Research of Note for February 2003

Print More

Some research provides new information that surprises us, while other studies help us think about why a well-accepted assumption is true. Results from a new study on working teens will surprise some youth workers. In contrast, findings from a study into the intelligence and abilities of young teen mothers confirm stereotypes, but raise questions about why. A third study unpersuasively attempts to answer an important question about the common use of drugs to treat hyperactivity.

Teen Jobs and Drug Use

The Relationship Between Employment and Substance Abuse Among Students Aged 12 to 17
Li-Tsy Wu, Sc.D., William Schlenger, Ph.D., and Deborah Galvin, Ph.D.
Journal of Adolescent Health, Vol. 32, No. 1, January 2003, pp. 5-15.
Available free from Wu at or Center for Risk Behavior and Mental Health Research, RTI International, P.O. Box 12194, Research Triangle Park, NC 27709-2194.

About one in six adolescents in the U.S. holds a job while going to school. This is the latest of several major research studies to show that the risks from teenage jobs can outweigh the benefits.

Using data from the National Household Surveys on Drug Abuse (conducted by the U.S. Substance Abuse and Mental Health Services Administration), the authors compared a national sample of more than 7,500 students ages 12 to 17 who did not have jobs with more than 1,000 who worked part time and 96 who worked full time.

Youths who worked either part time or full time were more likely to smoke, drink or use illegal drugs than those who did not have jobs. For example, 24 percent of students who worked full time had binged on five or more alcoholic drinks at one time in the past month, compared with 12 percent who worked part time and 6 percent who were not employed.

Results were similar for heavy drinking, defined as having five or more drinks on each of at least five occasions in the past month – a criterion met by 13 percent of youths working full time, 5 percent of part-timers and 2 percent of those who didn’t work. The same was true for heavy cigarette use (at least a pack a day), which was reported by 10 percent of full-timers, 3 percent of part-timers and 1 percent of nonworkers.

The differences were not as great, however, for heavy drug use, which was reported by 5 percent of full-time workers, and by 2 percent of part-timers and nonworkers. Heavy drug use was defined as daily marijuana use; heroin use at least once; or weekly use of cocaine, inhalants, pain relievers, sedatives, tranquilizers, stimulants or hallucinogens.

Is there something about work that causes drug use? Not exactly. School dropouts who were unemployed were the most likely to report heavy drug use (8 percent) and were similar to full-time workers in rates of binge drinking (21 percent), heavy drinking (12 percent) and heavy smoking (7 percent).

Because older students are more likely to drink, smoke and use drugs, and more likely to have jobs, that could explain the relationship between jobs and drugs. Fortunately, the researchers conducted statistical analyses taking age, race, ethnicity, mental health problems and other factors into account. They found that even when those traits were statistically controlled, youths who worked were more likely to report recent and heavy substance use.

The researchers also found differences for boys and girls. Among them: Boys who work full time are more likely to smoke, smoke heavily, binge on alcohol, drink heavily and use marijuana compared with those who don’t work. Among females, those who worked full time or part time were more likely to use alcohol than were other girls.

The reasons for the link between work and substance use is not known, but likely explanations include exposure to co-workers with those habits, the availability of drugs and alcohol through co-workers, and the availability of earnings to buy drugs
and alcohol.

In addition, research shows that students who work tend to have poorer grades and be less committed to doing well in school or going to college, so they might not be as motivated to avoid behaviors that could interfere with those goals.

The implications for adults who counsel, mentor or supervise teens are clear: Be aware of the risks that jobs can present for teens, and find ways to minimize those risks. If kids need the money or other benefits that work provides, try to help them find jobs that won’t expose them to negative influences. Encourage some controls on the earnings so they won’t be spent on alcohol, cigarettes and drugs.

Teen Pregnancy and Intelligence

Association of Early Childbearing and Low Cognitive Ability
Darlene Shearer, Beverly Mulvihill, Lorraine Klerman and others.
Perspectives on Sexual and Reproductive Health, Vol. 34, No. 5, September/October 2002, pp. 304-309.
Available free at

The stereotype of young teen mothers is that they tend to be from low-income families and not as bright as other teenagers. This national study indicates that the stereotype is correct.

The researchers focused on more than 2,000 14- to 22-year-old women who participated in the federal government’s National Longitudinal Survey of Youth and were followed for 16 years. They compared young women who had a first baby before the age of 18 with teens of the same age, race, ethnicity, geographic location and urban/rural status who did not have babies.

The young teen moms had lower cognitive scores on a standardized test than the other girls, and those who went on to have a second baby before the age of 20 had even lower cognitive scores. The test was the Armed Services Vocational Aptitude Battery, which was designed by the military to measure 10 areas, including reasoning, comprehension and knowledge.

Two-thirds of the early teen moms scored in the bottom half of national scores on the test, and only 11 percent were in the top quarter. Of those who had a second child before 20, 70 percent scored in the bottom half on the test and only 7 percent were in the top quarter.

