This study aimed to investigate the predictors of recidivism in first-time driving under the influence (DUI) offenders, analyzing variables derived from medico-legal and toxicological examinations. The research was structured as a comparative study for the period 2012-2019. DUI offenders with a blood alcohol concentration >0.5 were included in the study. The case group consisted of recidivist offenders, while the comparison group consisted of first-time offenders. Personal data, socioeconomics, and parameters linked to the DUI were compared between the two groups. Significance was determined by chi-square and Mann-Whitney tests. To prevent confounding effects, multivariate binary logistic regression analysis was performed. Our sample encompassed 1678 subjects (196 in the case group, 1482 in the comparison group). Gender, driving license category, education, and tobacco use resulted in significant differences between the groups. In a model including age at DUI, education, and smoking habit as independent variables, higher educational levels (high school, bachelor's) and older age protected against recidivism, whereas smoking >20 cigarettes/day was an independent risk factor for recidivism. Recidivist offenders have specific characteristics indicating different therapeutic programs and carefulness in driving license regranting. A higher tobacco consumption in recidivists suggests that the use of this substance could influence the risk of DUI for reasons that will need to be explored.
Keywords: alcohol impairment; driving under the influence; medico-legal ascertainment; recidivism; tobacco use.