Early Childhood Antisocial Behavioral Factors which predict Conduct Disorder and Substance Dependence in adolescence:

helpful hints for high school teachers/counselors.

Ais K.W. Murray, M.S. Health Administration, B.S. Psychobiology

Conduct Disorder can be established as early as age seven (7) and as late as one's twenties (20s). Individuals with CD are usually classified as troubled youth by the school system. However, they will most likely have a pattern of antisocial behavior which is not limited to the school environment.  

Individuals with CD may exhibit intra (poor self-esteem) and interpersonal (innability to form intimate relationships) problems. They often take more risks than their peers. Males and females with CD often exhibit different symptoms. Males are more likely to acquire CD than females. 

Genetics and environment may both contribute to a predisposition for an individual to develop CD. Behavioral indicators as early as kindergarten have been shown to have predictive value. Aggression, inattention, hyperactivity, impulsivity, poor school performance, and delayed skills development are predictive of future delinquent behavior. In addition, a bad homelife is also predictive of CD. In general, an earlier age of onset and more antisocial symptoms exhibited predict a worse prognosis.  

Substance Dependence can effect children and adolescents. It is a maladaptive pattern of substance use which may results in compulsive drug-taking behavior and subsequent consequences associated with dependence. Suicide and aggression are associated with SD. Like CD, males are more likely than females to develop SD. 

Both genetic and environmental influences have been shown for alcohol dependence. However, other drugs have not shown conclusive evidence for genetic and environmental contributions to a predisposition for dependence. Personal traits as early as three years of age have shown predictive value for later substance use. Aggressive behavior, negativity and frequent changes or swings in emotion or mood are the most common predictive traits. 

CD increases in adolescence. Nearly half of the individuals with CD are limited only to their adolescent years and tend to grow out of their antisocial behaviors. Those that do change often persist into adulthood. Adolescence is a time when young men and women are trying to become young adults. While milder forms of delinquency and substance experimentation are common at this stage of life, extreme levels of delinquency (CD) or substance use (SD) require intervention.  

CD and SD co-occur quite often with estimates ranging from forty to eighty percent. CD precedes SD in most cases. CD/SD is also associated with criminal behavior and self harm, including suicide.  

School professionals involved with CD prevention programs should take note of the successes in the field. The developmental framework provides a focus on multiple causal factors/contexts and more success in evaluation. Effective prevention programs may be able to reduce the prevalence of new cases, delay the onset of problem behaviors, or decrease the severity and chronicity of antisocial behaviors.  

Effective intervention/treatment is also being realized in addressing antisocial children and adolescents. A cognitive-behavioral, problemsolving skills training program has shown success with seven to thirteen year olds in reducing aggressive behavior. Psychopharmacological methods including the administration of lithium and haloperidol have also shown an effect on the same age group in reducing aggressive, hyperactive, hostile and unresponsive behavior. Long-term supportive and socializing environments have shown success as well by decreasing behavioral problems and future antisocial activity (i.e. re-arrests). The most effective techniques appears to be Multi Systemic Treatment and long-term supportive environments. 

Introduction: 

This paper provides information for school professionals to help them identify the symptoms associated with Conduct Disorder (CD) and Substance Dependence (SD) and to act appropriately to initiate intervention. The paper includes: 

· a presentation of the criteria necessary for a diagnosis of CD and SD. 

· early childhood indicators which predict future CD and SD for adolescents 

· information on CD and SD and CD/SD co-occuring in adolescence 

· a discussion of Adolescence-Limited vs Life-Course-Persistent Antisocial Behavior  

· a discussion of prevention, intervention and treatment in regards to CD 

Many of the behaviors identified in this article can result from multiple causes. CD and SD, as are all mental health disorders, are diagnosed from symptomatic behaviors which deviate from the norm and persist for a minimum period of time. Therefore, it is best not to generalize the items listed in the article to the population as a whole. CD and SD can be very debilitating for the individual, their family members and friends. It is important to identify problematic behaviors related to CD and SD and to initiate intervention to reduce the devastating effects which may result.  

Conduct Disorder: 

Conduct Disorder is characterized by a repetitive and persistent pattern of behavior in which the basic rights of others or major societal norms are violated. The Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (DSM IV) categorizes four groups of behavior for CD: aggressive, nonaggressive, deceitfulness or theft, and serious violation of rules.  

