A population-based case-crossover and case-control study of alcohol and the risk of injury.

OBJECTIVE This study investigated the relationship between alcohol and the risk of acute injury, examining both short-term (drinking over a few hours) and long-term exposures (past-month hazardous drinking and current alcohol use disorders). METHOD The study was a case-crossover and population-based case control study. Cases (N = 2,517; 1,432 men) were injured patients recruited from all three emergency departments in Boone County, Missouri. Each case's alcohol consumption in the 6 hours prior to injury was compared to his or her consumption the day before in a case-crossover analysis. Community controls (N = 1,856; 948 men) were recruited by telephone and matched to cases by age, gender, day of week and hour. Case-control analyses examined recent alcohol consumption (past 6 hours), past-month hazardous drinking and past-year alcohol use disorders. RESULTS The odds ratios (ORs) for injury associated with short-term alcohol exposure were similar, whether cases were compared with themselves (case-crossover analyses) or with others (case-control analyses). After 1 or 2 drinks in a 6-hour interval, the ORs were 1.8 (95% confidence interval [CI]: 1.3-2.6) and 1.5 (0.96-2.2), respectively. Similar dose-response curves were seen in the two comparisons, with ORs of 6.2 and 3.7, respectively, after 3 or 4 drinks and 9.5 and 13.5 after 5 or 6 drinks. For past-month hazardous drinking, the OR was 1.7 (95% CI: 1.2-2.3). Current alcohol dependence was associated with injury (OR = 1.9; 95% CI: 1.5-2.6), but alcohol abuse was not (OR = 0.9; 95% CI: 0.8-1.1). CONCLUSIONS Alcohol's effect on injury risk is related more strongly to acute exposure than to measures of long-term exposure. The risk is significant even at low levels of consumption.