Introduction

Research findings about intimate-partner violence (IPV) have focused mostly on women as victims of violence. However, recent studies show the importance of violence inflicted by women towards men or between same-sex couples (^{Carmo, Grams, & Magalhães, 2011}; ^{Lövestad & Krantz, 2012}). In a multinational study coordinated by the World Health Organization (WHO), it was found that the percentage of women who had ever had partners and who had suffered IPV at any time ranged from 15% to 71% (^{WHO, 2005}). A study of IPV between men and women in Portugal reported that, from the total number of victims who reported IPV between 2007 and 2009, 11.5% were men (^{Carmo et al., 2011}). In the United States, ^{Coker et al. (2002)} reported that 28.9% of women and 22.9% of men had been victims of IPV throughout their lives. Similarly, in 2011 it was reported that 35.6% of women and 28.5% of men had experienced IPV (^{Black et al., 2011}). A multinational study in Latin America on gender equity that included men and women between 18 and 59 years of age in cities in Brazil, Chile, and Mexico showed that 28% of women in Brazil and 31% in Chile and Mexico reported having suffered physical violence at some time from their partners; in men, these percentages varied slightly more, 17% in Mexico, 24% in Brazil, and 31% in Chile (^{Barker & Aguayo, 2012}).

Epidemiological research in Mexico (^{Instituto Nacional de Salud Pública [INSP], 2008}) indicated that 43.9% of women aged 15 years or older have experienced partner violence at some point in life and 25.6% in the last 12 months. Regarding the associated factors, more violence, especially physical and sexual violence, is reported for those who have a lower level of education and lower economic income (^{Instituto Nacional de Estadística y Geografía [INEGI], 2017}). Other studies report that economically active women, regardless of education or socioeconomic status, are more likely to suffer emotional, physical, and sexual violence (^{Instituto Nacional de las Mujeres [INMUJERES], 2008}). Regarding age, ^{Catalano (2012)} has reported that violence towards men is more often from women aged 18-35.

Alcohol consumption and substance abuse by the perpetrator is an important risk factor and facilitator for IPV. Alcohol is the most researched substance and is reported to cause the greatest amount of damage (^{Brighton, Moxham, & Traynor, 2016}; ^{Leonard & Blane, 1992}; ^{O’Leary & Schumacher, 2003}). Psychological or physical aggression was two to three times higher when the perpetrator had drunk compared to the days he had not drunk (^{Moore, Elkins, McNulty, Kivisto, & Handsel, 2011}). Alcohol abuse has served as an excuse to justify or forgive violence (^{Leonard, 2005}) and tends to mitigate the circumstances of aggression or to reduce sanctions (^{Tryggvesson, 2004}). In addition, the belief that men are more likely to use alcohol to inhibit the expression of aggression and to report alcohol consumption as a coping mechanism of negative emotions is still present. Conversely, in certain social contexts, women are encouraged to internalize their emotions and to drink with moderation (^{Catalano, 2012}; ^{Choo et al., 2014}).

Even though alcohol abuse is the main problem of consumption in Mexico (^{Medina-Mora, Real, Villatoro, & Natera, 2013}), drug use is increasingly visible, and socially, there is greater acceptance towards its consumption (^{Villatoro-Velázquez et al., 2012}). A nationwide survey found that a woman was 15 times as likely to be raped by her partner when he consumed drugs than when he consumed alcohol, which risk was 3.2 times greater than in cases without substance consumption (^{Natera & Juárez, 2010}).

Regarding the context of IPV, the place where women experience the most violence is at home (^{Miranda, Halperin, Limón, & Tuñon, 1998}). Between 55% and 95% of women who have been victims of physical violence acknowledged that they had never sought help from official services (^{WHO, 2005}). In Mexico, almost 80% of women who suffered physical or sexual violence from their current partner did not request support and did not file a complaint (^{INEGI, 2017}).

Although surveys in Mexico and previous studies (^{INEGI, 2017}; ^{INEGI & INMUJERES, 2013}) have revealed epidemiological data on violence against women, there is still no national study that includes men, nor the population of 12 to 65 years living in their homes. Therefore, the objective of this epidemiological study was to estimate the prevalence of IPV and its association with alcohol and drug consumption, in a representative sample of men and women in Mexico, through secondary data analysis.

