The Diagnostic and Statistical Manual (3rd ed., rev.) (DSM-III-R) diagnosis of conduct disorder assumes that all children who engage in three or more criterion antisocial behaviors for 6 months or more suffer from a mental disorder. It resists all contextual information about a child's developmental history, capacities, strengths and circumstances, and assumes that the antisocial behavior necessarily stems from an underlying disorder. In this review, we use Mark Twain's narrative of the lives of Tom Sawyer and Huckleberry Finn as a point of departure for questioning the reasonableness of this assumption, and for examining normal as well as pathological pathways to antisocial behavior. We begin by reviewing the status of earlier controversies about the mental disorder concept in the service of documenting the impressive progress of the field in conceptualizing disorder. Next, we examine Wakefield's (1992a, 1992b) recently introduced “harmful dysfunction” concept of mental disorder and employ its criteria to evaluate the hypothesis that chronic antisocial behavior in childhood as defined by DSM-III-R is caused by an underlying mental disorder. We also examine some of the difficulties in discriminating between disorder- and nondisorder-based antisocial behavior, and consider issues that warrant attention in future theoretical and empirical work. Finally, we explore the pragmatic rather than scientific basis for DSM-III-R's mental disorder claim and argue that regardless of its status as a mental disorder, this most troubling and harmful behavior syndrome of childhood deserves the intensive interest, concern, and resources of the scientific and public health communities.

A developmental approach to the classification of antisocial behavior is necessary for two reasons. First, although the continuity of antisocial behavior is strong for many individuals, the topography of antisocial behavior changes during the course of development. Second, antisocial behavior apparently develops in at least two separate pathways — child-onset versus adolescent-onset — that differ markedly regarding types of antisocial behavior displayed, persistence, and perhaps etiology. The development of antisocial behavior must also be understood within the context of co-occurring disorders and conditions. Comorbid attention-deficit hyperactivity disorder appears to be associated with greater aggression and a worse prognosis, and comorbid academic underachievement is also associated with a negative course. Emerging evidence also suggests that comorbid anxiety disorder is associated with level of aggression, but the direction of the correlation appears to differ at different ages. In all, full understanding of conduct disorder requires developmentally sensitive classification as well as consideration of comorbid conditions.

Many children meet criteria for multiple Diagnostic and Statistical Manual (DSM) categories, such as oppositional defiant disorder (ODD) and conduct disorder (CD). If each disorder has been well validated as a separate entity, statistically significant co-occurrence of different disorders may be highly informative. However, ODD and CD have not been well validated as separate entities. The very high rate of overlap between diagnoses of ODD and CD may therefore arise artifactually from the lack of a valid distinction between them, rather than from potentially informative comorbidity between two different disorders. Empirical research strongly supports a distinction between two syndromes that correspond to subsets of the DSM-III-R criteria for CD. Designated as delinquent behavior and aggressive behavior, these syndromes have been found to differ in biological correlates, heritability, developmental stability, course, response to interventions, and long-term outcomes. At this stage of our knowledge, empirically based assessment and taxonomic methods can be especially useful for distinguishing between syndromes, deriving norms, doing longitudinal studies, and detecting patterns of comorbidity. These methods do not preclude categorical taxa, which can be formed by imposing cutpoints on the distributions of syndrome scores. Categorical taxa can also be formed by cluster analyzing profiles of syndrome scores. Accelerated longitudinal designs can bring multiple analyses to bear on empirically derived syndromes to test complex developmental relations more quickly and powerfully than can traditional longitudinal designs. Follow-ups of high-risk groups can identify variables that predict good versus poor outcomes if standardized baseline and outcome measures are used. Interventions for conduct problems should be designed to counteract the risk factors found to predict poor long-term outcomes.

