The European Journal of Public Health Advance Access originally published online on February 13, 2006
The European Journal of Public Health 2006 16(3):278-284; doi:10.1093/eurpub/ckl016
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Injuries |
Alcohol and intimate partner violence: do we have enough information to act?
D. Gil-González1, C. Vives-Cases1,2, C. Álvarez-Dardet1,2 and J. Latour-Pérez1
1 Preventive Medicine and Public Health Area. Department of Community Nursing, Preventive Medicine and Public Health and History of Science,, University of Alicante, Spain
2 Network for Research on Gender and Health (RISG), Spain
Correspondence: Diana Gil-González, Área de Medicina Preventiva y Salud Pública, Departamento de Enfermería Comunitaria, Medicina Preventiva y Salud Pública e Historia de la Ciencia. Edificio de Ciencias Sociales, Campus Sant Vicent del Raspeig, Ap. 99. 03080 Universidad de Alicante, Alicante, Spain, tel: +34965909679, fax: +34965903964, e-mail: Diana.Gil{at}ua.es
Received July 3, 2005, accepted December 8, 2005
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
Background: Male alcohol consumption is one of the accepted risk factors for intimate partner violence. The aim of this study is to assess the magnitude of the association between male alcohol consumption and intimate partner violence against women and the quality of the evidence of published papers exploring this relationship empirically. Methods: Systematic review and meta-analysis of quantitative studies (1966–2004). Eight databases from Social and Behavioural Sciences, Clinical Medicine, and Life Sciences were reviewed. Studies with available 2 x 2 table or odds ratio were analysed using meta-analytic techniques. Results: A total of 22 studies fulfilled the inclusion criteria for the systematic review: 14 (63.6%) were cross-sectional studies, 6 (27.3%) case–series, 2 (9.1%) case–control studies. Ten studies analysed the relationship between alcohol and violence as their primary hypothesis and only two used a direct measure of alcohol consumption. Of them, 11 papers were included in the meta-analysis. The overall pooled odds ratio was 4.57 (95% confidence limits 3.30–6.35), but a high degree of heterogeneity was observed. The magnitude of the effect was inversely associated with the year of publication. The biggest odds ratios were obtained in the studies with the smallest sample sizes. Conclusions: The evidence about the relationship between alcohol consumption and intimate partner violence is of low quality in the study designs and maybe biased by publication of positive results. Currently there is not enough empirical evidence to support preventive policies based on male alcohol consumption as a risk factor in the particular case of intimate partner violence.
Keywords: alcohol, intimate partner violence, risk factor
Male alcohol consumption is one of the accepted risk factors for intimate partner violence.1,2 The scientific literature often supports this association, not only considering that alcohol use in males plays an important role in the occurrence of intimate partner violence,3,4 but also promoting rehabilitation programmes addressed to ameliorate this problem.5–7 Male alcohol consumption has also been integrated in the most accepted causal models as a causal risk factor.8–10 In addition, some countries take into account alcohol rehabilitation in their policies against gender violence.11,12
Despite its general acceptance, the causal relationship between male alcohol consumption and intimate partner violence remains controversial.13–15 Systematic review and meta-analysis could be of some value to assess the available empirical evidence about this issue and to shed some light on its controversial nature. An evidence-based approach could provide valuable information for the development of preventive policies based on the relationship between male alcohol consumption and intimate partner violence.16–19
In the present study, a systematic review and meta-analysis were performed to assess the magnitude of the association between male alcohol consumption and intimate partner violence against women and the quality of the evidence of those published papers which explore this relationship empirically.
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
The fields of Social and Behavioural Sciences, Clinical Medicine, and Life Sciences were searched from 2 April to 5 April, 2004, using the following data bases: ISI Current Contents (1995–March 2004), Medline (1966–March 2004), CINAHL (1982–January 2004), Psycinfo (1966–January 2004), Econlit (1969–February 2004), Francis (1984–November 2003), Sociological Abstracts (1986–March 2004), and Eric (1966–March 2004).
The keywords used were as follows: battered women and alcohol, violence against women and alcohol, domestic violence and alcohol, gender-based violence and alcohol, and gender violence and alcohol. From the identified papers, the quantitative empirical ones analysing the causal relationship between male alcohol consumption and physical intimate partner violence were selected for the systematic review.
