Alcohol-related Domestic Violence: a Household Survey in Brazil

OBJECTIVE: To describe situations of domestic violence committed by perpetrators under the influence of alcohol in the largest Brazilian cities. METHODS: A household survey was carried out in the 108 Brazilian cities with more than 200,000 inhabitants in 2005. A multistage probabilistic self-weighted sample stratified in terms of conglomerate units was performed in three selection stages: census tracts, households, and respondents (population between 12 and 65 years old). The instrument to collect the data was the Substance Abuse and Mental Health Services Administration, with questions on sociodemographics and psychotropic drug abuse.. RESULTS: The survey encompassed 7,939 households. In 33.5% of them there were reports of domestic violence, 17.1% out of which involving intoxicated perpetrators. The most frequently reported types of violence associated with the use of alcohol were: arguments among the people in the household (81,8%), loud arguments not aimed at a specific person (70.9%), threats of assault (39.5%), and breaking households objects (38.7%). The respondents also reported physical assault (27.8%), physical assault with weapon (5.5%), and sexual abuse (3.2%). More than half of perpetrators lived in the household and 88.8% were male. Most of the victims were female (63.9%); 33.9% were wives and 18.2% were children. In terms of recidivism, 14.1% of the cases lasted for a period between 1 and 5 years, and in 14.3% they lasted for over a decade. Most of the victims (86%) and perpetrators (77.9%) did not look for the help of either the health services or the police. CONCLUSIONS: In addition to the considerable number of Brazilian households with a history of violence involving intoxicated abusers, this kind of abuse has many specific characteristics. The low rate for the search for help at the health services/police stations point to the importance of actively identifying cases of domestic violence.


INTRODUCTION
Domestic violence affects the entire world population, thus representing a serious social and public health issue. 10,13,14The World Health Organization (WHO) considers that violence can be prevented and recommends prioritizing research in the field based on Resolution 56.24 issued during the World Health Assembly (WHA) held in 2003. a The association between domestic violence and alcohol intake has been addressed in studies carried out in countries such as the United States, United Kingdom, Australia and Switzerland.Researches carried out in North America show that from 50 to 70% of the cases of domestic violence between couples the husband had been drinking before the assault. 9,14,16,19nother study in the U.S. looked into calls to the police because of domestic violence and found that in 86% of the cases the aggressors had been drinking on the day of the assault. 3 Latin America, in a study involving eight cities (Salvador and Rio de Janeiro, Brazil; Santiago, Chile; Cali, Colombia; San José, Costa Rica; San Salvador, El Salvador; Caracas, Venezuela), it was found that 68% of the aggressors had been drinking before assaulting their wives and partners. 18In another study carried out in five countries (Colombia, Dominican Republic, Haiti, Nicaragua and Peru), women whose partners drunk on a regular basis were from 2.6 to 9.8 times more likely to being victims of assault compared to women whose partners did not drink. 8In Brazil, a household survey showed that aggressors committed assaults under the influence of alcohol in 52% of the households with a history of violence. 20cohol is consumed in almost all the regions of the world and in different cultural contexts.In Brazil, a national household survey with the population in the 12-65 age group estimated a prevalence of 74.6% of alcohol intake and 12.3% of alcohol dependence. 4though there are strong associations between domestic violence and alcohol abuse, there are not many studies addressing the specific features of these associations.
Thus, the objective of this study was to analyze situations of domestic violence that occurred due to the influence of alcohol.

METHODS
The data used in this study were obtained from the In the first stage, census tracts were randomly drawn and represented geographical areas containing approximately 200 to 300 households.The census tract is the smallest study unit for socioeconomic data at the Instituto Brasileiro de Geografia e Estatística (IBGE -the Brazilian Institute of Geography and Statistics).This information was used to determine, in each city, through multivariate statistics techniques, groups of homogenous tracts called strata.The number of census tracts per city varied from one to 60 depending on the city's population size.
In the second stage, a systematic random sample was selected and was made up of 24 households per census tract.In calculating the household sample, businesses shops, hospitals, factories, hostels, hotels, among others, were not considered.In the case of buildings, each apartment was counted as a household.
In the third stage, a respondent was selected to provide information on the situations of domestic violence.The selection was random, through an independent mechanism of the interviewer, according to Kish. 11 The Substance Abuse and Mental Health Services Administration (SAMHSA) questionnaire was applied.This instrument was adapted and validated for the Brazilian context in 1999, a with questions on sociodemographic data and on psychotropic drug use.Concerning domestic violence, a questionnaire on the situations of violence taking place in the household was used addressing: 20 frequency, duration and kinds of violence (verbal, physical and sexual) according to WHO standards.b Respondents were also asked about the state of the aggressor at the time of abuse (under the influence of alcohol), characteristics of aggressors and victims (age, sex and family relationship), and search for help at the health services/police authorities.The socioeconomic class (high, middle and low) was assessed based on a standardized rating scale issued by the Associação Brasileira de Institutos de Pesquisa de Mercado (ABIPEME -Brazilian Association of Market Research Institutes).The socioeconomic ratings were similar to the IBGE 2000 census, which ensures sample representativeness.
The team of interviewers for the 27 Brazilian state capitals received standardized training, aimed at making interview procedures more homogenous, in addition to providing information on applying and filling in the questionnaire.
The study was approved by the Ethics and Research Committee of the Universidade Federal de São Paulo (Registration N. 1895/06).The interviews were carried out in private with the respondent and without the presence of third parties.Respondents were informed about the goals and procedures of the study and were ensured anonymity, and informed of right to stop the interview at any time.

