Intimate partner violence and incidence of common mental disorder

ABSTRACT OBJECTIVE To investigate the association of intimate partner violence against women reported in the last 12 months and seven years with the incidence of common mental disorders. METHODS A prospective cohort study with 390 women from 18 to 49 years, registered in the Family Health Program of the city of Recife, State of Pernambuco; from July 2013 to December 2014. The Self Reporting Questionnaire-20 (SRQ-20) assessed mental health. Intimate partner violence consists of concrete acts of psychological, physical or sexual violence that the partner inflicts on the woman. Poisson regression was used to estimate crude and adjusted relative risks (RR) of the association between common mental disorders and intimate partner violence. RESULTS The incidence of common mental disorders was 44.6% among women who reported intimate partner violence in the last 12 months and 43.4% among those who reported in the past seven years. Mental disorders remained associated with psychological violence (RR = 3.0; 95%CI 1.9–4.7 and RR = 1.8; 95%CI 1.0–3.7 in the last 12 months, and seven years, respectively), even in the absence of physical or sexual violence. When psychological violence were related to physical or sexual violence, the risk of common mental disorders was even higher, both in the last 12 months (RR = 3.1; 95%CI 2.1–4.7) and in the last seven years (RR = 2.5; 95%CI 1.7–3.8). CONCLUSIONS Intimate partner violence is associated with the incidence of common mental disorders in women. The treatment of the consequences of IPV and support for women in seeking protection for themselves for public services is essential.


INTRODUCTION
Intimate partner violence (IPV) causes physical, sexual, or psychological harm, including physical assault, sexual coercion, psychological abuse, and controlling behaviors a . This occurs with more moderate situations in which diverse conflicts or episodes of frustration and anger occasionally erupt into aggression, and a more serious and chronic, progressive pattern 17 .
Acts of violence against women are not isolated events; this pattern of behavior violates the rights of women and girls, limits their participation in society and impairs their health and well-being b . The IPV complexity requires considering the structuring conditions of the Brazilian social fabric, validating the focus on the cultural context of the patriarchal system, which generates power asymmetry in human interactions, and intensifying conditions to worsening human rights violations 4 .
Dantas-Berger and Giffin 3 consider that a social order of patriarchal tradition has long "consented" to a certain pattern of violence against women, assigning man the "active" role in the social and sexual relationships, while restricting the woman to passivity and reproduction. The male power game comes from these beliefs that men have rights and privileges more than women 4 .
IPV originates social and economic costs, harming the whole society 6,15 . Women may suffer isolation, inability to work, loss of wages and lack of participation in regular activities, and limit their ability to care for themselves and their children a .
The World Health Organization states that IPV is the most common type of violence against women, affecting 30% of women. It is a serious public health problem since it can lead to immediate injuries, infections and mental disorder 5,13,b . The health sector has been slow to engage with violence against women.
Worldwide, mental health problems, emotional distress and suicidal behavior are common among women who have suffered IPV 7,c . Common mental disorders (CMD) symptoms are insomnia, fatigue, irritability, forgetfulness, difficulty concentrating and somatic complaints, allied to depression and anxiety. CMDs cause a high social and economic cost due to lost days of work, as well as increasing health services demand 8 .
Using the Self Reporting Questionnaire-20 (SRQ-20) 10 , intimate partner violence, experienced by about 50% of Brazilian women, was associated with the CMD under analysis. The authors found a higher CMD among women who reported violence than among those who did not report violence (49% versus 19.6%, respectively). In addition, CMD increases with the severity of violence, with 30.6% for women who were victims of physical violence alone and 62.9% for those who suffered all forms of violence.
Women who have suffered IPV in the past may be more likely to have current psychological disorders than women who have never experienced IPV. The IPV effects on mental health can be immediate and acute, but also have long-term consequences or even become chronic 16 .
Estimating CMD in the population helps to understand its distribution in different groups in relation to its different characteristics, as well as the risk factors associated with its occurrence. The diagnosis provides relevant information to guide the policies of intervention in mental health, reducing or preventing worsening.
This study aimed to investigate the association of intimate partner violence reported against women in the last 12 months and the last seven years with the incidence of common mental disorders in women. This work is the third stage of the study. We invited all the women from the second stage to participate in this new phase, and they answered the same questions about violence that we did in the other stages, with reference to the last seven years and the last 12 months.

This
We used a structured questionnaire, in face-to-face interviews from July 2013 to December 2014 to data collection. The interviewers had higher education and experience in dealing with "violence against women". There were weekly discussions with the interviewers in relation to the difficulties they found during the interviews. Most interviews occurred at the women's home. The questions related to IPV were based on the questionnaire of the Multi-Country Study on Women's Health and Domestic Violence, of the World Health Organization, validated in Brazil 18 .
The questions characterized physical assaults or the use of objects or weapons to cause injury as physical violence; as psychological violence, threatening behavior, humiliations and insults; and as sexual violence, sexual relations through physical force or threats and imposition of acts considered humiliating. Previous publication has details 10 . If the woman answered "yes" to at least one of the questions to each type of violence, we considered as a positive case.
The partner or ex-partner with whom the women lives or lived was her intimate partner regardless of formal union, including current boyfriends, as long as they had sexual relations.
The SRQ-20 instrument, developed by the World Health Organization, assessed mental health; it detects psychiatric problems in primary healthcare for developing countries 6 , with twenty yes-no questions, four physical symptoms and sixteen on psycho emotional disorders. The SRQ-20 was validated in Pernambuco with 62% sensitivity and 80% specificity 9 . In the data analysis, we assigned one point for each affirmative response and zero for each negative response. The SRQ-20 cut-off score was 7/8 6 and the groups were non-suspected CMD (score ≤ 7) and suspected CMD (score ≥ 8).
The questions related to their demographic and socioeconomic characteristics: age (24-27; ≥ 28); race/ skin color (white; non-white); living with a partner (yes; no); years of education (0-4; ≥ 5); productive insertion (unemployed, others); monthly income (none; < 1 minimum wage d ; ≥ 1 minimum wage). The data from the third Interview were typed in the EpiInfo program version 5.3.2, with double data entry and by different typists. Subsequently, the Validate software checked the typing errors and we cleaned and verified data consistency.
The Stata program version 10.1 for Windows performed statistical analysis. To test differences between proportions, we used the Chi-square test and considered those with p < 0.05 statistically significant. We used the Poisson regression to estimate the risk ratio (RR) and 95%CI of the association between CMD, socioeconomic and demographic characteristics of women, and types of IPV in the last 12 months and the last seven years.