Next, the selected studies were codified independently by two authors (C.V. and D.G.), using an ad hoc developed checklist. This checklist was made up of the following variables: degree of alcohol consumption (codified as ‘chronic’, ‘acute intoxication’, ‘not specified’), how alcohol consumption was measured (codified as ‘in batterers blood’, ‘by a third person’, and ‘no specified’), epidemiological design used (cohort, case–control, cross-sectional, ecological, case–series study), and sampling method (intentional, random, stratified, and no sampling); control of confounding variables; and, possible biases within the studies (selection, misclassification, and confounding) following the operative definitions given by Last in his Dictionary of epidemiology (4th edition).20
Since the present study aimed at assessing the magnitude of the association between male alcohol consumption and intimate partner violence, a descriptive statistical analysis was performed with the included papers, using SPSS-11 and Excel-2000 commercial software. Additionally, those studies from the systematic review that calculated crude odds ratio or had available 2 x 2 tables were analysed using meta-analytic techniques with Stats Direct v 4.1.21 In some studies, tables that only provided summary statistics could not be reconstructed. In order to include this information, a summary random effect meta-analysis was performed. On the one hand, the heterogeneity between studies was assessed graphically using the funnel plot (figure 1) of the standard error versus the observed effect. On the other hand, the relationship between the observed odds ratio and the study's year of publication was also examined through the forest plot (figure 2). Lastly, in order to further measure the degree of the possible association between the variables' year of publication and odds ratio of the studies, a Spearman's rank correlation was also calculated.
|
|
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
In this study, 1035 papers were identified for potential inclusion: 336 papers through the ‘battered women and alcohol’ keywords, 283 through ‘violence against women and alcohol’, and finally 416 through the ‘domestic violence and alcohol’ keywords. No papers were found when using the keywords ‘gender-based violence and alcohol’ and ‘gender violence and alcohol’. A total of 1013 of the studies (98%) were excluded, since the majority of them focused on other victims of violence, such as children, older people, and men (n = 580, 57.1%), while others consisted of intervention programme evaluations (n = 113, 11.2%) instead of the problem of violence against women. In addition, theoretical reviews, book reviews, editorials, qualitative studies, and summary of conferences were excluded. Table 1 outlines the reasoning behind this exclusion.
|
Systematic review
Only 22 of the studies published from 1996 to 2003 were included for the systematic review.22–43 All of them were related with physical intimate partner violence. 12 of them (54.5%)23–26,31–37,39 did not analyse the relationship between alcohol and violence as their primary hypothesis. Moreover, only 2 (9.1%) of the 22 studies33,41 made use of a direct measure of male alcohol consumption.
As for the epidemiological design of the papers, 14 (63.6%) of the selected studies consisted of a cross-sectional design,24,25,27,30,32,33,35–37,39,40–43 6 (27.3%) were case–series,23,28,29,31,34,38 and 2 (9.1%) were case–control studies.22,26 No cohort studies were found. Besides, intentional sampling was used in 13 (59%) studies,22,26–28,29–33,36,38,39,41,43 random sampling in 8 (36.6%) studies,23–25,27,34,37,40,42 and stratified sampling in 1 paper35 (4.5%).
As far as possible biases are concerned, all the papers specified clearly their exclusion and inclusion criteria. Also, the authors of the papers stated that confusion variables were controlled in 15 (68.2%) of the studies.22–27,30,35–37,39,40–43 A potential selection bias could not be identified in 19 (86.4%) studies.22–28,30–37,39–41,43 And no evidence of a potential misclassification bias of the outcome was found in 18 (81.8%) of the studies.22,24–32,35–41,43 Nonetheless, a potential misclassification bias in the exposure was observed in 15 (68.2%) of the studies.22–30,32,34,35,38,39,42
Meta-analysis
Table 2 outlines the most important features of the 11 studies included in the meta-analysis.22,24–26,30,33,35–37,39,40 An analysis of the role of male alcohol consumption as a predictive or determinant factor of intimate partner violence was found in all of them, of which two were case–control22,26 studies and nine were cross-sectional24,25,30,33,35–37,39,40 studies.
|
All of the studies which considered the causal association between male alcohol consumption and intimate partner violence showed a statistically significant risk excess. Nonetheless a high degree of heterogeneity regarding their most important features was observed (figure 1).
The relationship between odds ratio and the year of publication of the studies was also analysed. It could be observed that the different odds ratio decreased over time (figure 2). Thus, studies with bigger odds ratio and with broad confidence intervals were published in 1996 and 1998.30,22 In after years, the odds ratios decreased and smaller values could be found. For instance, the most recent study's38 (2003) odds ratio has a value near to the combined odds ratio, and is lower in comparison with the odds ratio of the oldest study analysed30 (1996).