Individuals with CD often initiate aggressive behavior towards other people or animals. Aggression can take the form of bullying, threatening, initiation of fights, use of weapons, and/or physical violence (rape, assault, or homicide). Deliberate destruction of other's property is an example of nonaggressive behavior (arson and vandalism). Deceitfulness or theft includes frequent lying, breaking and entering, and stealing. Serious rule violations include a pattern of disobeying parental curfew, truancy from school, and running away and staying away overnight. Individuals with CD usually exhibit the above behaviors in multiple settings (home, school, and community).  

Individuals with CD may exhibit common associated features. They may have little regard for the feelings or well being of others. When confronted they may readily try to blame others for their own misdeeds. They may exhibit poor self-esteem, frustration, irritability and temper outbursts. CD is associated with an early onset of drug use, early sexual behavior, recklessness, learning and communicative disorders, and increased suicidality. Attention Deficit/Hyperactivity Disorder (ADHD), Anxiety Disorders, and Mood Disorders may also be associated with CD.  

CD shows differential symptomology between males and females. Males with a diagnosis of CD frequently exhibit fighting, stealing, vandalism, and school discipline problems. Females with a diagnosis of CD are more likely to exhibit lying, truancy, running away, substance use and prostitution.  

Childhood CD, in particular, but also late onset CD is much more common in males. Prevalence rates of CD for males under the age of eighteen (18) range from 6% to 16%, whereas prevalence rates for females tend to be in the range of 2% to 9%. Prevalence rates tend to be higher in urban settings, as well. 

Individuals may possess a predisposition for conduct disorder and show early symptomology. Estimates from twin and adoption studies provide evidence that CD has both genetic and environmental components. Parental rejection and neglect, inconsistent child-rearing practices with harsh discipline, physical or sexual abuse, lack of supervision, early institutional living, frequent changes of caregivers, large family size, an adoptive parent with Anti-Social Personality Disorder (ASPD) and association with a delinquent peer group are all environmental factors which can predispose a child to the acquisition of CD. Children with a biological parent with ASPD, Alcohol Dependence, Mood Disorders, Schizophrenia, Attention-Deficit/Hyperactivity Disorder (ADHD), or CD, and children who have a sibling with CD are all at increased risk to develop CD versus those children without any of the above mentioned familial psychopathologies. 

Substance Dependence: 

Substance Dependence is a diagnostic term applied to all drugs including alcohol and tobacco. SD is characterized as a maladaptive pattern of substance use leading to clinically significant impairment or distress. Individuals with SD exhibit loss of control over their drug use. They may take larger amounts or use for a longer period of time then intended. They may have previous unsuccessful attempts to decrease or discontinue use. SD may include tolerance, withdrawal, and a pattern of compulsive use. Previously important activities are given up or reduced because of the substance. The substance use is continued despite persistent or recurrent problems caused by the substance. SD frequently leads to deterioration of general health resulting from improper diets and inadequate personal hygiene. Individuals with SD often take more than one substance, simultaneously, or sequentially, and may be intoxicated with one substance while withdrawing from another.  

According to DSM IV, the course of SD is usually chronic, lasting years, with periods of partial and full remission. On occasion, spontaneous, long-term remissions my take place. For example, twenty (20) percent of individuals with alcohol dependence become permanently abstinent, usually following a severe life stress (legal or medical in nature).  

Associated features of SD include suicide and aggression. According to the DSM IV, many individuals with SD commit suicide, often when experiencing a substance-induced mood disorder. Note that this statistic is for all individuals diagnosed with SD, of which, the majority are adults. Substance dependence is also associated with physical aggression and criminal activity.  

SD also shows differences between males and females. Males outnumber females in the prevalence of SD, but the ratios vary with class of substance.  

SD can occur at any age, as long as the criteria and time (minimum 12 months) have been met, but typically initial onset for most drug dependence is in the 20s, 30s, and 40s. SD appears to aggregate in families, but some of this effect may be explained by concurrent distribution of ASPD, which also runs in families and which may predispose individuals to the development of SD. Both genes and environment appear to contribute to a predisposition for alcohol dependence. However, other drugs have not shown conclusive evidence for genetic and environmental contributions to a predisposition for dependence.  

