By Rick Weiss
Washington Post Staff Writer
Sunday, November 11, 2007
Researchers at Ohio State University garnered little attention in February when they found that youngsters who lose their virginity earlier than their peers are more likely to become juvenile delinquents. So obvious and well established was the contribution of early sex to later delinquency that the idea was already part of the required curriculum for federal "abstinence only" programs.
There was just one problem: It is probably not true. Other things being equal, a more probing study has found, youngsters who have consensual sex in their early-teen or even preteen years are, if anything, less likely to engage in delinquent behavior later on.
That new analysis, a reworking of the same data the Ohio team used, is one of several recent instances in which a more precise parsing of data has begun to turn long-standing societal presumptions on their head. By bringing evidence to bear on complex social issues, these studies are forcing individuals and policymakers to rethink such hot-button topics as the benefits of breast-feeding, the risks of teen child-bearing and, in the latest example, the harms long presumed to result from teen sex.
Like many of the newer studies, the latest one -- led by Paige Harden, a doctoral candidate in psychology at the University of Virginia in Charlottesville -- used the powerful techniques of behavioral genetics. The field specializes in studies on twins, research that can help tell whether behavioral traits are the result of genes or the social environment, and that has periodically stirred controversy when it has focused on the genetic underpinnings of criminality and intelligence.
But the specialty's analytic methods can also help tell whether one behavior, such as early sex, is merely correlated with or actually causes a second behavior that is often found with it, such as delinquency. If two behaviors often exist in the same people but are found not to be connected by cause and effect, then a third factor is likely to be causing both.
That kind of finding can help identify better targets for prevention efforts, experts say.
"Behavioral geneticists have long sought to establish causal links between genes and complex behaviors. So it's fascinating to see them use the tools of their trade to dispute widely held beliefs" about the social roots of some of those behaviors, said Erik Parens, a senior research scholar who has tracked the field intensively at the Hastings Center, a Garrison, N.Y., science and ethics think tank.
The latest example started when Dana Haynie, a sociologist at Ohio State, and her then-graduate student, Stacy Armour, published a study in February in the Journal of Youth and Adolescence. They analyzed data collected from more than 7,000 children as part of the National Longitudinal Study of Adolescent Health, a federally funded survey that in 1994 began gathering information about the health-related behavior of U.S. schoolchildren who were then in grades seven through 12.
Haynie and Armour divided the children into three groups based on when they first had sex: when they were younger, about the same age or older than the age at which most of their local peers lost their virginity. (It varies by region, but on average, U.S. children lose their virginity at age 16.) They also compiled information on graffiti-painting, shoplifting, drug-selling and other "problem behaviors" by those young people in later years.
Their conclusion: One year after losing their virginity, children in the early category were 20 percent more likely than those who started having sex at the average age to engage in delinquent behavior, even when several other relevant factors such as wealth, race, parental involvement and physical development were taken into account.
Those findings supported the widely held notion that loss of virginity at a relatively young age appears to, as Haynie and Armour wrote, "open the doorway to problem behaviors."
Harden, at the University of Virginia, didn't believe it.
Looked at from a similarly high altitude, she said, people might conclude that red meat is a health food, since people live longer in countries where more is eaten. Only when the issue is studied within one country does red meat's link to chronic diseases appear.
Suspecting such an error in the Haynie study, Harden and three colleagues, including her adviser, Eric Turkheimer, an expert in behavioral genetics, studied more than 500 pairs of twins in the same national survey analyzed by the Ohio team. Because twin pairs share similar or identical genetic inheritances (depending on whether they are fraternal or identical) and the same home environment, twin studies are useful for seeing through false cause-and-effect relationships.
The team looked at identical twin pairs in which one twin initiated sex younger than the other, then team members tallied subsequent problem behaviors. If sex really adds to the chances of delinquency, then early-sex teens should end up delinquent more often than their later-sex twins.
