Youth Violence Prevention Conference Explores Risk Factors, Interventions

December 1, 2004

Although a lot remains to be learned, much is already known about the early childhood characteristics that predispose children to antisocial behavior and violence in adolescence and early adulthood, an expert in psychopathology said recently at a National Institutes of Health conference.

Psychiatric Times

December 2004

Vol. XXI

Issue 14

Although a lot remains to be learned, much is already known about the early childhood characteristics that predispose children to antisocial behavior and violence in adolescence and early adulthood, an expert in psychopathology said recently at a National Institutes of Health conference.

Benjamin B. Lahey, Ph.D., professor of psychiatry and chief of psychology at the University of Chicago, was one of 21 investigators and clinicians involved in violence prevention and adolescent health issues who discussed research studies and findings with an independent panel at the NIH's State-of-the-Science Conference on Preventing Violence and Related Health-Risking Social Behaviors in Adolescents. The conference was co-sponsored by the National Institute of Mental Health, the Office of Medical Applications of Research and several other federal agencies.

A 13-member panel that comprised experts in psychiatry, pediatrics, behavioral health, economics, juvenile justice and pediatrics, among others, was tasked with evaluating the current research on youth violence and recommending areas for future investigation. Robert Johnson, M.D., associate professor of psychiatry and professor and chair of pediatrics at New Jersey Medical School, served as panel chair.

Early Predictors

"Clearly, different experiences and circumstances are profoundly important in determining which children will grow up to be violent, but so are some characteristics of those individuals in childhood," Lahey said. Those characteristics include male gender, lower intelligence, delayed language development, low resting heart rate, specific personality traits, and gene-environment interactions and correlations.

Most studies of personality and violence are cross-sectional, but some preliminary studies have suggested that a child's temperament robustly predicts later antisocial behavior and violence, according to Lahey. Antisocial youth tend to respond to threat and frustration with intense negative emotions (neuroticism) and are unlikely to be sympathetic to others and to respect rules. They also tend to be unreliable, careless and bold (low conscientiousness). Conversely, shy, inhibited children are at low risk for antisocial behavior.

Work by Terrie Moffitt, Ph.D., professor of psychology at the University of Wisconsin, and colleagues, Lahey said, is largely responsible for showing us that individuals who have shown low intelligence in childhood are at increased risk for antisocial behavior and violence.

"It is possible that particular facets of intelligence are more important than global measures of intelligence," Lahey said. "For example, executive functions may be particularly important. There is emerging but very strong evidence that slow language development is a good predictor of antisocial behavior and violence."

Lahey added that gender differences in antisocial behavior and violence may be explained by gender differences in language development (development being slower in males than females). Other cognitive factors that are correlates and perhaps predictors of violence, he said, are positive attitudes toward antisocial behavior and violence, and cognitive processing variables, such as attributing hostile intent.

With regard to early aggression and childhood mental disorders, Lahey explained in a conference syllabus that "at least among males, oppositional defiant disorder (ODD), but probably not attention-deficit/hyperactivity disorder (ADHD), increases the risk for childhood conduct disorder (CD), and CD and aggression robustly predict antisocial behavior and violence."

At the biological level of analysis, clear evidence exists that there are genetic influences, Lahey said, explaining that aggression and violence are at least moderately heritable across all sociocultural groups.

It is likely that genetic influences operate through both gene-environment interactions and gene-environment correlations, Lahey added. In the conference syllabus, Lahey explained that two studies suggested adolescent antisocial behavior is related to childhood adversity and alleles of the gene influencing monoamine oxidase-A (MAO-A) activity in a gene-environment interaction (Caspi et al., 2002; Foley et al., 2004). The role of MAO-A is particularly interesting, he said, because it degrades serotonin and other neurotransmitters.

"Metabolites of serotonin in cerebral spinal fluids have been associated with very large effect sizes to a pattern of violence called impulsive violence and particularly in individuals below the age of 30," he told the panel.

