Research article
A case-crossover study of alcohol consumption, meals and the risk of road traffic crashes
-
* Corresponding author: Stefano Di Bartolomeo stefano.dibartolomeo@sanita.fvg.it
1 Agenzia Regionale della Sanità del Friuli Venezia Giulia/Cattedra di Epidemiologia, DPMSC, Università degli Studi di Udine, Via Pozzuolo 330, 33100 Udine, Italy
2 SOC Istituto di Igiene ed Epidemiologia, Azienda Ospedaliero-Universitaria di Udine, Via Colugna 50, 33100 Udine, Italy
3 S.O.C. Pronto Soccorso, Azienda Ospedaliero-Universitaria di Udine, Pzza SM Misericordia, 33100 Udine, Italy
4 Cattedra di Epidemiologia, DPMSC, Università degli Studi di Udine, Via Colugna 50, 33100 Udine, Italy
5 Cattedra di Epidemiologia, DPMSC, Università degli Studi di Udine and SOC Istituto di Igiene ed Epidemiologia, Azienda Ospedaliero-Universitaria di Udine, Via Colugna 50, 33100 Udine, Italy
BMC Public Health 2009, 9:316 doi:10.1186/1471-2458-9-316
Published: 1 September 2009Abstract
Background
The case-crossover (CC) design has proved effective to investigate the association between alcohol use and injuries in general, but has never been applied to study alcohol use and road traffic crashes (RTCs) specifically. This study aims at investigating the association between alcohol and meal consumption and the risk of RTCs using intrapersonal comparisons of subjects while driving.
Methods
Drivers admitted to an Italian emergency room (ER) after RTCs in 2007 were interviewed about personal, vehicle, and crash characteristics as well as hourly patterns of driving, and alcohol and food intake in the 24 hours before the crash. The odds ratio (OR) of a RTC was estimated through a CC, matched pair interval approach. Alcohol and meal consumption 6 and 2 hours before the RTC (case exposure window) were compared with exposures in earlier control windows of analogous length.
Results
Of 574 patients enrolled, 326 (56.8%) reported previous driving from 6 to 18 hours before the RTC and were eligible for analysis. The ORs (mutually adjusted) were 2.25 (95%CI 1.11-4.57) for alcohol and 0.94 (0.47-1.88) for meals. OR for alcohol was already increased at low (1-2 units) doses - 2.17 (1.03-4.57) and the trend of increase for each unit was significant - 1.64 (95%CI 1.05-2.57). In drivers at fault the OR for alcohol was 21.22 (2.31-194.79). The OR estimate for meal consumption seemed to increase in case of previous sleep deprivation, 2.06 (0.25-17.00).
Conclusion
Each single unit of acute alcohol consumption increases the risk of RTCs, in contrast with the 'legal' threshold allowed in some countries. Meal consumption is not associated with RTCs, but its combined effects with sleepiness need further elucidation.