Method

The data from this study come from the 2011 Encuesta Nacional de Adicciones (ENA, National Survey on Addictions). The description of the methodology of the survey has been published in other studies (^{Villatoro-Velázquez et al., 2012}). Key features are described below. It is a household survey with national representation that includes people aged 12 to 65 who live in rural and urban areas. The design of the sample was probabilistic, stratified with cluster selection in various sampling stages. The response rate at the national level was 73.3%.

For the purposes of this secondary data analysis, the population that had a partner at the time of the interview was included (n = 11,096). There were no other eligibility criteria in this section.

Statistical analysis

The overall prevalence of violence at any time and in the last year was calculated by sex, with 95% confidence intervals (95% CI). Subsequently, 95% CIs were calculated to analyze the sociodemographic characteristics of the population that reported IPV by sex. Likewise, for the bivariate analysis, the chi-squared test was applied to find the differences by sex and the characteristics of the situation in which each reported incident of violence occurred. Kaplan–Meier curves and Cox regressions were generated to calculate the probabilities of the onset of violence, alcohol consumption, and drug use. Finally, to obtain the factors related to violence with the current partner in the last year, logistic regression models were performed, stratified by sex, yielding odds ratios (ORs) with 95% CIs. The analyses were done in ^{STATA Version 13 (2013)} and were adjusted by the design of the survey using the command “SVY.” Normalized weighting was used to perform the analyses (n = 11,096), and the population was weighted to draw the frequency tables (N = 54,370,267) (Figure 1). For more specific data of the sample size calculation, it is suggested to consult the ENA 2011 report (^{Villatoro-Velázquez et al., 2012}).

Figure 1 Flow diagram of sample selection process. 

Results

Factors associated with IPV

Regarding the factors associated with IPV in the last year (Table 4), it was found that men aged 18 to 34 years (OR = 1.4; 95% CI [1.0, 1.9]), those currently or ever married (OR = 2.2; 95% CI [1.3, 3.8]), and those living in urban communities (OR = 1.5; 95% CI [1.1, 2.1]) were more likely to suffer violence from their partner. The men also reported that if one of the members of the couple consumed alcohol, their risk of suffering violence was 4.6 times higher than when alcohol was not consumed, and this increased to 10.7 times if both consumed alcohol. If one of the two partners used drugs, the risk of men experiencing violence increased 1.5 times, and if both consumed, this risk increased 6.8 times.

Table 4  Factors Associated with Intimate-Partner Violence in Men and Women. Mexico 2011  

	Violence in the last year
	Men			Women
	OR	95% CI		OR	95% CI
Age
35 to 65	1.0	-		1.0	-
18 to 34	1.4a	1.0-1.9		1.1	.9-1.4
12 to 17	1.5	.7-3.2		2.0a	1.1-3.7
Marital status
Single	1.0	-		1.0	-
Married or ever married	2.2a	1.3-3.8		4.8a	2.8-8.2
School level
Bachelor’s degree or more	1.0	-		1.0	-
Highschool or equivalent	1.2	.7-1.8		1.3	.9-2.0
Middle school	1.2	.7-2.0		1.3	.9-2.1
None or Elementary	1.3	.8-2.1		1.3	.9-2.0
Labor situation
Does not work	1.0	-		1.0	-
Is dedicated to the home	3.8	.9-15.6		.8	.5-1.3
Works	1.2	.8-1.9		1.0	.6-1.6
Family monthly income
More than 6 minimum wages	1.0	-		1.0	-
From 3 to 6 minimum wages	.9	.5-1.6		1.1	.6-1.9
Up to 2 minimum wages	1.1	.6-2.3		1.9a	1.1-3.4
Type of population where he/ she lives
Rural	1.0	-		1.0	-
Urban	1.5a	1.1-2.1		1.1	.8-1.5
Couple’s alcohol consumption
None consumes	1.0	-		1.0	-
1 of the 2 consumes	5.6a	2.5-12.9		3.1a	2.0-4.7
Both consume	11.7a	4.9-27.7		4.9a	3.2-7.5
Couple’s drug use
None consumes	1.0	-		1.0	-
1 of the 2 consumes	2.5a	1.8-3.5		2.7a	1.7-4.2
Both consume	7.8a	1.7-36.9		6.9a	2.5-19.3

^aLevel of significance < .05.