Lastly, a possible relationship between the sample size of the studies and their odds ratio was also assessed. A smaller odds ratio (figure 2) was observed in studies with big samples; while the highest odds ratio were identified in studies with smaller sample sizes and broader confidence intervals.
Spearman's rank correlation
The Spearman's rank correlation between the year of publication and odds ratio obtained a value rs = 0.782 (P < 0.05).44
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Evidence linking male alcohol consumption to violence against women is not strong enough. The lack of designs with enough inferential power (e.g. cohort studies) and the possibility of a publication bias stand in its way. Consequently, decisions on the development of male rehabilitation programmes as a result of intimate partner violence prevention policies are not evidence-based. Therefore, a better understanding about the causes of the problem is crucial for the effectiveness of the policy. New studies are urgently needed to explore the empirical evidence of the relationship between intimate partner violence and other risk factors.16–19
Because grey literature has not been searched in this study and it is widely known that policy-makers make use of this information when designing their action programmes, this fact might be regarded as a limitation on the present study. The different cultural and political contexts of the countries where the intimate partner violence is studied in the papers reviewed, the number of these finally included in the meta-analysis, their heterogeneity, and their low quality seriously damages the credibility of the pooled odds ratio. In this respect, publication biases are probably one of the main sources of heterogeneity. The first two papers published on this subject (1996 and 1998)22,30 highlighted a strong association between alcohol and violence. Then, the large studies suggested weak associations compared with the strong associations proposed by smaller studies. For example, one of the highest odds ratios found (12.94) was obtained by one of the smaller samples (46 people).22
As stated in the introduction to this paper, the treatments and the efficacy of the preventive programmes developed by public institutions are often based on the scientific evidence. For this reason, the possibility of a publication bias suggested in this study could be specially relevant, since a low validity of study findings usually precludes the institutional and financial efforts to ameliorate the problem of intimate partner violence against women.45–47 This finding has also been observed in other meta-analysis studies on other fields of knowledge (e.g. genetics).48,49 It is mentioned that the possible bias and the size of the study reduce drastically the credibility of the associations. Statistical approaches of the studies found in databases are not adequate to analyse the causal influence. Besides, studies which cannot establish strong associations are not often published, hence they are difficult to locate.
The political implications of these findings are crucial for clinicians and policy-makers. The role of substance consumption such as alcohol and its effect upon the occurrence of intimate partner violence are widely accepted in scientific literature.50–52 However, now we know that the evidence in support of this statement is weak. As a result, when treating people with alcohol problems clinicians should not forget that on doing so maybe they are not solving the social problem of intimate partner violence.
In order to determine any real association between male alcohol consumption and intimate partner violence more case–control and cohort studies are necessary. Case–series, though, should be excluded because of their low inferential power. Lastly, current information about this issue should also be carefully re-examined before it is put into practice.
Key points
|
|
Acknowledgments |
|---|
The authors thank the Institute of Women of Spain for its logistic and financial support to this research (Grant I+D+I 106/02) and the Network for Research on Gender and Health (RISG) on its support. The authors also thank Dr Daniel La Parra, from the Sociology II Department, University of Alicante, for his valuable comments on earlier versions of the manuscript. Finally, the authors thank Irma Muñoz, from Education Faculty, University of Alicante, for her valuable support in the English edition on the last version of the paper.
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
1 Collins JJ, Kroutil LA, Roland EJ, et al. Issues in the linkage of alcohol and domestic violence services. Recent Dev Alcohol 1997;13:387–405.[Medline]
2 Murphy CM, Winters J, O'Farrell TJ, et al. Alcohol consumption and intimate partner violence by alcoholic men: comparing violent and non violent conflicts. Psychol Addict Behav 2005;19:35–42.[CrossRef][Web of Science][Medline]
3 Field CA, Caetano R. Ethnic differences in intimate partner violence in the U.S. general population: the role of alcohol and socio-economic status. Trauma Violence Abuse 2004;5:303–17.[Abstract]
4 Smith JW. Addiction medicine and domestic violence. J Subst Abuse Treat 2000;19:329–38.[CrossRef][Web of Science][Medline]
5 O'Farrell TJ, Murphy CM, Stephan SH, et al. Partner violence before and after couples-based alcoholism treatment for male alcoholic patients: the role of treatment involvement and abstinence. J Consult Clin Psychol 2004;72:202–17.[CrossRef][Web of Science][Medline]
6 O'Farrell TJ, Fals-Stewart W, Murphy M, et al. Partner violence before and after individually based alcoholism treatment for male alcoholic patients. J Consult Clin Psychol 2003;71:92–102.[Web of Science][Medline]
7 Stuart GL, Moore TM, Kahler CW, et al. Substance abuse and relationship violence among men court-referred to batterers' intervention programs. Subst Abus 2003;24:107–22.[CrossRef][Medline]
8 Anderson K. Gender, status and domestic violence: An integration of feminist and family violence approaches. J Marriage Fam 1997;59: 665–9.