Girls who had children early also had sexual intercourse for the first time an average of almost two years earlier than their classmates (15.2 years compared with 17.5 years). They were also less likely to have had a sexuality education course (47 percent compared with 65 percent) and were less likely to know when during the menstrual cycle women are most fertile (25 percent compared with 36 percent).

Sexual experience is only part of the reason for the difference in pregnancy rates. On average, the girls with the lowest intelligence scores had sexual intercourse for the first time 1.4 years earlier than those with the highest cognitive scores, which is not an enormous difference.

Early teen moms also tended to grow up in families below the poverty level, and their educational goals were two years lower than those of other girls. Not surprisingly, lower cognitive scores were linked to poverty and low aspirations. However, the link between cognitive scores and teen pregnancy was still significant when poverty and aspirations were controlled.

There are many possible explanations for the link between teen pregnancy, intelligence and knowledge, and this study can’t determine which are most important. For example, smarter girls and girls who do better in school might be more likely to use birth control correctly, or to be more motivated to avoid pregnancy or childbirth so that they can succeed in educational or career goals.

A less intelligent or accomplished girl might be more easily influenced by her sexual partner, especially if he is smarter than she is. This is consistent with research on age differences, which shows that girls with much older boyfriends are more likely to have sex and to become pregnant.

In addition, less capable girls, especially those who are mentally retarded or borderline retarded, would be expected to have poorer judgement and be less able to consider the consequences of their sexual activity.

Although it raises interesting questions, this study has a rather important shortcoming: The data were collected between 1979 and 1990. As sexuality education has become more available and as unmarried mothers have become more accepted, it is possible that cognitive abilities are less important in predicting teen motherhood. Unfortunately, it is not possible to tell from this study.

Do Prescription Drugs Result in Later Drug Abuse?

Does the Treatment of Attention-Deficit/ Hyperactivity Disorder With Stimulants Contribute to Drug Use/Abuse? A 13-Year Prospective Study
Russell Barkley, Ph.D., Mariellen Fischer, Ph.D., Lori Smallish, M.A., and Kenneth Fletcher, Ph.D.
Pediatrics, Vol. 111, January 2003, pp. 97-109.
Available from Barkley at or College of Health Professions, 19 Hagood Ave., Charleston, SC 29425.

The dramatic increase in the use of Ritalin and similar stimulants for Attention-Deficit/Hyperactivity Disorder (AD/HD) has raised concerns about its long-term impact. Does a child’s dependence on drugs increase the chances of that child using or abusing drugs? This study says no. Unfortunately, there are serious questions about the study design.

The study is based on 147 “rigorously diagnosed” AD/HD children ages 4 to 12 in 1979-80. They were evaluated again in 1987-88 and in 1992-96, when they were young adults. Of the 119 people for whom medication information was the most available, 98 had been treated with Ritalin or similar stimulants while in elementary school (usually for several years), while 32 reported that they had taken these medications while in high school.

Those who were medically treated for AD/HD were more likely to use drugs than those with AD/HD who had not been so treated, but on most measures these differences were not statistically significant. For example, 52 percent of the medically treated group had tried cigarettes (compared with 30 percent of the others), 20 percent had tried marijuana (compared with 6 percent), and 5 percent had tried cocaine (compared with 0 percent).

When differences are not statistically significant, the assumption is that they occurred by chance. However, it is important to acknowledge (which the authors do not) that the sample size is modest and that differences of this magnitude would have been statistically significant if the sample had been larger.

The next analysis compared young adults who were given Ritalin and similar stimulants in high school with those who had not. Only 32 were treated with these drugs in high school, compared with 115 that were not.

One out of four (26 percent) of adults who had been treated with AD/HD drugs in high school reported having ever used cocaine, compared with 5 percent of those not treated. That was the only statistically significant difference in drug use between the two groups.

The differences in cocaine use are worrisome, but the researchers found that if severity of AD/HD symptoms was taken into account, any differences were no longer significant. In fact, when comparisons were made between children who took AD/HD drugs for more than a year and those who took it less than a year, the link between medication and cocaine use also disappeared.

The researchers explain that children or youth with more severe symptoms are more likely to use cocaine. Children and youth with more severe symptoms are more likely to be treated with drugs and are more likely to use cocaine, but if those drugs are effective, the chances of using cocaine are reduced.

This study is meant to be reassuring, but is it? The sample of children who were not treated with drugs is too small to provide the kind of information this study is attempting to determine. It is not possible to control meaningfully for other important variables, such as severity of hyperactive symptoms, when there are only 21 children who were not treated. Similarly, the number of high school students who were treated with AD/HD drugs is also rather small (32). As a result, it is difficult to know if the drug abuse differences found occurred by chance, or are real differences that would be significant if the number of children studied were larger.

Most children who are diagnosed with AD/HD are treated with drugs, especially if their symptoms are severe and if the drugs are effective for those children. That makes it very difficult to study what happens to untreated children. On the other hand, there isn’t much point in doing a study like this if the sample is too small to evaluate persuasively.