Early Childhood Predictors of CD: 

Childhood aggression is highly predictive of later juvenile delinquency and adult criminality. The more variety, frequency and seriousness of early aggressive behavior predict a greater risk of antisocial and criminal behavior continuing into late adolescence and adulthood. The particular acts which are characterized by delinquency are those that warrant police and court action. This legal term is very highly correlated with diagnostic criteria utilized for CD (see subsequent section on delinquency in this article).  

ADHD diagnostic categorical criteria like inattention, hyperactivity, and impulsivity also predict delinquency. In one study, Tremblay and colleagues found that boys with high impulsivity ratings in Kindergarten were more at risk for early onset of frequent delinquent behavior than those with low impulsivity ratings. In addition, the relative level of delinquency among groups based upon their Kindergarten personality measures was stable at ages eleven (11) and thirteen (13). Please note that most studies on childhood behavioral problems related to CD were done with males. As noted earlier, males diagnosed with CD greatly outnumber females.  

Delayed skills development and poor school performance may also be predictive of CD. As noted in the associative features section of CD above, learning disorders and communicative disorders (particularly reading and verbal skills) may be associated with CD.  

Early home atmosphere is strongly associated with the development of delinquency. Experience of chronic conflict or abuse/neglect was associated with delinquent behavior in youth. In addition, violent delinquents were found to have cold, harsh, cruel, disharmonious, poorly supervising, and criminal parents.  

Early Childhood Predictors of SD: 

Traits at three (3) years or age which predicted substance use by age 14 include: 

· aggression 

· domination 

· blaming others 

· teasing others 

· expressing negative feelings 

· being emotionally labile 

· being sulky or whining 

Traits at three (3) years of age which are negatively associated with later drug use include:  

· cooperation 

· eagerness to please 

· being physically cautious 

· neat and orderly 

· shyness and reserve 

Children who have peers who use substances, and exhibit school problems, rebelliousness, novelty seeking and risk taking behavior are at a greater risk to use substances. Adult men with alcohol problems had been described in their youth as being hot-headed, cutting up, and having strong personalities. 

Adolescent CD: 

Once any individual, especially a male, reaches adolescence, delinquency almost becomes the norm. Moffit cited studies that showed one third (1/3) of males are arrested during their lifetime for a serious criminal offense and four fifths (4/5) have police contact for some minor infringement. Most police contacts happen in adolescence. Moffit conducted a study and found that only seven (7) percent of the eighteen (18) year old males interviewed denied all delinquent activities.  

The extreme level of delinquency (CD) includes risk-taking and rule breaking. CD affects 4-10% of Teens, with a three (3) to one (1) prevalence ratio for males versus females. Nearly half of the highly antisocial children recover, while the rest progress onto adult ASPD. A steep incline in antisocial behavior between ages seven (7) to seventeen (17) is mirrored by a steep decline in antisocial behavior between ages seventeen (17) to thirty (30). When official rates of crime are plotted against age, the rates for both prevalence and incidence of offending appear highest during adolescence. The rates peak sharply at age seventeen (17) and drop precipitously into young adulthood.  

The individuals who recover from adolescent antisocial/behavioral problems in their late teens and early twenties are termed Adolescence-Limited Antisocial Behavior types. They are categorized by the following: 

· Temporal instability - abrupt stop in delinquency or periods without delinquency (yrs) 

· Situational instability - delinquency does not cross situations (home, school, etc.) 

According to the DSM IV, individuals with antisocial behaviors beginning after age ten (10) are less likely to display aggressive behaviors and are more likely to have normative peer relationships. 

The adolescents who do not recover from adolescent antisocial/behavioral problems are termed Life-Course-Persistent Antisocial Behavior types. They are categorized by the following:  

· Early age of onset  

· Temporal stability - behavioral pattern is consistent from childhood into adulthood. 

· Situational stability - behaviors are consistent across situations (home, school, etc.) 

Individuals who display at least one criteria for CD by age ten (10) frequently display physical aggression towards others and have disturbed peer relationships. These individuals may have been diagnosed Oppositional Defiant during early childhood and usually have symptoms that meet the diagnosis for CD prior to puberty. Early onset predicts a worse prognosis and an increased risk for ASPD and SD.  

Moffit proposes that during adolescence, adolescence-limiteds are mimicking the life-course-persistents. Delinquency shifts from an individual psychopathology in childhood to a normative group behavior during adolescence, and then back to individual psychopathology in adulthood. Moffit suggests that every curfew violated, car stolen, drug taken, and baby conceived during adolescence is a statement of personal independence and thus a reinforcer for delinquent involvement.  