At the level of peripheral psychophysiology, resting heart rate is inversely related to antisocial behavior and predicts future antisocial behavior and violence, according to Lahey.

"Some longitudinal studies [Farrington, 1997] show low resting heart rate in children predicts violence itself," he said. "Dampened autonomic reactivity (sympathetic activation measured by skin conductance) has also been linked to antisocial behavior and violence, although the evidence is less clear."

Cortisol has been shown in several studies to differentiate children with CD from children without CD, Lahey said. But the direction of the difference has been inconsistent, and there is no clear evidence of a link between cortisol levels with violence itself.

Sex steroids (e.g., testosterone, estradiol and estrogen) have been found in many studies to be related to dominance and aggression, yet the meaning of those relationships is unclear, Lahey said, because there are bidirectional relationships between engaging in aggression and achieving dominance and changes in the sex steroids.

Lahey ended his presentation by warning, "These predisposing factors cannot be equated with destiny. These early behavioral and biological predictors clearly do not predict violence on an individual basis. You cannot measure a child's behavior and say this child will or will not engage in violence; you can only do that at the group level. But I think it is clear that early predictors can be used to identify very high risk groups of children for targeted preventive interventions if the practical and ethical implications of doing so are carefully considered."

Among the other presenters to the panel were psychiatrist Felton Earls, M.D., professor of social medicine and of human behavior and development at Harvard University, and Delbert Elliott, Ph.D., director of the Center for the Study and Prevention of Violence at the University of Colorado.

Prevalence Rates

Earls pointed out that youth surveillance programs, such as the Youth Risk Behavior Survey (YRBS) and Monitoring the Future Study, have shown a "noticeable and significant reduction in the amount of violence" but not in the seriousness of the violence (amount of injury and death occurring).

He cited data from the Centers for Disease Control and Prevention (2004) and YRBS:

Two of the national health objectives for 2010 are to reduce the prevalence of physical fighting among adolescents to ≤32% and to reduce the prevalence of carrying a weapon by adolescents on school property to ≤4.9%.

To examine changes in violence-related behavior among high school students in the United States from 1991 to 2003, the CDC analyzed data from the YRBS. The results can be seen in the Figure.

Surveillance data on victimization are also available, according to Earls, who discussed the 1995 National Survey of Adolescents (Kilpatrick and Saunders, 1997). This national probability sample survey of 4,023 young people (ages 12 to 17) sought to test specific hypotheses illustrating the relationships among serious victimization experiences, the mental health effects of victimization, substance abuse/use and delinquent behavior in adolescents. Lifetime prevalence of physical assault was 17.4% of the sample; sexual assault, 8.1%; one or more incidents of physically abusive punishment, 9.4%; and having personally witnessed one or more serious incidents of violence, 39.4%. In some demographic subpopulations, more than half had witnessed violence.

In the survey, researchers found that exposure to interpersonal violence increased the risk of posttraumatic stress disorder, major depressive episode and substance abuse/dependence after controlling for demographic factors and family substance use problems (Kilpatrick et al., 2003).

Earls went on to point out that for detailed information on psychiatric disorders in children and adolescents and comorbidities, he relies on data from the Ontario Child Health Study, which looked at the six-month prevalence of four child psychiatric disorders (CD, hyperactivity, emotional disorder and somatization) (Offord et al., 1987), and the Great Smoky Mountain study, a randomized multistage study that examined the prevalence and development of psychiatric disorders in 1,420 children and adolescents residing in western North Carolina (Costello et al., 2003).

For Earls, the Great Smoky Mountain study is a landmark study in psychiatric epidemiology, because it examined three-month prevalence of any psychiatric disorder and because it was a cross-sectional study that turned into a longitudinal study. The researchers found that by the time children in the study reached 16 years of age, one in three currently was experiencing or had had at least one psychiatric disorder.

