Research of Note for July-August 2004

Print More

New Findings about Risky Behaviors

Youth Risk Behavior Surveillance – United States, 2003
Jo Anne Grunbaum, Ed.D., and colleagues
Morbidity and Mortality Weekly Report, May 21, 2004,
Vol. 53, No. SS-2
Centers for Disease Control and Prevention, U.S. Department of Health and Human Services
Available free at

Adolescents engage in many risky behaviors, and youth workers talk about those behaviors all the time: excessive drinking, drug abuse, violence and unprotected sex, to name a few. But this new study by the Centers for Disease Control and Prevention (CDC) shows that the behaviors that kill adolescents and young adults are not exactly what most of us think they are. It also reveals substantial differences in teen behavior among different states and communities, which provides reassuring evidence that these behaviors are not inherent in adolescence and can be changed.

The report starts with this rather surprising fact: Seven in 10 deaths of anyone between 10 and 24 years old are from four causes: motor vehicle accidents (32 percent), other unintentional injuries (12 percent), homicide (15 percent) and suicide (12 percent).

When we talk about teens’ risky behaviors, we tend to focus on drug and alcohol use, unprotected sex and violence, not car crashes. Of course, the age range for such behaviors extends beyond adolescence, and many of the deaths in these four categories could be related to drug and alcohol use. Nevertheless, I wasn’t sure how to interpret these mortality statistics.
Fortunately, the report is full of very specific information about high school students that provides important insights for youth workers.

Unfortunately, however, the report also includes some interpretations of data that seem more ideological than factual. For example: In the very brief summary of the report, the researchers point out that high schoolers engage in four behaviors that increase their risk of death. Here are the four top risk behaviors engaged in over the previous 30 days, according to the summary:

• Thirty percent rode in a motor vehicle driven by someone who had been drinking.
• Seventeen percent had carried a weapon.
• Forty-five percent had drunk alcohol.
• Twenty-two percent had used marijuana at least once.

The first two were obviously dangerous, but it was not clear how drinking alcohol or smoking marijuana once necessarily increased the risk of death. Isn’t drunk driving more dangerous than smoking marijuana? Where’s the evidence that these one-time acts are deadly? It is not in this report.

However, the report contains plenty of real findings of interest for the reader who sticks with the data instead of relying on the interpretation of the data.

This study is based on the Youth Risk Behavior Survey, which was completed anonymously and voluntarily by more than 15,000 public and private high school students in grades 9 through 12 between February and December 2003.
The report includes a wide range of information about risk behaviors. This column focuses on those that are most relevant to the statistics about risk behaviors that can kill.

Unintentional Injuries: Seat Belts, Bike Helmets and Drunk Driving

The survey found that 18 percent of these students rarely or never wear a seat belt when they are passengers in a car. That seems high, but is actually down from 26 percent when the same survey was conducted 12 years ago.
The percentage of non-use is higher among males (22 percent) than females (15 percent), and higher among black males (26 percent) than black females (16 percent). Failure to use seat belts varies among states, from 6 percent (Utah) to 23 percent (Mississippi).

While seat belt use is high, bicycle helmet use is rare. Of kids who ride bikes, 86 percent rarely or never use helmets. Non-usage is higher among blacks (95 percent) and Hispanics (90 percent) than whites (84 percent). Boys are less likely to use helmets than girls. There is tremendous variation among the states, ranging from 52 percent in Vermont to 96 percent in Mississippi that rarely or never use helmets.

The percentage of youth reporting that in the past 30 days they had ridden in a car with a driver who had been drinking was dangerously high: 30 percent for all youths, and somewhat higher among Hispanics (36 percent) than blacks (31 percent) or whites (29 percent). The rates were higher among Hispanic girls (40 percent) than Hispanic boys (32 percent), white girls (30 percent) and black girls (30 percent ).

Youth were more than twice as likely in some states to be passengers in vehicles with drivers who had been drinking: The range was from 18 percent (Utah) to 43 percent (North Dakota).

What about all the education efforts to “not let friends drive drunk”? Well, the number of high schoolers who in the past month had ridden in a car driven by a driver who had been drinking has actually decreased, to 30 percent from 40 percent 12 years ago. The percentage who admitted to having driven after drinking has also decreased, to 12 percent from 17 percent.

The drivers who had been drinking were more likely to be boys (15 percent) than girls (9 percent). The racial differences are very small. Again, there is tremendous variation among different states, ranging from 7 percent (Utah) to 27 percent (North Dakota).