It has been proposed that Teens transitioning from middle (16) to late (18) adolescence are attempting to form their personal identity, and adjust to their adult social role (including separate from the family of origin). During this time period, the peer group is the primary source of identification and a powerful source of influence on Teen behavior. It appears that adolescent-limited antisocial behavioral types are able to make the transition to adulthood, while life-course-persistent antisocial behavior types cannot.  

Moffit also cites examples of adolescents who never committed any antisocial behavior as having delayed puberty, respected adult roles (often cultural or religious), environments that limit exposure to delinquency, and/or personal characteristics that limit them from delinquent peer networks ( nervous, withdrawn, isolated, and lacking interpersonal skills). 

Adolescent SD: 

Drug experimentation is common in adolescence, but 10-30 percent of teens experience problems that significantly impact later life development. Adolescent SD is associated with academic, legal, social and familial problems.  

Adolescent CD/SD: 

CD/SD tend to co-occur. When SD begins in early adolescence, it is often associated with CD and failure to complete school. Brown cited estimates of CD and substance abuse in forty (40) to sixty (60) percent of clinical cases, while the Clark study, found a co-occurance rate of eighty one (81) percent. Young cited a Hubbard et al., 1985 study with seventy (70) percent of the boys and sixty (60) percent of the girls in substance treatment as delinquent. Myers found that seventy (70) percent of the substance treatment hospitalized teens had CD. 

CD often proceeds SD. The Young study, in particular, found that most of the sixty boys in the study had one or more conduct disorder symptoms before initial drug use and met the threshold of three lifetime conduct disorder symptoms shortly before they began regular drug use. All boys had CD while 53 out of 57 who completed an in treatment clinical instrument had SD. The boys in the Young study typically reported: 

· Alcohol use began at age 11, while weekly use began at the age of 13.  

· Drug use, other than alcohol, began at 13.7 years, while regular use began at 14.5 yrs.  

· Multiple drug use, other than alcohol, with an average of 3.3 substances used. 

· Dependence on at least one (96%) or two (75%) substances, in addition to alcohol.  

Among adolescents with CD/SD, males were more likely to have Attention Deficit Disorder in addition to CD/SD, while females exhibited Emotional Disorders (depression, anxiety) more often in addition to CD/SD. 

Those adolescents with severe cases of CD or SD were more likely to inflict self injury, including attempt suicide.  

Myers proposes that CD not only proceeds, but also causes SD. Myers wrote that delinquent youth have disturbed interpersonal relationships, associate with delinquent peers, have disrupted family environments, and have symptoms of hyperactivity. These factors tend to promote stress that tax the individual's coping mechanisms. The overwhelming stress increases attraction to drugs as a method of coping. Peer pressure and minimization of risk lead to high risk taking (use of more or dangerous types of drugs). The drugs are reinforcing because they alleviate the stress. This pattern continues until the individual develops dependence upon the drug(s). Self-critical statement questions and interpersonal difficulties, showing decreased coping skills, were correlated with CD. 

Delinquency: 

Delinquency is often associated with troublesome behaviors and with personal and social difficulties. Problem behaviors which are common reasons for police/court action include aggression, acquisitive, drugs, vehicular, avoidance, and others. The aggression includes disorderly, vandalism, verbal assault or threat, fighting and use of obscenity. Acquisitive behaviors include breaking and entering, shoplifting, larceny, robbery, and theft. Drug behavior includes possession, sale or use. Vehicular behaviors include unauthorized use, tampering, and can include traffic violations with motorbikes. Avoidance behaviors include runaway, truancy, and escape. The other category includes arson, gambling, loitering, use or possession of weapon, misuse of telephone, and curfew violation (to name a few).  

CD Prevention: 

Major theoretical advances have been applied to the prediction and prevention of antisocial behavior. First, a developmental approach provides a framework which emphasizes the importance of the interplay of normal developmental tasks and the interaction of a multiplicity of causal factors. As a result, prevention can occur through the reduction of the prevalence of new cases, delayed onset of problem behaviors, or a decrease in severity and chronicity of antisocial behaviors. Second, studies are being conducted that investigate causal factors and related interventions that emphasize multiple influences. The multiple influences include individual characteristics (gender and ethnicity), close interpersonal relationships, social settings, and societal constraints. In particular, an emphasis is being placed upon how antisocial behavior unfolds in specific cultural and community contexts. Third, the accumulating knowledge provided by the developmental framework has promoted greater sophistication in sampling, analytic methods, and evaluation of external validity.  