A discussion of gender and conduct disorder must first answer the basic questions of whether or not there are any differences in prevalence, symptoms, and correlates of conduct disorder by sex. Several epidemiologic studies have found no difference in the prevalence of conduct disorder in adolescence by sex. Correlates of conduct disorder in girls are similar to those in boys (including aggression and internalizing disorders), once base rates of the correlates are accounted for. A major problem in studying conduct disorder in girls is the lack of appropriate criteria; the Diagnostic and Statistical Manual (3rd ed., rev.) criteria are not appropriate for girls. A case is made for sex-specific criteria for conduct disorder that take into account known differences in male and female childhood cultures and base-rate differences in aggression and criminality. Until basic issues of diagnosis and prevalence are resolved, other issues such as risk factors and developmental pathways cannot be successfully addressed.

Antisocial behaviors in females may differ from more prototypically “male” patterns of aggression, violence, and criminality that dominate criteria for conduct problems in diagnostic systems. This raises questions of how to define and investigate conduct problems in females as well as how to identify differential childhood antecedents. A developmental psychopathology perspective is advanced as one useful approach to understanding adaptive and maladaptive social patterns in males and females that may lead to different developmental trajectories and influence the forms of psychopathology that develop. The utility of de-emphasizing serious physical aggression as one important criteria for conduct problems is questioned. Recommendations that particular forms of deviance in females be considered as markers of their antisocial patterns (e.g., somatic complaints, friendlessness, underachievement) are also called into question. Rather than adopt sex-specific criteria to assess conduct problems, it is necessary to expand and broaden the diagnostic categories to include serious externalizing problems regardless of whether they occur in males or females.

Epidemiology, the study of patterns of disease distribution in time and space, offers a powerful set of theories and methods for understanding and preventing illness. The disruptive behavior disorders (DBDs) can be studied from the point of view of whether or not epidemiological methods could help to understand and prevent them. We suggest that the best approach to the DBDs is one that takes into account both the development of the child and the development of the disorder. Developmental epidemiology is the attempt to use recent advances in developmental psychopathology to help answer questions about how developmental processes increase or decrease vulnerability to DBDs, how risk factors change across childhood, and how the development of the disorder is reflected in rates of onset and patterns of symptomatology. Using examples of primary and secondary prevention studies, we discuss the importance of appropriate sampling strategies and a clearly articulated theoretical background, if intervention trials are to achieve their full utility as tests of a developmental theory of the DBDs.

Developmental sequences in disruptive behavior from childhood to adolescence are traced retrospectively and prospectively in two community samples of boys. Three developmental pathways are distinguished: (a) an early authority conflict pathway, consisting in sequence of stubborn behavior, defiance, and authority avoidance; (b) a covert pathway, consisting of minor covert behaviors, property damage, and moderate to serious forms of delinquency; and (c) an overt pathway, consisting of aggression, fighting, and violence. The overlap among the three disruptive pathways is examined. Those boys who escalated in the overt pathway were more likely to escalate in the covert pathway than boys escalating in the covert pathway showing an escalation in the overt pathway. Escalation in the authority conflict pathway was not associated with escalation in either the overt or the covert pathways. Boys' rate of self-reported delinquency was highest for those in triple pathways (covert-overt-authority conflict) or in certain dual pathways (covert-overt, covert-authority conflict). However, by age 16 the highest rate of offending was displayed by those in the triple pathways. The rate of violent offenses was also highest for those in the triple pathways and for those in the overt and covert pathways. Results from the rate for court petitions largely supported these findings. Lowest rates of offending were observed for boys in the overt and authority conflict pathways. Implications are discussed for clinical practice and future research.

This article reviews evidence from neuropsychological tests that brain dysfunction is a correlate of conduct disorder. Most studies report consistent findings of differential neuropsychological deficits for antisocial samples in verbal and “executive” functions. Neuropsychological measures are related to some of the best indicators of poor outcome for children with conduct symptoms, such as early onset, stability across time, hyperactive symptoms, and aggressiveness. Neuropsychological tests statistically predict variance in antisocial behavior independently of appropriate control variables. This article argues that neuropsychological variables warrant further study as possible causal factors for conduct disorder and presents one developmental perspective on how neuropsychological problems might contribute risk for conduct disorder.

