In a new study publishes in the May issue of PLoS Medicine, Joel Ray and colleagues establish the relative risk of assault per volume of alcohol sales from retail outlets in Ontario. The study focuses specifically on adult victims of assault.
In a population-based case-crossover analysis, Joel Ray and colleagues find that the risk of being a victim of serious assault increases with retail alcohol sales, especially among young urban men.
Ray and colleagues sought to add to the existing evidence on the link between alcohol sales and risk of assault. Survey data have been the preferred source of many previous studies, and the authors state that such methods tend to lack adequate controls. For this reason the authors used computerised medical records and accurate sales data from Ontario, Canada. Using an empirical case-crossover method seldom used in studying alcohol and violence, the authors aimed to elucidate further the link between alcohol sales and risk of being a victim of assault.
The authors chose a case-crossover design as a way to address a number of concerns regarding potential bias in previous studies. The case-crossover method is used to address the high likelihood of between-person confounding in previous empirical work, where uncontrolled differences among individuals can affect results. Epidemiologists have long favoured such a study design, whereby a case’s exposure prior to or during the event in question is compared with that case’s exposure at other times. This essentially allows a case to act as his or her own control. In Ray and colleagues’ study, the exposure period was the day before an assault case’s hospitalisation and the control period was seven days earlier. The volume of alcohol sold at the store in closest proximity to an assault victim’s home on the day before the assault was compared to the volume of alcohol sold at the same store seven days earlier.
The authors found that increasing alcohol sales were associated with a 13% (95% confidence interval [CI] 2–26) higher relative risk of hospitalisation for assault, for every 1,000 litres more alcohol sold per store per day. A 1,000 litre rise in alcohol sold per day equates to an approximate doubling in the level of usual daily sales. The relative risk was highest for spirits and lowest for beer. While a 1,000 litre rise in the daily sales of all alcohol proved significant for the relative risk of males being hospitalised (relative risk 1.18, 95% CI 1.05–1.33), this was not found for females (relative risk 0.89, 95% CI 0.68–1.18).
Source: Ray JG, Moineddin R, Bell CM, Thiruchelvam D, Creatore MI, et al. (2008) Alcohol sales and risk of serious assault. PLoS Med 5(5): e104. doi:10.1371/journal.pmed.0050104t
Dr Geoff Lowe comments:
This research/study shows that the risk of being hospitalized from a violent assault increases shortly after periods of higher alcohol sales near the victims’ homes, and that young men seem to be particularly vulnerable.
One of the main strengths of this study is the use of a population-based, case-crossover analysis, accessing detailed data from comprehensive records provided by Ontario’s remarkable monitoring systems.
There is already evidence regarding the link between alcohol sales and risk of assault. For instance, in Norway researchers found increasing rates of violence in those areas with greater density of licensed premises (1). However, another study in Sweden noted slightly increased alcohol sales following increased availability in alcohol retail shops, but no increases in assault cases (2). More recently, researchers in Australia observed that the increase in number of alcohol outlets was linked with increases in alcohol-related violence and assault in the local area (3). Somewhat unexpectedly, however, most of this increased violence occurred in private homes rather than at licensed premises. (The location of assault was something that the Ontario study did not address).
In some other countries, where alcohol is sold in far more locations (than in Ontario, Canada), any specific links between alcohol sales and violent assault would be much harder to identify.
The Ontario study certainly adds to our knowledge and understanding of these potential links. It does so in a methodologically impressive way – despite some limitations. For instance, although alcohol sales may be high shortly before hospitalization, the researchers provide no evidence that blood-alcohol-levels are high.
The authors acknowledge one anomaly occurring around Christmas - with sales of alcohol increasing (as expected), but accompanied by a relative decline in assault cases. This suggests that social context and expectancies may well be important influences.
Other commentators have suggested that the observed increases in alcohol sales might be part of the ‘pre-loading’ phenomenon, whereby drinkers consume more at home before going out to pubs or clubs. At the end of the night their intoxication levels are likely to be higher (than ‘non-pre-loaders’), and it’s this excessive consumption, rather than alcohol per se, that contributes to violence and aggression.
Many studies have shown that links between alcohol and aggression are complex, with aggressive traits likely to be more problematic than alcohol itself. It is generally acknowledged that certain individuals are more prone to violence under the influence of alcohol (and many studies highlight the susceptibility of young, urban men). So questions still remain about what kinds of individuals end up being violently assaulted. In what sorts of places do these assaults occur, and under what conditions (e.g. sporting events, nightclubs, neighbourhood rivalry, etc)?
There is still a long way to go, but gradually, by pulling together all the various methodological approaches, we are gaining further insights into the links between increased alcohol consumption and violence.
(1) Norström, T. Outlet density and criminal violence in Norway, 1960-1995. J.Stud.Alcohol (2000), 61, 907-911.
(2) Norström, T. & Skog, O-J. Saturday opening of alcohol retail shops in Sweden: An impact analysis. J.Stud.Alcohol (2003), 64, 393-401.
(3) Livingston, M. Alcohol outlet density and assault: A spatial analysis. Addiction (2008), 103, 619-628.
