THURSDAY, July 24 (HealthDayNews) -- A new study has identified several factors that seem to independently predict which people will end up in abusive relationships as adults.

Not surprisingly, children who had behavior problems or witnessed their parents being violent toward one another were likelier to be involved in an abusive relationship later. But the study also found that, to a lesser degree, kids who were harshly punished were also headed down the path to abuse in adulthood.

About 20 percent of men and 20 percent of women report that they have abused their partner, with about the same percentage reporting being on the receiving end. For more serious violence (for instance, which might result in an injury and come to the attention of a medical professional), about 5 percent of women and 5 percent of men report being on each side -- victim and perpetrator.

Despite the prevalence of the problem, existing treatment programs for domestic violence do not seem to be working, says Miriam Ehrensaft, lead author of the new study, appearing in the August issue of the Journal of Consulting and Clinical Psychology. The current findings, Ehrensaft says, argue for early preventive measures in childhood.

Others take issue with the fact that the study only looked at the mother's parenting styles. "We have had concerns with studies that look just at the mother's parenting," says Maria Jose Angelelli, program policy coordinator for the Texas Council on Family Violence in Austin. "We're not disputing that that's not true, but how the father fits into that cannot be omitted. It can't."

Researchers already knew that three factors were likely to increase the risk that someone would end up in an abusive relationship as an adult: conduct disorder (a group of behavioral and emotional problems in children), exposure to parental violence, and "power assertive parenting" (which involves controlling, forceful tactics, including physical punishment).

"Each was assumed to be important in the development of partner violence, but we didn't know how each of these things might contribute independently," says Ehrensaft, who is an assistant professor of clinical psychology at Columbia University College of Physicians and Surgeons in New York City.

Now researchers have some insight into which factors may be more important.

"Conduct disorder is the most serious risk factor for partner violence, and that's closely followed by exposure to domestic violence," Ehrensaft says.

The risk from power-assertive parenting is much lower than the risk from the other two factors.

What's surprising, Ehrensaft says, is that "you don't actually need a history of physical abuse or exposure to actual physical violence between parents [to end up in an abusive relationship]."

The study authors followed 543 children from two upstate New York counties for 20 years, from childhood into adulthood. The youths and their mothers were first contacted in 1975 and were interviewed again on three additional occasions: in 1983, 1985-86, and 1991-93. In 1999, the participants were asked to fill out a questionnaire on recent life changes, work history, aggressive behavior, intimate partner history, and partner violence.

Children with conduct disorder were four times more likely to perpetrate violence against a partner and more than twice as likely to be on the receiving end.

Children who had witnessed violence between their parents were more than two times as likely to perpetrate and almost three times as likely to receive it.

Child abuse seemed to be a strong predictor of who might end up injuring a partner, but not who would end up receiving violence. That was another surprise, Ehrensaft says.

In fact, though, the two are highly related. Most people who perpetrate domestic violence are also recipients. "Most cases of aggression are going both ways," Ehrensaft says.

Another important piece of the puzzle -- and one that has implications for prevention and treatment programs -- is that the predictors for men and women are largely the same.

Angelelli is concerned with some of these findings.

In her experience, witnessing abuse has not been a link among victims of domestic violence. And childhood behavior disorders may be an issue but not at the top of the list.

"We would argue that a major factor in preventing partner violence would be prevention programs and providing safety to the victims and community resources to the victim and the perpetrator," she says. "Treating the child with a disorder we do not believe is a major factor. It's one of the factors."

Angelelli does agree that programs should be geared to girls and boys.

"We need to be addressing girls' risk factors as well," Ehrensaft says. "People have shied away from that because it will look like women blaming. For me, it's really a matter of protecting girls more thoroughly. If there is a risk factor for girls, it doesn't behoove us politically to ignore it."