Alexander and colleagues (1992) also found that various risky behaviors predicted later injury in their rural cohort. Participants who used alcohol or marijuana were more likely to report serious injury, as were those who generally exhibited a high risk-taking behavior pattern and those with school discipline problems. Indeed, substance use and general risk-taking behavior have been identified as antecedents of numerous specific injury types, as will be detailed in the following sections. Certain extracurricular activities were also associated with a greater risk of injury. Adolescents who played on sports teams and those who were employed were more likely than their peers to report later injury.

A case-control study, based on a population of 16- to 19-year-olds in Gwinnett County, Georgia, suggests that risky driving behaviors are associated with an increased risk of being in an MVC (CDC, 1994b). This study compared the self-reported risky driving behaviors among a group of 64 adolescents who were involved as drivers in an injury-producing MVC, with those among a group of 224 comparison adolescents. The comparison group included adolescents who were between the ages of 16 and 19 years, were licensed drivers, were never involved in a police-reported crash, and were otherwise randomly selected from the enrollment files of the Gwinnett County public high school. Adolescents who had been involved as a driver in an injury-producing MVC were more likely than adolescents in the comparison group to report driving more than 20 miles per hour over the speed limit, passing a car in a no-passing zone, and taking driving risks to enhance driving pleasure. These findings suggest that reckless driving behaviors are associated with an increased risk of being in an injury-producing MVC. Data from the 1997 National Survey of Speeding and Other Unsafe Driving Actions, the first national survey of its kind, suggest that risky driving behaviors are quite common, especially among adolescent and young adult drivers. For example, 82% of 16-20 year-old drivers reported driving 10 miles per hour faster than most other vehicles at some point during the past year, while 44% drove through a red light, 39% cut in front of another car to make a turn, and 26% raced another driver (NHTSA, 1998).

It is well known that driving after drinking increases the risk of an MVC. Some researchers have estimated the extent to which various blood alcohol concentrations (BAC) increase this risk. Zador, Krawchuk, and Voas (2000), for example, examined national data from the Fatal Accident Reporting System (FARS), which compiles information on almost all of the fatal MVCs on public roads in the U.S. The distribution of BAC of drivers involved in fatal single-vehicle crashes was compared with that of the general population of drivers, as determined by interviews and breath tests of a random sample of drivers. The relative risk of dying in a single-vehicle crash rose with increasing driver BAC. Males experienced a steeper increase in risk with each increase in BAC, when compared with females. Young drivers, ages 16 to 20 years, exhibited a sharper rise in risk with increasing BAC than did older drivers. The authors estimate that the risk of a fatal single-vehicle crash among male drivers, ages 16 to 20 years, more than doubles with each 0.02% increase in BAC. According to the 1999 YRBS, driving after drinking and riding with a driver who has been drinking are fairly common among high school students today. One-third of all high school students reported having ridden at least once during the 30 days prior to the survey with a driver who had been drinking (Kann, et al., 2000). About 13% reported driving after drinking themselves, a behavior that was more common among male students (17%) than among female students (9%). White students were more likely to report drink driving behavior than were black or Hispanic students. The prevalence of driving after drinking was higher among older than among younger students.

Adolescents who exhibit high levels of risk-taking behavior in general show higher levels of motor vehicle injuries in cross-sectional studies. Jelalian, Alday, Spirito, Rasile, and Nobile (2000) conducted two cross-sectional studies to evaluate the relationship between behavioral factors and motor vehicle injury. One study population was derived from one urban and two suburban high schools and included 1,576 students in the 9th through 12th grades. The ethnic distribution varied across the three schools, but was at least 50% white in each. Overall, the sample was about evenly split across genders, and represented a range of socioeconomic backgrounds. The second study population included 573 boys attending a parochial high school with a predominately white population. In each study population, adolescents who scored higher on a risk-taking scale, based on the frequency with which they reported engaging in each of six risky behaviors, were more likely than were their peers to have been injured in an MVC.

Finally, marijuana use has been associated with drinking and driving among adolescents. One short-term, longitudinal study followed 1,359 10th and 11th grade drivers, primarily white and middle class, for five to six months (Klepp & Perry, 1990; Klepp, Perry, & Jacobs, 1991). Marijuana use at baseline was found to predict drinking and driving behavior in this study population.