"It turns out that there was no positive relationship between age of first sex and delinquency," Harden said.
The way to reconcile that with the previous evidence of a link is to conclude that some other factors are promoting both early sex and delinquency, she said. In an e-mail, Haynie agreed. And the Virginia study, to appear in the March 2008 issue of the Journal of Youth and Adolescence, offers some clues.
It found that identical twins, who have the same DNA, were more similar to one another in the ages at which they lost their virginity than were fraternal twins, whose DNA patterns are 50 percent the same -- an indication that genes influence the age at which a person will first have sex. Other twin studies have found the same pattern for delinquency.
Together, those findings suggest that some genes -- perhaps, for example, those that increase impulsivity and risk-taking -- may underlie both behaviors.
"You need to have some appetite for risk-taking to be a delinquent. And the same if you're 13 and going to have sex for the first time," Harden said.
Efforts to prevent delinquency can hardly take aim at people's genes. But the Virginia study also indicates that social factors, as yet unidentified but perhaps involving relationships with family and friends, have an even bigger impact than genes on whether a child will become delinquent. Those are the things that should be identified and targeted by delinquency-prevention programs, said Jeanne Brooks-Gunn, co-director of Columbia University's National Center for Children and Families.
"I wouldn't be focusing on early sexuality . . . to alter rates of delinquency," she said.
Perhaps most surprising, the Virginia study found that adolescents who had sex at younger ages were less likely to end up delinquent than those who lost their virginity later. Many factors play into a person's readiness for sex, but in at least some cases sexual relationships may offer an alternative to trouble, the researchers say.
Even then, there are emotional and physical risks. Young adolescents, in particular, are less likely to use condoms and so are vulnerable to sexually transmitted diseases and unwanted pregnancies.
But those are risks that other nations have mitigated with education, Harden and Turkheimer said, while U.S. educators wanting a piece of the nation's $200 million "abstinence only" budget must adhere to a curriculum that links sex to delinquency and explicitly precludes discussion of contraception.
The new study "really calls into question the usefulness of abstinence education for preventing behavior problems," Harden said, "and questions the bigger underlying assumption that all adolescent sex is always bad."
Similar re-analyses have begun to undermine other conventional notions about health.
A recent study by Scottish researchers asked whether the higher IQs seen in breast-fed children are the result of the breast milk they got or some other factor. By comparing the IQs of sibling pairs in which one was breast-fed and the other not, it found that breast milk is irrelevant to IQ and that the mother's IQ explains both the decision to breast-feed and her children's IQ.
In another example, Arline Geronimus, a University of Michigan professor of health behavior who is now a fellow at Stanford University's Center for Advanced Study, knew that babies born to teenagers are more likely to die in their first year of life than those born to older women.
"But that is an apples-to-oranges comparison," she said. In New York City, for example, far more teen mothers live in Harlem than on the Upper East Side, she said, and "there are a lot of differences between those groups."
So Geronimus looked more closely and got a different answer.
"If you compare Harlem teen moms to Harlem older moms, you find that the kids of the teen moms are actually less likely to die," she said. The reasons include the fact that, unlike older women, poor teenagers are generally not juggling jobs and have older relatives to help.
It can make sense for poor women to have children when they are quite young, Geronimus concludes, and any effort to change that ought to treat it as an economic problem, not a health education problem.
In a different re-analysis, Geronimus made another counterintuitive finding. While it is true that, in general, teen mothers are less likely to breast-feed their babies than older moms, it is not true among poor women. Poor teenagers are actually more likely to breast-feed than poor older moms, in large part because the older women have jobs that don't grant them the time to breast-feed or pump milk.
Because of that misconception, programs promoting breast-feeding have targeted teens instead of older women, Geronimus said. And they have taken aim, in part, at a concern that teenagers were believed to have: the cosmetic effects of breast-feeding on their breasts.
"So you've targeted the wrong population," Geronimus said, "and come up with the wrong kind of intervention."