In the case of women, the ones who were more likely to suffer violence by their partner, were those between 12 and 17 years (OR = 2.0; 95% CI [1.1, 3.7]), those who were married or ever married (OR = 4.8; 95% CI [2.8, 8.2]), and those with the lowest salaries (OR = 1.9; 95% CI [1.1, 3.4]). In women, alcohol consumption by their partner increased the risk 2.1-fold, and alcohol consumption by both increased it 3.9-fold. The risk of experiencing violence also increased when one or both partners used drugs (ORs = 2.7 and 6.9, respectively).

Finally, the results of the probability analysis of the initiation of IPV showed that the lowest age of onset of IPV was 11 years for men and women, although it had a faster growth in women than men with increasing age (OR = 1.567, p < .001). On the other hand, the lowest age of onset for drug use by sex was reported at six years in women and at eight years in men, and the probability that drug use had a faster growth in men than women with increasing age (OR = 4.852, p < .001). The age of onset of alcohol consumption was six years for both; however, men had a faster increase in the likelihood of starting and maintaining alcohol consumption with increasing age (OR = 2.121, p < .001). The growth rate in the likelihood of initiating alcohol consumption was the fastest, followed by being a victim of IPV and drug use (Figure 2).

Figure 2 Probabilities associated with the onset of intimate partner violence, onset of drug use and onset of alcohol consumption by sex. 

Discussion and conclusion

This is one of the first epidemiological studies in Mexico to record violence not only towards women but also towards men living with their partners. It is a national panorama between both sexes, in both rural and urban areas, a first approach was made at ENA 2008 (^{Natera & Juárez, 2010}). Violence against women continues to be significantly more common, although in these data from 2011, a lower percentage was obtained, 26.9%, than in a similar survey in 2008 (^{Natera & Juárez, 2010}) that found that 46% of the women had experienced violence from their partner at some point in life. National surveys have mainly evaluated IPV against women, which has seen a reduction from 39.2% in 2006 to 27.3% in 2011 (^{INEGI & INMUJERES, 2013}). These figures encompass the estimate that, on average, approximately 35% of women have suffered violence (physical or sexual) by their partner or a third party in Mexico (^{WHO, 2017}). In the present study, one in five men had been the victim of IPV, in contrast with one in four women.

We do not know why IPV decreased from 2008 to 2011, althougt the scale used was the same and the methods were similar. One answer could be that there has been more information in the media that violence is a problem to be denounced, and the health sector is promoting mechanisms to prevent violence against women and encouraging the creation of health care centers, though these are still not enough. The General Law for Access to Women to a Life Free of Violence has also been promoted (^{Diario Oficial de la Federación, 2007}), which defines domestic violence against women as a crime. It is also possible that social movements, such as feminist movements, have had an influence in empowering women to defend themselves and protect themselves from violence, and in turn, it is possible that men are responding to these changes.

However, it seems that these social changes have not had enough impact on the youngest. Strikingly, we found a significant prevalence of violence in young people aged 12 to 16 years. This change in landscape, intimate relationships, or sexual life is increasingly found at earlier ages (^{Saldívar, 2018}), which makes it a group at risk of following patterns of generalized violence in adult life. It is important for prevention to emphasize the promotion of conflict resolution without violence and the effects that substance use can have on violent behavior among young people.

When comparing the curves of the onset of violence, drug consumption, or alcohol consumption, we observed that the latter began at younger ages, was more prevalent in men than in women, grew with age more rapidly than drugs. This indicates that alcohol consumption is one of the highest-risk substances at early ages. IPV grew less rapidly, the age of initiation is at 11 years of age and, above all, it was mostly directed at women.

We know (^{Esquivel-Santoveña & Dixon, 2012}; ^{Stalans & Ritchie, 2008}) that alcohol and drugs are factors that are constantly present in IPV towards women and men. This study showed the same. We found that when both drugs and alcohol were used, violence increased homogenously for both sexes.