RESULTS
We interviewed 60.9% (644) of the postpartum women who answered the second stage of the cohort. Six women had died. We lack to find 390 women due to change of address and 17 refused to remain in the survey. However, these 390 postpartum women did not show a statistically significant difference in relation to IPV and demographic and socioeconomic variables.
We excluded 254 women out of 644 in the current database who had CMD during pregnancy, resulting in a sample of 390.
In agreement with the socioeconomic and demographic characteristics, most women were over 28 years of age, non-white, living with a partner and having five or more years of education. Most women were also unemployed and received less than 1 minimum wage. The incidence of CMD in the last seven years was higher in women with monthly income less than 1 minimum wage (Table 1).
In the 12 months prior to this study, 20.8% of women were victims of some type of violence. In relation to the last seven years, the victims of some type of violence totaled 26.2%. Among women who reported IPV, the most frequent violence was psychological violence. Currently, 21.3% of all women were affected by CMD ( Table 2). The incidence of common mental disorders was 44.6% among women victims of violence in the last 12 months and 43.4% in those who reported violence in the last seven years ( Table 3).
The risk ratio (RR) was higher in physical and sexual violence with or without psychological. The incidence of CMD showed a strong association with violence in the last 12 months (RR = 3.1; 95%CI 2.1-4.7) and in the last seven years (RR = 2.5; 95%CI 1.7-3.8), maintained even after adjusting RR for possible confounding factors (Table 3).

DISCUSSION
This study estimated the incidence of common mental disorder in women victims of IPV, in which psychological violence was associated with CMD even when violence occurred without physical or sexual violence and after adjusting for other variables. The greatest association occurred in situations of physical and sexual violence with or without psychological violence in the last 12 months and in the last seven years, corroborating the findings of Ludermir et al. 10 , regarding the fact that CMD increases with the severity of violence. The small number of women victims of physical or sexual violence, similar to other studies 10,20 , made it impossible the analysis of these variables in isolation. To our knowledge, this is the first Brazilian cohort study estimating the incidence of CMD in women victims of IPV seven years after pregnancy.
According to our study, psychological violence was more common than physical and sexual violence and in women with low education and living with lower income, corroborating with other studies 6,10,11,15 . In fact, the frequency of IPV we found may reflect the characteristics of the territory of the study. Frequencies of violence also depend on the socioeconomic conditions of the population and on the sociocultural contexts in which the gender hierarchy is more or less legitimized, which contributes to the increase or decrease of violence reports 20 .  Exposure to violence is a common feature of women living in developing countries and is significantly associated with mental health problems 12,14 . This may happen because developing countries are conservative and patriarchal societies that reinforce gender inequality 2 . Women who are victims of IPV are more likely to have CMD symptoms 2 . In addition, a systematic review reported a moderate or strong association between IPV and depression, suggesting that women victims of IPV are about three times more likely to develop depression 19 .
Studies report that the more severe the aggression, the greater the impact on women' s mental health 12,15 . These studies are similar to our results, as we verified that the highest incidence of CMD occurred in women who reported more than one type of violence. This association may be a consequence of a greater frequency of violence against women, as well as of the more serious nature of violence, more common in situations with combined forms of violence 1 .
One of the limitations of this study is the information bias, regarding the underreporting of violence. Because it is a sensitive issue, stigma and shame can induce women to omit violence, leading to underreporting of the cases 10 . In addition, women's difficulties and blockades to face the trauma and to remember this painful experience may interfere with their willingness to speak 20 . Thus, the results may be underestimated, leading to an underestimation of the strength of the association between violence and mental disorders.
The role of health professionals in identifying violence against women is still a matter of debate 20 . There is evidence that IPV is associated with the high frequency of demand for primary healthcare services 19 . Health professionals should be aware of women suffering from CMD as this may indicate that they are victims of IPV, as these two issues are strongly associated with the results in this study. In addition, IPV is an aggravation of compulsory notification of public health in public and private health services throughout the national territory, contemplated by Law 10.778/2003, which establishes compulsory notification of violence against women e .
Sometimes physical violence occurs with psychological violence, so health professionals who are treating victims of intimate partner physical violence should be sensitive to symptoms of potential mental health problems and refer them to appropriate mental health services 12 .
This study reinforces that IPV, often reported by women in Brazil, is associated with CMD. Public policies and strategies aimed at reducing gender-based violence can help prevent and reduce CMD symptoms among women who are victims of violence. It is also essential to address the IPV consequences and to support women in seeking protection in public services.