Predictive factors associated with driving under the influence among Brazilian drug-using drivers.

The incidence of driving under the influence of psychoactive substances (DUI) and its recidivism can be curtailed by the proper identification of specific and predictive characteristics among drug users. In this sense, interpersonal violence (IV), psychiatric comorbidity and impulsivity seem to play an important role in DUI engagement according to previous studies. There are, however, limited data originated from low and middle income countries. In the present study, drug-using Brazilian drivers reporting DUI (n = 75) presented a higher prevalence of bipolar disorders (BD; DUI: 8% vs. non-DUI: 0%, p < 0.001), lower prevalence of obsessive-compulsive disorder (OCD; DUI: 0% vs. non-DUI: 12.6%, p < 0.001), and higher prevalence of childhood trauma (DUI: 65.3% vs. non-DUI: 46.8%, p = 0.022) than those not reporting DUI (n = 79). The evaluation of impulsivity though the Barratt Impulsivity Scale, which give impulsivity scores ranging from 30 to 120, showed higher impulsivity scores in the DUI group (80.4 ± 8) than in the non-DUI group (77.2 ± 10, p = 0.045). In general, subjects were young adults (mean age of 36 ± 9 years), Caucasians (58.4%), not married (61.0%), and with elementary schooling (40.3%) with no significant differences in demographic characteristics between drivers with and without DUI behavior. A multiple Poisson regression model showed that individuals reporting IV as perpetrators and history of childhood trauma were more likely to report DUI (PR: 1.66, 95%CI 1.22-2.7; PR: 1.57, 95%CI 1.02-2.42, respectively). The overlapping of violent situations (childhood trauma, IV and DUI) in some individuals presented here corroborates literature data suggesting that DUI can be an externalizing expression of a range of risky behavior, such as impulsiveness and aggressiveness. Moreover, while BD and higher impulsivity scores seem to act as risk factors for DUI, OCD was shown as a protective factor. These results corroborate the hypothesis that individuals with high risk for DUI could probably be identified by multidimensional assessment of cognitive, risky taking, and personality traits, which perhaps could facilitate the development of focused interventions.