Predictors of Drinking-Related Traffic Incidents in a High-Risk Sample

Alcohol is a factor in nearly one third of the traffic-related deaths in the United States. The age group at highest risk for being involved in a fatal crash with a blood alcohol concentration higher than 0.08% is 21-24 years (Centers for Disease Control). Understanding early factors that may help prevent drinking and driving in early adulthood is of great importance.

The Michigan Longitudinal Study, ongoing for over 28 years, is a population-based study of over 450 families at high risk, based on parental alcoholism. Assessments of mental health and behavior occur every three years and include the HOME scale for parenting, parent monitoring as reported by adolescents, and parent alcoholism. Young adult offspring reported getting in accidents due to alcohol use and receiving a citation for drinking and driving. Two outsides sources of information were also obtained. Teacher reports of child behavior were collected once the child entered school. In addition, driving records were obtained from the state Department of Transportation and included alcohol-related convictions. The present study combined the self-report of alcohol-related tickets and accidents with state data and examined distal predictors of these alcohol-related driving incidents.

This study included 756 offspring (71% male, average age 22.9, SD=3.0). Of these, 11.8% had a conviction for an alcohol-related traffic violation (including drunk driving as well as under-age possession of alcohol) and 14.5% reported any ticket or accident while drinking. Combining self-report and state data showed that 18.0% of the sample had an alcohol-related incident.  Boys were more likely than girls to have either self-reported incident or conviction (20.5% vs 11.8%, chi-square=8.0, p < .01).  Logistic regression showed that the child’s alcohol-related incidents were predicted by parent alcoholism measured at ages 3-5 (chi-square=15.0, p <.001). Parent report of externalizing behavior problems in childhood, family socioeconomic status and nurturing quality in the preschool home environment were unrelated to the likelihood of a drinking related traffic incident (logistic regression, all p > .1). Teacher reported delinquency at age 12-14 predicted higher likelihood of an alcohol related traffic incident in young adulthood (logistic regression, odds ratio = 1.06, p < .05). This effect was partially mediated by the adolescent's self-report of parent monitoring at age 15-17 (logistic regression, odds ratio = 0.58, p < .05).

These results show that risky driving is predicted by familial, presumably genetically-mediated processes, as well as by parent and child behavior. It particularly points to the importance of parent monitoring help reduce drinking and driving.

Supported by NIAAA grant R01 AA07065 (RAZ & MMH)