Intentional Injuries: Weapons

Many high school students carry a weapon such as a gun, knife or club, but the 17 percent who said they had carried a weapon in the previous 30 days was lower than the 26 percent who had done so 12 years ago. The report pointed out that carrying of weapons has remained at about 17 percent since 1997.

Carrying a weapon was more likely among boys (27 percent) than girls (7 percent). Despite stereotypes to the contrary, carrying weapons was slightly more likely among white males (27 percent) than black males (25 percent) or Hispanic males (24 percent). Black girls (10 percent) and Hispanic girls (8 percent) were more likely to carry a weapon than white girls (6 percent).
Carrying a gun was more common among boys (10 percent) than girls (2 percent). The racial differences were minor for both genders.

From freshman to senior year in high school, there was no pattern of increase or decrease in carrying weapons or guns. Carrying a weapon was least likely among youth in Rhode Island (12 percent) and most likely in Wyoming (25 percent), while carrying a gun was least likely in Massachusetts (3 percent) and most likely in West Virginia (11 percent).


The survey asked whether the youths had seriously considered suicide, made plans to commit suicide or attempted suicide. Seventeen percent had seriously considered committing suicide within the past year, and 17 percent said that they had made a plan to commit suicide.

This was one of the few risky behaviors that was higher among girls (21 percent) than boys (13 percent). Hispanic girls (23 percent) and white girls (21 percent) were more likely to seriously consider suicide than black girls (15 percent). Among boys, the racial differences in seriously considering suicide were small, ranging from 10 percent and 13 percent.

Compared with other risk behaviors, there was less variation among states, ranging from 14 percent (Alabama, Mississippi, New York, North Dakota and Rhode Island) to 21 percent (Wyoming) for seriously considering suicide, and from 11 percent (New York, North Dakota, Rhode Island) to 17 percent (Nebraska) for making a suicide plan.

Nine percent of the youth reported having attempted suicide at least once in the previous 12 months. Suicide attempts were more than twice as high among girls (12 percent) than boys (5 percent). Attempts were also higher among Hispanics (11 percent), than blacks (8 percent) and whites (7 percent), and especially high among Hispanic girls (15 percent).

Smoking, Drug Use and Obesity

Some risky behaviors are rarely deadly in the short-term but can be very deadly in the long-term. The drug statistics indicate a lot of lifetime drug use and recent drug use for a wide range of substances, but the frequency of drug use was not reported in ways that were important for understanding the long-term health implications.

For example, the study reports that 16 percent of the youths had smoked cigarettes every day for at least 30 days, but it doesn’t say if they smoked one cigarette a day or two packs. The study also says that on the days that they had smoked within the past 30 days, 3 percent smoked more than 10 cigarettes – but it doesn’t specify whether they smoked on one day or every day in those 30 days.

However, the study does include some frightening information, such as 3 percent saying they’ve used heroin at least once, 11 percent saying they’ve used Ecstasy at least once, and 6 percent saying they’ve used non-prescribed steroid pills or shots.

Not surprisingly, the report shows that very few teens ate the recommended five portions of fruits and vegetables every day for the past week, and many do not take physical education classes or get regular exercise. A few specific statistics are clearly dangerous. For example, only 17 percent drank at least 3 glasses of milk each day for the past week. That could mean a substantial risk of bone fractures now and osteoporosis in later years, unless the youths ate lots of yogurt or cheese, or drank orange juice fortified with calcium instead. It seems unlikely, but that information wasn’t reported.

There’s something for every youth worker in this comprehensive report. It covers many other topics of interest, including sexual behavior and contraceptive use, alcohol or drug use on school property, dating violence, dieting and use of diet pills, TV-viewing and sports participation. Although some of the information is available in other government surveys, it is rare to have a study published within six months of collecting the data.

How representative of high school students nationwide are these findings? Of course, these are the kids who are still in school, and the study does not include kids of the same age who are home-schooled or who dropped out of school or died prior to the study. However, the researchers picked public and private schools that would be representative, and achieved a very good response rate, with 81 percent of the schools agreeing to participate, and 83 percent of the students filling out the surveys.

To make these findings more useful to youth workers, it would be helpful to evaluate why youth in some states engage in risky behaviors less often than in other states, and see what we can learn.

Diana Zuckerman, Ph.D., is president of the National Center for Policy Research for Women and Families. Contact: Information about many Research Watch columns is available at