Risk factors for antisocial behavior are social-cognitive factors including: 

· family relationship characteristics  

· parenting practices 

· information-processing skills 

· individual beliefs 

· cognitive skills 

· peer relationships 

Protective factors include: 

· communication skills 

· family management of external stress 

· changing schools norms toward reinforcement for reduced aggression 

· prosocial peer networks 

Prevention is being measured in terms of its effect upon attitudes, beliefs, and social information-processing skills in addition to behavioral differences. Risk models have moved from identifying general dispositional risk factors to emphasizing person-environmental interactions.  

Controlled prevention trials are needed to evaluate the efficacy of strategies in changing the prevalence or course of antisocial behavior. The trials should also examine mediators and moderators of change. Prevention effectiveness needs to look at the understanding of how effects vary by setting and population and under what conditions benefits can be realized. The ultimate effectiveness depends upon usefulness in real world settings, the transportability of the essential components to other settings, their compatibility with existing service delivery systems, and the acceptance by the targeted community.  

There is a need for multicomponent, multicontext prevention programs. The most promising programs focus on changing the individual, the influence of close interpersonal relationships, and the contexts of development. It appears that preventions aimed at multiple risk factors are more effective and seem more cost efficient. Whether universal interventions are cost efficient compared to focused interventions for selected populations has not been empirically evaluated.  

Intervention: 

One key to effective intervention is to start with the appropriate participants. This includes an adolescent who clearly has CD or SD. The only way to determine if the adolescent has a problem is to have them evaluated. It is best to choose a professional that is easily accessible to the adolescent and his/her family, and ideally one who has worked with the family previously. This will have a two fold effect. The access barrier can be reduced and the family involvement barriers might be reduced. The family may be a challenge to attract, motivate and maintain their involvement. As described earlier, an adolescent with CD or SD often times has parents or siblings who are also suffering from antisocial behavior or substance-related disorders.  

The second key is to start at the appropriate level. Intervention involves an attempt to apply the least restrictive approach in a manner that is productive enough to overcome the problem. Intervention can lead to treatment and usually does so by a pre-determined path. The path progresses from little involvement by others in community settings like the home or school (level one). The second level (two) involves more professional involvement in a setting more removed from the community, like a day treatment program or a foster parent program. The final and most restrictive level (three) is provided in a treatment setting like a residential hospital. The level of professional involvement is very high and the individual is removed from the community.  

What ever level of intervention is currently being utilized, the adolescent will need professional support for healthy functioning in the community or for transitioning back into the community. School teachers and counselors can greatly assist the adolescent in the school environment. This professional support may be required for a long period of time (years). 

While there is no proven intervention that prevents CD or reverses the course of CD in children, there are some psychosocial and pharmacological treatments which have been shown to reduce the symptoms of CD. A problemsolving skills training program (cognitive-behavioral) showed significant improvement for aggressive children ages seven (7) to thirteen (13) in regards to their behavior at home and school. Also, hyperactivity, aggression, hostility, and unresponsiveness improved for aggressive children with CD ages six (6) to thirteen (13) years of age with the administration of lithium or haloperidol. In addition, structured environments with immediate rewards and punishments and few peer inducements for problem behavior have also been shown to reduce antisocial behavior. 

It has been found that delinquency can be reduced by early childhood intervention programs that promote competence across multiple systems. Multi Systemic Treatment (MST) is one such program. It utilizes interventions that are present-focused and action oriented. It directly addresses intrapersonal (cognitive) and systemic (family, peer and school) factors that are known to be associated with adolescent antisocial behavior. It utilizes individualized and highly flexible techniques, especially during interviewing and service delivery. For example, sessions are usually held in the family's home and in community locations (school and recreation centers). The sessions are scheduled at convenient times and are time limited.  

MST focuses on youth's social systems throughout treatment and intervenes directly in those systems. MST attempts to empower the youth and their family members within a context of supportive skills building. MST also emphasizes the need for behavioral changes across all key systems linked with antisocial behavior. The success of MST may be linked to its explicit focus on key factors associated with delinquency including: behavior problems, parental disturbances, problematic family relations, association of deviant peers and poor school performance. MST has been shown to decrease psychiatric symptomology in mothers and fathers, while also decreasing behavioral problems and re-arrest histories in youth participants.  