RESULTS
From the 9,528 households drawn, 7,939 were included in the study.The loss rate was 16.7%, of which 9.6% corresponded to refusals and 7.1% were due to safety issues of the survey team (e.g.households in drug trafficking areas).
Table 1 shows the sociodemographic characteristics of the 7,939 respondents.The distribution of interviewees in terms of sex and marital status (married/single) was similar.More than half of the respondents were age 35 and above.In terms of schooling, 28.3% were illiterate or had not finished elementary school.Among the 7,939 households surveyed, in 2,661 (33.5%) some kind of violence was reported, out of which 1,361 (17,1%) households respondents reported the aggressors were under the influence of alcohol at the time of the assault.Among the cases of violence reported, 49.4% took place in the year prior to the survey.
Table 2 shows the various kinds of violence that took place in the households after aggressors had been drinking.The most common kinds of violence were verbal assault (scolding/argument 81.8% and loud argument 70.9%), followed by threats of: breaking objects in the household (38.7%), physical assault (39.5%), and assault with objects in the household (27.9%).Additionally, the survey also considered violent reports of assault involving weapons, sexual abuse and theft of money and objects from the household.All of the respondents reported more than one kind of violence on the same occasion.
Concerning recurrence of violence, among the households with a background in alcohol-related violence, in 14.1%, the violence lasted from one to five years, and in 6.8%, from six to ten years, and in 14.3%, they lasted more than a decade.The characteristics of the 1,306 aggressors are shown in Table 3.The survey found a higher proportion of aggressors who lived in the surveyed household.Most aggressors were male, and belonged to the 31-59 agegroup.Only 10.9% of aggressors looked for some kind of help in cutting down on alcohol intake or quitting.Table 4 shows the characteristics of the 1,173 victims identified during the survey.The rate of women was more than two times the rate of men (63.9% versus 31.0%) and 49.5% were between 31 and 59 years of age; in 18.2% the children were the victims.Only 8.3% of the victims looked for some kind of help of the police and 3.7% resorted to the health services.

DISCUSSION
Due to the complexity of the topic, it is possible that the cases of violence have been under-reported.Nevertheless, there was a high prevalence of households with a history of violence associated to drinking.In more than half of the cases of violence reported, the aggressor was under the influence of alcohol.The rate of violent episodes involving alcohol intake corroborates other studies carried out in Brazil and other countries. 6,8,17,20,21,24e different forms of violence associated to alcohol use found in the present study have also been described in the literature.Studies have showed that episodes of domestic violence involving alcohol use tend to be more severe and increase the likelihood of co-occurrence of other forms of violence. 7,12,19,22,23sed on the pharmacological model, alcohol produces disinhibition and impairs judgment, which in some situations can facilitate or justify more aggressive behaviors. 1,7,8,10,16,24Alterations on the levels of monoaminergic neurotransmitters such as dopamine and serotonine may also be associated with alcohol-related assaults. 2,5wever, among the many assumptions proposed to explain violence, excessive alcohol use is one of the most controversial.There is no consensus on whether this is a causal relationship or alcohol consumption is an excuse for violent behavior. 15It is likely to be a complex relationship involving several other biological, psychological and social factors. 2,5,12,17 regard to violence recurrence, alcohol use by aggressors seems to increase the impact of violence, its consequences to family health and violence perpetuation.The belief that alcohol use makes people engage in aggressive behaviors alleviates the aggressor's feeling of guilt and increases the victim's tolerance, which may facilitate new violence episodes. 15In addition, the chronic pattern of drinking can be a major factor for aggression recurrence, which may be aggravated with alcohol dependence. 9,12,22,24 for the characteristics of people involved, more than half of aggressors lived in the same household.It was found a higher proportion of male aggressors and female victims, of which 33.9% were their partners.Many studies on intimate partner violence have reported that most aggressors are males who engage in violent behaviors under the influence of alcohol. 3,7,9,12,24verall studies address the association of alcohol and intimate partner aggression but they do not take into consideration other significant family relationships, such as parent-child.It seems that children witnessing violence, in addition to their suffering, are more likely to reproduce this same violent behavior during their adult life. 2,3,9,24Moreover, alcohol-related violence is often perpetuated throughout childhood and/or adolescence, which are key years for a healthy human development.Further studies and more specific actions targeting alcohol-related violence in the parent-child relationship are needed.
A small proportion of alcoholic aggressors (10.9%) sought help to reduce or stop alcohol use.This finding shows that dependence treatment services can help preventing violence, as suggested in other studies. 15,19,23,24Yet again it is evident the need for preventive actions to address the overlapping issues of domestic violence and substance dependence.
In addition to the considerable prevalence of violence situations involving aggressors under the influence of alcohol in Brazilian cities, the present study showed many peculiarities common to these violent behaviors.
Families should be given more support, including active search of cases, for example through the Programa Saúde da Família (Family Health Program).And given that victims usually seek help at health services or police stations, providers at these services should be more aware of that as it may be the only opportunity to intervene. 14,16Other aspects of domestic violence associated to alcohol use should be further explored such as the amount of alcohol aggressors consumed before engaging in domestic violence and whether victims have also consumed alcohol.

Table 2 .
Kinds of violence taking place involving aggressors under the influence of alcohol (n= 1,361).Brazil, 2005.

Table 3 .
Characteristics of aggressors under the influence of alcohol and kinds of help aggressors looked for (n= 1,306).

Table 4 .
Characteristics of the victims of aggressors under the influence of alcohol and kind of help victims looked for (n= 1,173).Brazil, 2005.