Predictors of driving under the influence (DUI) and non-DUI events among a multiethnic blue-collar cohort

This study assesses the contribution of substance use, demographic, and occupational factors to the likelihood of driving under the influence (DUI) and non-DUI events (e.g., speeding, failure to stop) among a multiethnic sample of urban transit operators (n=1282). Using an historical prospective study design, survey and medical exam data obtained from workers who participated in the 1993-95 San Francisco MUNI Health & Safety Study were matched against Department of Motor Vehicle records through 2000. Covariates were demographic factors (age, race/ethnicity, marital status) and occupational factors (years driving as a transit operator; full or part-time work status). Substance use measures were quantity-frequency of alcohol consumption, the CAGE screener for alcohol dependence, and smoking status (i.e., current, former, and never). During follow-up, 28 workers had a DUI, and 143 workers had a non-DUI event. Multivariable logistic regression results showed that alcohol consumption predicted DUI (OR=4.10; 95% CI 1.82, 9.24) and non-DUI events (OR=1.43; 95% CI 1.04, 1.96). African American transit operators were more likely to have a DUI (OR=3.10; 95% CI 1.22, 7.89) or a non-DUI event (OR=2.56; 95% CI 1.72, 3.82) compared to transit operators of other racial/ethnic groups. Current smokers were more likely to have a non-DUI event than non-smokers (OR=1.59; 95% CI 1.05, 2.43). Years driving as a transit operator were negatively associated with likelihood of having a non-DUI event. Occupational factors did not predict DUI. Further research is needed to investigate if aspects of the work environment contribute to increased drinking, and thereafter risk of non-occupational driving violations.