But new research from the 91Ě˝»¨ suggests that troubling behaviors exhibited by abused children can be predictors of later criminal activity, and that those indicators differ between boys and girls.

The , published Aug. 11 in the Journal of Interpersonal Violence, found that elementary-aged boys who show “externalizing” behaviors such as arguing, disobedience and fighting are more likely to commit crimes as adults, but girls who similarly acted out were not. Instead, it found elementary-aged girls who exhibited signs of “internalizing” behaviors such as being depressed or withdrawn were more likely to commit crimes as adults, while boys who showed the same characteristics were less likely to.

The link between , and among children, has been documented in numerous studies. But less is known about how gender relates to problem behaviors among abused children and whether the likelihood of crime linked to those behaviors might vary between boys and girls.

In particular, the connection between internalized behavior and crime has been largely unexamined, said co-author , a 91Ě˝»¨social work professor.

“There’s this sense that kids who are depressed and withdrawn and tend to isolate themselves from other people aren’t necessarily at risk for engaging in criminal behavior later on,” Herrenkohl said.

Children who are abused, particularly girls, are at risk of becoming repeat victims of violence throughout their lives. Herrenkohl said abused girls who repeatedly internalize their feelings may eventually reach a threshold at which their repressed emotions turn outward and aggressive — perhaps pushing them to the point of criminal behavior.

Child abuse can be also predictive of relationship violence, Herrenkohl said, and women who were abused early in life may fall into relationships in which violence or criminal behavior is the norm.

“A woman who has been abused may end up partnering with somebody who is involved in criminal behavior, and that reinforcement in the context of that relationship might drive her to manifest other types of behaviors that wouldn’t necessarily have come to the fore,” he said.

The findings, lead author Hyunzee Jung said, underscore the risk that problem signs in abused girls could be overlooked.

“People might think that they’re just quiet girls and not causing any problems. Those internalizing behaviors really need to be paid attention to,” said Jung, a research scientist at the UW-based .

While previous studies have found that abused boys and girls exhibit different problem behaviors, the 91Ě˝»¨study found that abuse is linked to both externalizing and internalizing behaviors at elementary school age, regardless of gender. And while other research has pointed to adolescent problem behavior as a predictor of later crime, the 91Ě˝»¨study found no such connection. Instead, the research concludes that problem behavior in abused children of elementary school age was a stronger determinant of longer-term criminal behavior.

“We tend to think that adolescents who are involved in delinquent behavior have a higher risk for adult crime. And in some cases, that’s true, but we really need to go back and look at what their childhoods were like,” Jung said.

The findings stem from a that started in 1976 and tracked abused children in two Pennsylvania counties from as young as 18 months to around age 18, in three separate assessments. The children’s parents were asked about their children’s behavior and about the physical and emotional disciplining strategies they used, which ranged from threatening to send the child away to slapping and kicking. The children were also surveyed in adolescence about their own behaviors.

In 2010, the 91Ě˝»¨researchers tracked down about 80 percent of the study’s original participants, then 36 years old on average, and asked whether they had committed any crimes in the previous year.

Slightly more than one-quarter said they had; the most common offenses were hitting someone or threatening to do so, and stealing money or other items from family members. The researchers then compared the self-reported crime data to the earlier reports of abuse and related problem behaviors to reach their conclusions.

Herrenkohl said the findings, and the mixed results from prior studies, underscore the need for additional research to unravel the complex trajectory from problem behaviors in abused children to adult crime, particularly as they differ between genders. Deeper knowledge of that path could lead to more targeted and effective interventions, he said.

“We need to find ways of identifying these kids and wrapping services around them and their families so that we’re reducing the likelihood that their behaviors are going to worsen over time,” he said.

The National Institute of Justice and the National Institute of Child Health and Human Development provided funding for the research. Other co-authors are Martie Skinner at the Social Development Research Group, Jungeun Olivia Lee at the University of Southern California, and Sheryl Hemphill and Jessica Heerde at Australian Catholic University.

“We find that children who were involved in child welfare services are at high risk of adult crimes, but once we accounted for childhood socioeconomic status and later marital status and education, many of those effects went away,” said co-author Todd Herrenkohl, an investigator at the , part of the UW’s School of Social Work.

“So the causal relationship isn’t necessarily there in our data.”

In other words, child abuse and neglect don’t always set up children for a life of crime.

The paper was recently published in the .

The data come from the Lehigh Longitudinal Study, which began in the mid-1970s. Participants were recruited from child welfare abuse and protective service programs, as well as daycares, private nursery programs and Head Start classrooms. Children came from diverse economic backgrounds.