9 Heise L. Violence against women. An integrated, ecological framework. Violence against women 1998;4:262–90.
10 Fagan JA, Stewart DK, Hansen KV. Violent men or violent husbands: background factors and situational correlates. In Finkelhor D, Gelles RJ, Hotaling GT, editors. The dark side of families. Beverly Hills, CA: Sage, 1983, pp. 49–67.
11 O'Farrell TJ, Van Hutton V, Murphy CM. Domestic violence before and after alcoholism treatment: a two-year longitudinal study. J Stud Alcohol 1999;60:317–21.[Web of Science][Medline]
12 Murphy CM. Treating perpetrators of adult domestic violence. Md Med J 1994;43:877–83.[Medline]
13 Gelles RJ. Alcohol and other drugs are associated with violence: they are not its cause. In Gelles RJ, Loseke DR, editors. Current controversies on family violence Thousand Oaks, CA, Sage Publications, 1990, 203–24.
14 O'Leary KD, Schumacher JA. The association between alcohol use and intimate partner violence: linear effect, threshold effect, or both? Addict Behav 2003;28:1575–85.[CrossRef][Web of Science][Medline]
15 Fals-Stewart W, Leonard KE, Birchler GR. The occurrence of male-to-female intimate partner violence on days of men's drinking: the moderating effects of antisocial personality disorder. J Consult Clin Psychol 2005;73:239–48.[CrossRef][Web of Science][Medline]
16 Schumacher JA, Fals-Stewart W, Leonard KE. Domestic violence treatment referrals for men seeking alcohol treatment. J Subst Abuse Treat 2003;24:279–83.[CrossRef][Web of Science][Medline]
17 Browne A. Violence against women by male partners. Prevalence, outcomes and policy implications. Am Psychol 1993;48:1077–87.[CrossRef][Medline]
18 Blanco P, Ruiz-Jarabo C, García de Vinuesa L, et al. Violence of couple and women's health. Sespas Report 2004. Gac Sanit 2004;18:182–8.
19 Lee WV, Weinstein SP. How far have we come? A critical review of the research on men who batter Recent Dev Alcohol 1997;13:337–56.[Medline]
20 Last JM. A dictionary of epidemiology Fourth edn. Oxford: Oxford University Press, 2001.
21 Stats Direct. Available from: http://www.statsdirect.com (accessed on 15 September 2004).
22 Kyriacou DN, McCabe F, Angglin D, et al. Emergency department based study of risk factors for acute injury from domestic violence against women. Ann Emerg Med 1998;31:502–6.[CrossRef][Web of Science][Medline]
23 Subramaniam P, Sivayogan S. The prevalence and patterns of wife beating in the Tricomalee District in Eastern Sri Lanka. Southeast Asian J Trop Med Public Health 2001;32:145–8.
24 Jewkes R, Levin J, Penn-Kekana L. Risk factors for domestic violence: findings from a South African cross-sectional study. Soc Sci Med 2002;55:1603–17.[CrossRef][Web of Science][Medline]
25 Vizcarra MB, Cortés J, Bustos L, et al. Violencia conyugal en la ciudad de Temuco. Un estudio de prevalencia y factores asociados. Rev Med Chil 2001;129:1405–12.[Web of Science][Medline]
26 Kyriacou DN, Anglin D, Taliaferro E, et al. Risk factors for injury to women from domestic violence. N Engl J Med 1999;16:1892–8.
27 Johnson H. The role of alcohol in male's partners assaults on wives. J Drug Issues 2000;30:725–40.
28 Wilson P, McFarlane J, Malecha A, et al. Severity of violence against women by intimate partners and associated use of alcohol and/or illicit drugs by the perpetrator. J Interpers Violence 2000;15:996–1008.
