Stephen J.LurieMD, PhD, Senior EditorIndividualAuthorJody W.ZylkeMD, Contributing EditorIndividualAuthor
In Reply: Mr Komanoff correctly points out that the information about BACs presented in our article is inadequate for gauging the detailed distribution of BACs in the study subjects. For the 16 case bicyclists who tested positive for alcohol, BACs ranged from 0.02 g/dL to 0.35 g/dL, with a mean of 0.18 g/dL; for the 10 control bicyclists who tested positive for alcohol, BACs ranged from 0.02 g/dL to 0.13 g/dL, with a mean of 0.07 g/dL. The range of positive BACs was indeed wide, especially in the case bicyclists. We estimated the odds ratio of bicycling injury according to BAC 0.02 g/dL or greater and BAC 0.08 g/dL or greater, with BAC 0.02 g/dL or less being the referent group. We defined a positive BAC result as 0.02 g/dL or greater. A BAC of 0.08 g/dL is widely used in research and legislation as the cutoff point for being alcohol impaired. Given the modest sample size, we believe that our approach to the data analysis is scientifically sound. It would be desirable to determine the exact "dose-response" relation between BAC and the risk of bicycling injury in a considerably larger sample.
Li G, Baker SP. Elevated Blood Alcohol and Risk of Injury Among Bicyclists—Reply. JAMA. 2001;285(19):2445-2446. doi:10.1001/jama.285.19.2445