As adults, participants were asked about 29 kinds of criminal offenses ranging in seriousness, including fighting, involvement with gangs, forcing sexual relations, theft and dealing drugs.

In this study of 357 adults, 79 percent who had been maltreated as children reported committing at least one crime in their lifetime, compared with 66 percent of children who had not been maltreated. About 46 percent of the maltreatment group had been arrested at some point, compared with 30 percent in the comparison group.

Lead author Hyunzee Jung, a 91Ě˝»¨research scientist, said researchers were not surprised to find a link between officially recorded child maltreatment and later crime. But what was surprising was that that link faded after they took into account other variables: socioeconomic status, gender, race, marital status, age and education. Poverty especially is a major predictor of crime and incarceration. The data lead researchers to believe it’s a combination of factors, not just maltreatment, which leads to later involvement in crime.

The researchers found that graduating from high school or earning a GED and being married were protective factors that reduced a child maltreatment victim’s later involvement with crime. A high school diploma reduced the odds of arrest by 53 percent and odds of incarceration by 66 percent. Being married lowered the odds of crime within the past year by 40 percent and the odds of ever being convicted of a crime by 58 percent. Those who had been convicted of crimes were convicted fewer times if they were married.

“These later life experiences – graduating high school, getting married – can still have a very profound effect on an individual’s life,” Herrenkohl said. “Many individuals, through their own determination and the support of others, find a way to move beyond adverse experiences and appear to go on to lead productive lives.”

He said being married appears to be a consistent protective factor for a variety of outcomes, demonstrating the potential for strong, committed relationships to change lives for the better.

“It’s an indication of the power of positive relationships, of positive bonding,” he said. “We all benefit from having individuals in our lives who can provide us with guidance and support. These results show that strong adult relationships can make a difference for those who were not treated well as children.”

Co-authors are and at the 91Ě˝»¨ and at the University of Montana. The study was funded by the National Institute of Child Health and Human Development and the National Institute of Justice.

“In other words, people are as likely to smoke whether or not they were sexually or physically abused, but they’re inclined to smoke more if they were abused and have a history of smoking,” said , a professor in the 91Ě˝»¨School of Social Work.

The paper is published online in the .

Herrenkohl and co-authors probed the Lehigh Longitudinal Study, which began in the mid-1970s. Participants were recruited from child welfare abuse and protective service programs, as well as day care programs, private nursery programs and Head Start classrooms in Eastern Pennsylvania.

91Ě˝»¨researchers looked specifically for any connection between physical or sexual abuse and adolescent and adult smoking. They found that boys who had experienced either type of abuse and were smokers, smoked more than those who hadn’t been abused as a child. For girls who smoked, only those who had been sexually abused smoked more as adolescents. That frequency of adolescent smoking by both girls and boys, in turn, led to increased smoking in adulthood, especially among women.

Lead author , a doctoral candidate in social work, found the difference between boys and girls to be one that requires more study.

“There may be other factors at work that we need to disentangle,” she said. “I think the big ‘Aha’ finding is the one on gender differences. Hopefully this will encourage other researchers to look at gender differences in smoking among teens and adults.”

In the Lehigh study, slightly more than 50 percent of the participants said they had smoked in adolescence – that’s about five times the national average for children ages 12-17. Fifty-seven percent of males and 44 percent of females reported smoking in adolescence. Researchers said they don’t know why the rate of smoking was so high in this study. Herrenkohl theorizes that the reasons could have been socioeconomic, geographical, or the fact that participants in this study were already at relatively high risk.

When study participants were evaluated as adults, 49 percent reported smoking in the past year (at nearly equal rates for men and women).

Kristman-Valente said what is of great concern is the fact that so many women who were abused as children were smoking while raising children, and that women who smoke frequently also are less successful in smoking cessation programs.

Since tobacco use often begins in adolescence, researchers say it’s important that public policies are in place to try to prevent kids from lighting up a cigarette in the first place.

“Early adversity can persist throughout a person’s life, so early intervention or prevention of child abuse can potentially lead to long-term public health benefits,” Kristman-Valente said. “I hope our findings encourage more focus on the connection between child maltreatment and smoking in particular. Not many people look at this consequence, even though smoking is the number one cause of preventable death in the U.S.”

The other co-author is , a 91Ě˝»¨research assistant professor in social work. The research was funded by the , , the , and the .

###

For more information, contact Kristman-Valente at ankv@uw.edu, or Herrenkohl at 206-221-7873 or tih@uw.edu.