"The Great Smoky Mountain study is important for looking at comorbidity," Earls added. "For example, conduct disorder is comorbid with depression in females; depression and substance abuse disorder co-occur more often in males than females; and the presence of an anxiety disorder inhibits conduct disorder."

Earls pointed out that these studies in circumscribed areas are particularly important because there are no nationally based studies of psychiatric disorders in children. He concluded his presentation with a series of recommendations to advance the field, including the greater use of case histories and innovations in quantitative reasoning and statistical approaches.

Effective Interventions

Elliott emphasized that strong evidence of effectiveness exists for some interventions addressing children and youth across the developmental and risk involvement spectrum. However, "there is a lot of confusion in the field, because different agencies have published lists of what they consider to be model or effective programs, and those lists differ from one another."

To reduce the confusion, Elliott used criteria established by the Working Group for the Federal Collaboration on What Works, an informal group that comprised representatives from the U.S. Department of Justice, U.S. Department of Education, National Institute on Drug Abuse and Center for Substance Abuse Prevention. Programs designed to reduce substance abuse, delinquency or violence were classified by whether or not they met the following criteria:

  • A randomized controlled trial (RCT);

  • Statistically significant positive effect;

  • Effect sustained for at least one year post intervention;

  • At least one external RCT replicating the results;

  • RCTs adequately addressed threats to internal validity, such as questions of selection and quality of measurement; and

  • No known health-compromising side effects associated with the intervention.

Programs were classified "effective" if they met all of those standards. Programs were classified "effective with reservation" if the replication was an internal rather than external evaluation-randomized controlled trial.

Elliott found four programs that met the standards for being effective: Functional Family Therapy (FFT), Multisystemic Therapy (MST), the Incredible Years: Parent, Teacher, and Child Training Series (IYS) and the Life Skills Training program (LST).

The IYS is designed to treat children ages 2 to 8, at risk for and/or presenting with conduct problems (high rates of aggression, defiance, oppositional and impulsive behaviors). The LST is three-year intervention designed to prevent or reduce gateway drug use, such as marijuana or alcohol.

The panel specifically emphasized the effectiveness of FFT and MST, noted for reducing arrests or violence precursors (NIH, 2004). Functional Family Therapy treats high-risk youth and their families using a short-term, family-based prevention and intervention program. Participating youth and families usually attend 12 one-hour sessions over three months, although up to 30 sessions may be available for difficult cases. Program evaluations demonstrated reductions in rearrest rates and in out-of-home placements; these reductions were sustained over four years. In addition, clinical trials demonstrated that FFT can effectively treat adolescents with CD, ODD or disruptive behavior disorder, as well as alcohol and other drug abuse disorders (Center for the Study and Prevention of Violence, 2004).

Multisystemic Therapy provides community-based clinical treatment for violent and chronic juvenile offenders who are at risk for out-of-home placement (NIH, 2004). Included in the average four-month treatment period is approximately 60 hours of therapist-family contact. Case loads are kept low (four to six families), and therapists are available 24 hours per day, seven days per week. Program evaluations demonstrated reductions in long-term rates of rearrest and in out-of-home placements; these results were maintained for nearly four years after treatment ended.

Although Elliott identified 11 programs as "effective with reservation," only six were highlighted in the NIH panel's report: Big Brothers, Big Sisters of America, reduction in hitting; Multidimensional Treatment Foster Care, reduction in incarceration; Nurse-Family Partnership, reductions in arrests and crime; Project Towards No Drug Abuse, reduction in weapon carrying; Promoting Alternative Thinking Strategies, reduction in peer aggression; and Brief Strategic Family Therapy, reductions in CD and socialized aggression. The others mentioned by Elliott were the Midwestern Prevention Project, Guiding Good Choices, Cognitive-Behavioral Therapy, Iowa Strengthening Families Program, and Athletes Training and Learning to Avoid Steroids. Information on many of these programs can be obtained from the Blueprints for Violence Project at: www.colorado.edu/cspv/blueprints.