In this commentary, we (a) discuss logical and empirical limits on the transactional model in accounting for the etiology of any developmental psychopathology, including conduct disorder (CD), and (b) review evidence bearing on whether or not frontal lobe lesions can directly produce CD behavior. Logically, transactions can both decrease and increase phenotypic variance; moreover, there is a mathematical limit (50%) to the amount of variance for which they can account. Empirically, documenting a transactional effect requires (a) that we have unconfounded measures of a child's biotype and social environment, (b) that the biotype and social environment are correlated, and (c) that we have a design (such as an adoption design) that is capable of separating the contribution of this correlation to outcome variance from the main effects of either biotype or social environment considered separately. Given these limits, we should also look for main effects in the etiology of CD. We argue that early damage or dysfunction in the frontal lobes may be one such plausible main effect on CD.

Undersocialized aggressive conduct disorder is conceptualized within the framework for personality and motivation of Jeffrey Gray. The disorder is seen as reflecting a dominance of the reward system over the behavioral inhibition system. Evidence for this conceptualizing coming from behavioral, psychophysiological, biochemical, and pharmacological studies is reviewed. Relevant findings from these studies include perseverative responding for reward, indices of inefficient noradrenergic and serotonergic functioning, and electrodermal underresponding. Additional research to test the proposed hypothesis is suggested.

An understanding of the interactions among developmental psychobiological and psychological and contextual aspects of conduct disorder has the potential for contributing new insights on the etiology and stability of conduct problems. The purpose here is to present theoretical and methodological issues germane to integrating a psychophysiological theory of conduct disorder with psychological developmental perspectives. These issues are theoretical concerns germane to a developmental perspective on conduct disorder, stability and change in conduct disorder, and bidirectionality in biological and psychological influences on conduct disorder.

This paper presents information pertaining to attachment processes as risk factors in the development of disruptive behavior in young children. In recognition of the fact that attachment is not the only or necessarily most important risk factor in the prediction of behavior problems, attachment is considered in the context of other domains of variables, including child biologic factors, family ecology, and parental management and socialization practices. Within the attachment domain, we describe three complementary processes that may lead to disruptive behavior: the information-processing aspects of affective-cognitive structures, the function of observable attachment patterns, and the motivational consequences of attachment security. The indirect effects of maternal representations of attachment on child disruptive behavior are also considered. Examples of protypical risk factor combinations involving attachment and other domains are provided. The implications of the attachment perspective for research and clinical work with young disruptive children are discussed.

Attachment theory and research have traditionally been subspecialties in infant social development. Recent work has extended the relevance of attachment theory and assessments well into childhood and has established firm ties to work with adults. Many of the same variables associated with the development of disruptive behavior problems also influence the development of attachment. In addition, recent data point to consistent relations between attachment status and disruptive behavior problems. This paper reviews attachment theory, summarizes mechanisms through which attachment might be related to disruptive behavior problems, and discusses the relevance of attachment theory to prevention and therapy. We emphasize the diversity of possible relations between attachment and disruptive behavior problems and the fact that incorporating attachment theory into research on disruptive behavior problems does not mean interpreting every disruptive behavior as attachment related or every attachment-related disruption as serving the same function.

Motivational ideas in psychology and their relation to causes are reviewed. Key ideas include hedonism, arousal, rational decision-making, reinforcement, and impulsivity. Motivations proposed for different acts classified as conduct disorder or delinquency are discussed as well as motivational constructs in delinquency theories. It is concluded that theories should include energizing, directing, inhibiting, and decision-making stages at a minimum. Empirical studies of motivations for delinquency that involve asking youth to give reasons for their acts are also reviewed. The most common motivations reported are utilitarian (seeking money or material goods) and hedonistic (seeking excitement or pleasure). It is pointed out that the validity of motivational theories is a separate issue from the validity of verbal reports of motivations. People may be aware of immediate situational motivations but unaware of longer term biological, psychological, or social factors. A program of research on the energizing, directing, and inhibiting of antisocial behavior is recommended and on the relative effectiveness, in reducing conduct disorder and delinquency, of programs targeting these different stages.

