence towards their partner and of reciprocal IPV.33 Our study, one of the very few from a low-income country, indicates a clear association between childhood experience of interparental violence and own victimisation later in life, which is in line with most other studies in this field.

Attitudes towards violence

Several researchers have suggested that such an association could be explained by a learned behaviour,1034 that is, that women who have witnessed violence between parents may perceive such violence as a normal part of family life, resulting in a higher acceptance of such violence and aggression.35 In this way, boys learn to use violence and girls learn to tolerate it, or at least to tolerate aggressive behaviour.1034

Women with a traumatic childhood experience may feel powerless in an abuse situation and may generalise such perceptions across situations, whereas women who do not experience feelings of powerlessness may be better able to cope with such events36 and have a higher sense of self-efficacy.37

In this study, women who had witnessed interparental violence to a higher extent reported acceptance of the husband’s use of violence. This conclusion rests on the finding that more women in the “witnessing group” agreed with the six attitudinal statements following on from “Does a man have a good reason to hit his wife if…” than those who had not witnessed any interparental violence.

For the two items describing events that occur fairly frequently (not completed housework satisfactorily and disobeying the husband), however, a statistically significant difference between the groups was also found, whereas for the other items this was not the case. Our interpretation is that, for such commonplace life events (not completed housework, disobedience) in which one would not expect any woman to accept violent behaviour, it seems the women with a history of interparental violence did so. However, for items on sensitive relationship matters (refusing to have sex, infidelity matters), the difference between the groups was smaller. This is interpreted as mirroring cultural norms embraced by all women and reflects men’s preferential right of interpretation in such matters,14 rather than acquired acceptance behaviour based on one’s own life experiences.

In support of this finding and interpretation is a study by Luke and colleagues16 from Vietnam, where 80% of the participating women and 63% of the men reported at least one situation in which they considered violence against women to be justified. The situations reported by both women and men referred to wives’ misconduct, including disrespect or talking back to a husband, disobedience and being unfaithful. The authors further found that the majority of men and women in their study continued to hold traditional attitudes to support inequitable gender relations, that is, husbands are to be the major decision-makers and breadwinners in the family and wives hold responsibility for family and domestic harmony and have to respect husbands’ supremacy.

Drawing on the findings of Haj-Yahia,38 who observed that a history of interparental violence had a stronger impact on females than on males with respect to feelings of hopelessness, it could be argued that hopelessness contributes also to a “giving-up” attitude, possibly resulting in a higher acceptance of violence victimisation.

Conclusions

This study was performed in a low-income country and confirms earlier findings from primarily high-income countries of an apparent risk of exposure to violence later in life among women who experienced their father’s use of physical or sexual violence towards their mother. There is further modest support for linking this to a higher level of acceptance of violence among the “witnessing” women than among those with no such childhood experience.

From this study it is not possible to establish the exact mechanisms behind this association but our hypothesis is that it should be understood within the frame of existing theories on social learning, learned behaviour and the gender socialisation process. This is not equivalent to putting the blame for the violence on the victimised women but it is important to describe it as it draws attention to the intergenerational effect and the role of women as repeatedly being the victim. This study emphasises the need for appropriate preventative strategies to improve women’s autonomy and gender equality, and the need for effective legislation that stops all kinds of violence against women.

Acknowledgments

We would like to acknowledge and thank the women who participated in the study and who gave their time to answer our questions and share their life experiences with us. We gratefully acknowledge the Health System Research Programme, Fila Bavi, and the interviewers, field supervisors and other office and field staff in Bavi District, Hatay Province, Vietnam, for their participation in the study. We also would like to thank the Department of Community Medicine and Public Health, Sahlgrentska Academy at the University of Gothenburg, Gothenburg, Sweden, for all the support given during the analysis phase of this study.

REFERENCES

View Abstract

Footnotes

  • Funding This study was supported by the Violence Against Women project and the Health System Research Programme funded by Sida/SAREC, Sweden.

  • Competing interests None.

  • Ethics approval The ethics committees at the University of Gothenburg, Sweden, Hanoi Medical University and the Ministry of Health, Vietnam, gave their approval for this research project.

  • Contributors NDV was responsible for data collection, analyses and writing. GK holds overall responsibility for this survey and played an active role in all parts of this project.