Another intervention technique that has shown success is Long-term Supportive Environments. One type is focused on the individual in their community (level one). It involves a strong emphasis on education for the individual, community functioning, job training, and sustaining healthy relationships. The individual is taught how to best relate with their parents despite the parents emotional, drug, and psychological problems. A second type is focused on the individual in a foster family environment (level two). It involves selecting caring supportive parents and providing training, consultation, financial assistance and monitoring. The focus of this approach is to reward appropriate behaviors and deter delinquent behaviors.  

Primary Bibliography: 

American Psychiatric Association: Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Washington, DC, American Psychiatric Association, 1994. 

Borduin CM, Mann BJ, Cone LT, Henggeler SW, Fucci BR, Blaske DM, Williams RA, Multisystemic treatment of serious juvenile offenders: Long-term prevention of criminality and violence, Journal of Consulting Clinical Psychology, 63:569-578, 1995. 

Brown SA, Conduct Disorder among adolescent alcoholics and drug abusers, Journal of Studies on Alcohol, 57(3):314- , 1996. 

Crowley TJ, and Riggs PD, Adolescent substance use disorder with conduct disorder, and comorbid conditions, Adolescent Drug Abuse: Clinical Assessment and Therapeutic Interventions, National Institute on Drug Abuse Research monograph Series #156 (NIH Publication No. 95-3908):49-111, 1995.  

Moffit TE, Adolescence-limited and life-course-persistent antisocial behavior: A developmental taxonomy. Psychology Review 100(4):674-701, 1993. 

Tolan, PH, Guerra, NG, and Kendall, PC. A developmental-ecological perspective on antisocial behavior in children and adolescents: toward a unified risk and intervention framework, Journal of Consulting and Clinical Psychology, 63 (4): 579-584, 1995. 

Young SE, Mikulich SK, Goodwin MV, Hardy J, Martin CL, Zoccolillo MS, Crowley TJ. Treated delinquent boys' substance use: onset, pattern, relationship to conduct and mood disorders, Drug Alcohol Dependence 37:149-152, 1995. 

Secondary Bibliography: 

Boyle MH, Offord DR. Psychiatric disorder and substance use in adolescence, Canadian Journal of Psychiatry, 36:699-705, 1991. 

Clark, DB, Jacob, RG and Mezzich, A. Anxiety and conduct disorder in early onset alcoholism. Annuals of New York Academy of Science 708: 181-186, 1994. 

Cohen, HL. BPLAY-A Community Support System, Phase One, Chapter 8 in Analysis of Delinquency and Aggression, 1976, Lawrence Erlbaum Associates, Hillsdale, New Jersey. 

Grove WM, Eckert ED, Heston L, Bouchard TJ, Segal N, Lykken DT. Heritability of substance abuse and antisocial behavior: a study of monozygotic twins reared apart, Biological Psychiatry, 27:1293-1304, 1990. 

Klug, W.S. and Cummings, M.R. Essentials of Genetics, Second Edition, 1996, Prentice Hall, Upper Saddle River, New Jersey. 

Myers MG. Preadolescent conduct disorder predict relapse and progression of addiction for adolescent alcoholics and drug abusers, Alcoholism: Clinical Experiment and Research, 19(6):1528- , 1995.  

Regier DA, Farmer ME, Rae DS, Locke BZ, Keith SJ, Judd LL, Goodwin FK. Comorbidity of mental disorders with alcohol and other drug abuse: results from the Epidemiologic Catchment Area (ECA) study, JAMA 264(19):2511-2518, 1990. 

O'Donnell, J, Hawkins, JD and Abbott, RD. Predicting serious delinquency and substance use among aggressive boys. Journal of Consulting and Clinical Psychology 63(4): 529-537, 1995. 

Robins LN. Sturdy childhood predictors of adult antisocial behavior: Replications from longitudinal studies. Psychological Medicine 8:611-622, 1978. 

Tremblay RE, Pihl RO, Vitaro F, Dobkin PL. Predicting early onset of male antisocial behavior from preschool behavior, Archives of General Psychiatry 51:732-739, 1994. 

Wolf MW, Braukmann CJ, Ramp KA. Serious delinquent behavior as part of a significantly handicapping condition: Cures and supportive environments. Journal of Applied Behavior Analysis, 20:347-359, 1987.