In this paper, both longitudinal and treatment studies relevant to conduct disorder (CD) are reviewed, and a developmental approach to its prevention is presented. Outcome studies for the treatment of CD and antisocial behavior are first reviewed to demonstrate that, although none have been entirely successful, many interventions have powerful effects on various symptoms that comprise the disorder, highly predictive antecedents, and risk factors. Second, the development of CD and the potency and interrelationship among antecedent and mediating variables is traced through the preschool and early elementary school years. Third, an attempt is made to synthesize the developmental and treatment research findings to suggest possible integrations of interventions that are promising for future preventive trials in the preschool and elementary school periods. It is concluded that, whereas before the entry to school preventive interventions targeted entirely in the family setting may prove successful, after the transition to school multisetting interventions will be essential. Finally, three examples of new and multisetting prevention trials are briefly described.

Two major transitions – initial school entry and transition to middle school – are emphasized as the points in development most amenable to preventing conduct disorder. As a complement to Reid's analysis of the child and family foci for prevention efforts, this paper discusses the importance of considering social context factors in prevention. In the early school years, peers inadvertently reinforce aggressive and coercive behavior and, thus, contribute to the coercive cycle Patterson describes in families. Middle schools in inner-city contexts have peer social network characteristics that also support delinquent and violent behavior more directly, in contrast to the general suppositions of social control theories of delinquency. The impact of neighborhoods and the larger societal tolerance of violence reflected in the media are also discussed. Prevention strategies for addressing these contextual factors at both developmental periods are outlined in the paper.

The present paper charts the course for future research on the treatment of conduct disorder. The course builds on current advances both in understanding conduct disorder and its treatment and in the clinical care provided to youth and their families. The paper highlights substantive and methodological advances in contemporary research on both treatment and developmental psychopathology more generally. In light of the current status of treatment research, several priorities are identified, which include expanding the criteria used to evaluate outcome, expanding the range of research questions asked about treatment, testing a broader range of treatments and treatment combinations, evaluating long-term treatment effects, and expanding the models to develop and to evaluate treatment. A plan is presented to guide the progression of research to identify and to develop effective interventions. The special opportunities that a developmental perspective provides for designing treatments for youth of different ages and stages of development are also discussed.

The thesis of this paper is that a reciprocal relation must develop between basic research on the developmental psychopathology of conduct disorder and applied treatment studies. Basic research can guide treatment design, and treatment outcomes can test developmental theories. The nature of conduct disorder seems to be one of multivariate components that act in self-perpetuating ways across development. These components include family, child-cognitive, peer group, and ecocommunity systems. Interventions that are directed toward just one component may be successful in producing proximal changes in the targeted domain, but they are not likely to be successful in long-term prevention of serious conduct disorder because other forces counteract these changes. The goal of treatment research needs to be long-term conduct disorder prevention. Two kinds of treatment studies are needed, one kind that is directed toward developing a technology of successful change procedures for individual processes and a second kind that uses these multiple change procedures in a comprehensive effort to prevent serious conduct disorder.

“Whether we form habits of one kind or of another from our very youth…makes a very great difference, or rather all the difference,” claimed Aristotle, wrestling with notions of conduct disorder (CD) and antisocial behavior two thousand years ago (Bk. II, chap. 1). As I respond to the invitation of John Richters and Dante Cicchetti to reflect on “where research on antisocial behavior has been, is now, and is going,” I think about the implications of Aristotle's claim. On the one hand, Aristotle points to the training received during childhood as key to understanding subsequent misbehavior. On the other, he believed that individuals must be responsible for their behavior and that, therefore, they must be able to make choices regarding how to act.