A number of strategies also appear to be effective, said Elliott, who was senior scientific editor for the U.S. Surgeon General's Report to the Nation on Youth Violence. These strategies include: self-control and social competency programs when cognitive/behavioral methods are employed, individual counseling when used with noninstitutionalized juvenile offenders, behavior modeling and behavior modification, restitution with probation or parole, and the use of multiple services.

What Doesn't Work

Programs that rely on "scare tactics" to prevent children and adolescents from engaging in violent behavior are not only ineffective, but may actually make the problem worse (NIH, 2004):

Such evidence as there is offers no reason to believe that group detention centers, boot camps and other 'get tough' programs do anything more than provide an opportunity for delinquent youth to amplify negative effects on each other.

At the press conference announcing the panel's conference statement, panelist Leon Eisenberg, M.D., a child psychiatrist at Harvard University, warned that many programs take the child out of the family. "Whatever they may do or not do for the child while he's in the institutional setting, [they] leave him completely adrift when the treatment is over. Some of these programs are, frankly, quite dreadful."

In its conference statement, the panel noted that the CDC has reviewed evidence indicating that laws that increase the ease of transferring juveniles to the adult judicial system "are counterproductive and lead to greater violence in the juveniles moving through the adult systems without deterring juveniles in the general population from violent crime."

With regard to violence prevention research, the panel made several recommendations including establishment of a national population-based adolescent violence registry; creation of long-term cohort studies that measure a rich set of risk factors (from the individual to the contextual level) in diverse populations and that are analyzed using state-of-the art statistical methods; development of systematic procedures to adapt established intervention protocols for diverse communities; and creation of more research studies on the cost-effectiveness of interventions to prevent violence.

References

Caspi A, McClay J, Moffitt TE et al. (2002), Role of genotype in the cycle of violence in maltreated children. Science 297(5582):851-854 [see comment].

CDC (2004), Violence-related behaviors among high school students-United States, 1991-2003. Morbidity and Mortality Weekly Report 53(29):651-655.

Center for the Study and Prevention of Violence (2004), Blueprints Model Programs, Functional Family Therapy. Available at: www.colorado.edu/ cspv/blueprints/model/programs/FFT.html. Accessed Oct. 23.

Costello EJ, Mustillo S, Erkanli A et al. (2003), Prevalence and development of psychiatric disorders in childhood and adolescence. Arch Gen Psychiatry 60(8):837-844 [see comment].

Farrington DP (1997), Biosocial bases of aggressive behavior in childhood: resting heart rate, skin conductance orienting, and physique. In: Biosocial Bases of Violence, Raine A, Brennan PA, Farrington DP, Mednick SA, eds. New York: Plenum Press, pp89-105.

Foley DL, Eaves LJ, Wormley B et al. (2004), Childhood adversity, monoamine oxidase a genotype, and risk for conduct disorder. Arch Gen Psychiatry 61(7):738-744.

Kilpatrick DG, Ruggiero KJ, Acierno R et al. (2003), Violence and risk of PTSD, major depression, substance abuse/dependence, and comorbidity: results from the National Survey of Adolescents. J Consult Clin Psychol 71(4):692-700.

Kilpatrick DG, Saunders BE (1997), Prevalence and Consequences of Child Victimization: Results from the National Survey of Adolescents Final Report. Charleston, S.C.: Medical University of South Carolina, National Crime Victims Research and Treatment Center.

NIH (2004), State-of-the-Science Conference Statement, Preventing Violence and Related Health-Risking Social Behaviors in Adolescents. Draft statement, Oct. 15. Available at: http://consensus.nih.gov/ta/023/023youthviolencepostconfintro.htm. Accessed Oct. 15.

Offord DR, Boyle MH, Szatmari P et al. (1987), Ontario Child Health Study. II. Six-month prevalence of disorder and rates of service utilization. Arch Gen Psychiatry 44(9):832-836.