29 Hutchison IW. Alcohol, fear and woman abuse. Sex Roles 1999;40: 893–919.[CrossRef][Web of Science]
30 Ramanathan S. Violence against women. Indian Med J 1996;3:145–8.
31 Roberts AR. Psychosocial characteristics of batterers: a study of 234 men charged with domestic violence offences. J Fam Violence 1987;2:81–93.
32 Figueredo AJ, Corral-Verdugo V, Frías-Armenta M, et al. Blood, solidarity, status and honour. The sexual balance of power and spousal abuse in Sonora, Mexico. Evolution and Human Behaviour 2001;22:295–328.[CrossRef][Web of Science]
33 Caetano R, Cunradi CB, Clark CL, et al. Intimate partner violence and drinking patterns among white, black and Hispanic couples in the U.S. J Subst Abuse 2000;11:123–38.[CrossRef][Web of Science][Medline]
34 Logan TK, Walker R, Leukefeld CG. Rural, urban influenced and urban differences among domestic violence arrestees. J Interpers Violence 2001;16:266–83.
35 Savona-Ventura C, Savona-Ventura M, Drengsted-Nielsen S, et al. Domestic abuse in central mediterranean pregnant population. Eur J Obstet Gynecol Reprod Biol 2001;98:3–8.[CrossRef][Web of Science][Medline]
36 Cunradi C, Caetano R, Schafer J. Religious affiliation, denominational homogamy, and intimate partner violence among U.S. couples. J Sci Study Relig 2002;41:139–51.[CrossRef][Web of Science]
37 Koenig MA, Lutalo T, Zhao F, et al. Domestic violence in rural Uganda: evidence from a community-based study. Bull World Health Organ 2003;81:53–60.[Web of Science][Medline]
38 Maffli E, Zumbrunn A. Alcohol and domestic violence in a sample of incidents reported to the police of Zurich City. Subst Use Misuse 2003;38:881–93.[CrossRef][Web of Science][Medline]
39 Coker AL, May Smith P, Mckeown RE, et al. Frequency and correlates of intimate partner violence by type: physical, sexual, and psychological Battering. Am J Public Health 2000;90:553–9.
40 Johnson H. Contrasting views of the role of alcohol in cases of wife assault. J Interpers Violence 2001;16:54–72.
41 Barnett O, Fagan R. Alcohol use in male spouse abusers and their female partners. J Fam Violence 1993;8:1–25.
42 Dent D, Arias I. Effects of alcohol, gender and role of spouse's attributions and evaluations of marital violence scenarios. Violence Vict 1990;5: 185–93.[Medline]
43 Fagan R, Barnett O, Patton J. Reasons for alcohol use in martially violent men. Am J Drug Alcohol Abuse 1998;14:371–92.
44 Kirkwood B R. Essentials of medical statistics. Oxford: Blackwell, 1998;151–2.
45 Dickersin K. The existence of publication bias and risk factors for its occurrence. JAMA 1990;263:1385–9.
46 Dickersin K, Min YI. Publication bias: the problem that won't go away. Ann N Y Acad Sci 1993;703:135–46.[Web of Science][Medline]
47 Sackett DL. Bias in analytic research. J Chronic Dis 1979;32:51–63.[CrossRef][Web of Science][Medline]
48 Loannidis JPA, Trikalinos TA, Ntzani EE, et al. Genetic associations in large versus small studies: an empirical assessment. Lancet 2003;361: 567–71.[CrossRef][Web of Science][Medline]
49 Trikalinos TA, Ntzani EE, Contopoulus-loannidis DG, et al. Establishments of genetics associations for complex diseases is independent of early study findings. Eur J Hum Genet. Published online 23 June Available from: http://www.nature.com/cgi-taf
50 Room R. Gender roles and interactions in drinking and drug use. J Subst Abuse 1996;8:227–339.[CrossRef][Web of Science][Medline]
51 Gorney IH. Domestic violence and chemical dependency: dual problems, duals interventions. J Psychoactive Drugs 1980;21:229–38.
52 Gondolf EW, Foster RA. Wife assault among VA alcohol rehabilitation patients. Hosp Community Psychiatry 1991;42:74–79.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  C. Vives-Cases, D. Gil-Gonzalez, and M. Carrasco-Portino Verbal Marital Conflict and Male Domination in the Family as Risk Factors of Intimate Partner Violence Trauma Violence Abuse, April 1, 2009; 10(2): 171 - 180. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||