While it cannot be concluded that the consumption of alcohol and drugs causes IPV, associations are important and are sufficient reasons to include the prevention of drug and alcohol abuse in interventions to prevent IPV. On the other hand, the perceived risk of the influence of the consumption of substances on violent behaviors differs between men and women: Men perceive a higher risk of violence when there is alcohol consumption by either one or both members of the couple, and women perceive a lower risk of violence if both have consumed, but not if only one consumes them. Here, women’s perception is that if both of them are intoxicated with drugs, depending on the type of drug, violence will not necessarily be generated. The percentage of drug use in general was so low that it did not allow a distinction by type of drug for the partner-violence variable (^{Natera & Hernández, 2013}), although marijuana was the most consumed.

IPV varied with age. The fact that males aged 18-34 report being the most abused by women leads them to think that because they are young, they are living in a seemingly more open environment of gender equality, even where men accept and report being abused by women, leaving aside cultural aspects of “machismo.”

The causes of violence against women have been studied (^{Muñoz & Echeburúa, 2016}). However, the causes or motives of violence by women towards men have not been studied; it has been presupposed that this occurs as a reactive response of women to men’s violence or is bilateral. The exact reasons will be important to study.

Regarding the differences between urban and rural populations, we observed that in the urban population, the risk of violence by men tended to be greater; however, more alcohol is consumed in rural populations, which explains much of the IPV, while the consumption of drugs was considerably lower. Women consumed less in rural areas, and the consumption of drugs or alcohol in women in Mexico is perceived as a greater transgression in rural areas than in urban areas (^{Natera, 2017}).

Despite the controversies that may exist about gender-based violence in that internationally, women have almost always been reported as victims, there is no denying that, increasingly, men are the victims of social violence, especially perpetrated by their partners (^{Carmo et al., 2011}). This leads us to consider the problem of IPV not only as a matter of gender but also as a social and cultural problem that needs to be addressed from this perspective.

In Mexico, the cultural mandate that women should tolerate men’s consumption of alcohol, including drunkenness, is associated with the machismo part of men’s identity and the accepting or tolerating part of women’s identity (^{Natera & Holmila, 1990}), which have resulted in the justification and legitimization of violence linked to alcohol consumption by men and emphasized submission and tolerance by women, leading the problem of violence to be kept hidden when there is substance use and discouraging the woman from asking for help (^{Natera & Hernández, 2013}).

As observed in this study, a low percentage (10.7% men and 15.8% women) requested help, and the help was mostly sought from the health care sector, where a woman can obtain legal support if she wishes to report. If the injuries are serious and endanger the life of the woman, the attending physician is obligated to report the perpetrator. Although male power is ceasing to be globally hegemonic (^{WHO, 2005}), in our country, there is still great inequality; men have the right to correct female behavior and to make themselves feel socially superior in the home, in private places, and in public environments such as in restaurants and bars, where violence can occur. However, in this study, it is in the public context where women exerted greater violence towards men, which could be interpreted as a response of self-affirmation of women to an act of masculine (macho) violence and because they perceive being more protected in public spheres. However, the home continues to be the main arena for settling disputes.

It is clear that more explanatory factors of violence towards both men and women must be found, such as the influence of social determinants, not only the consumption of substances. Prevention should pay attention to the dynamics of the beginning of the consumption of substances at increasingly younger ages, especially alcohol in the youngest.

This study has an epidemiological basis: the problem of IPV and the use of substances at the national level. Future studies should inform us more precisely about the act of violence, amount of consumption per occasion, the type of drug, the reasons for gender-based violence at different ages and the type of relationship, as well as the reasons why victims did or did not seek help or report the violence. In addition, they should investigate the impact or results of the actions carried out by both civil and governmental organizations to reduce IPV at the national level and its association with substance abuse. Both quantitative and qualitative studies are necessary to achieve a better understanding of the subject.

Programs that promote equality to reduce violence are necessary, though maybe never enough, and should be based on a framework of respect and human rights. This is a long process that requires learning to navigate it. Women are learning to set limits, and it is necessary for both men and women to learn to express conflict and seek solutions without violence.