Gender Differences in the Relationship Between Alcohol and Violent Injury: An Analysis of Cross-National Emergency Department Data * (Report‪)‬

CONSISTENT EVIDENCE HAS SHOWN THAT a pattern of heavy drinking, alcohol use before injury, and positive and high blood alcohol concentrations (BAC) (i.e., exceeding .08, or 8 mg of alcohol/100 ml of blood) detected among injury patients in emergency departments are significantly associated with a greater likelihood of violent injury, as compared with nonviolent injury (Cherpitel, 1993; Macdonald and Wells, 2001; Macdonald et al., 1999, 2005). This evidence implicates the consumption of alcohol in violent injury episodes. Drinking impairs cognitive and psychomotor abilities, decreasing the ability to recognize, avoid, or escape danger (see Abbey et al., 2004; Logan et al., 2002; Steele and Josephs, 1990; Stritzke et al., 1995). Additionally, the effects of alcohol may increase the likelihood that the victim will be chosen as a target for aggression because of increased vulnerability (Miczek and Barry, 1977). Drinking by the victim also may result in behaviors that precipitate violence, such as reacting with physical aggression (Streifel, 1997) or engaging in risky or provocative behavior (see McClelland and Teplin, 2001). An important question is whether the association between alcohol and violent injury is different for men and women. Some evidence suggests that alcohol may contribute to violent injury more for men than for women. Steen and Hunskaar (2004) found that male victims of violence visiting an emergency department were more likely than female victims to be under the influence of alcohol, as judged by the attending physician. In an analysis of data from the Emergency Room Collaborative Alcohol Analysis Project (ERCAAP), consisting of studies from 30 emergency departments from seven countries, attributable risks for violent injuries associated with having a positive BAC, self-reported alcohol consumption before the injury, and self-reported frequent heavy episodic drinking (five or more drinks at least monthly) were greater for men than for women (Cherpitel et al., 2005). Notably, however, because attributable risks are computed based on both relative risks and the prevalence of the risk factor (i.e., alcohol), this evidence may at least partly reflect that drinking and heavy drinking are simply more common among men than among women (World Health Organization, 2004). That is, if more men than women drink alcohol and are heavy drinkers, it stands to reason that a larger proportion of violent incidents experienced by men than by women will involve alcohol and will be reported by heavy drinkers. To illuminate whether gender differences exist in the explanatory role of alcohol in violent injury, tests for gender by alcohol interaction effects in the prediction of violent injury are required. No studies were found in the empirical literature on alcohol and violent injury that have conducted such tests.