Oppositional Defiant and Conduct Disorders
Joseph M. Rey
Garry Walter
Cesar A. Soutullo
… His academic performance declined, he became more sullen and argumentative, was occasionally truant or caught lying about his whereabouts, did not adhere to curfews, and had joined in with other troubled teenagers. His mother was sure that he had been stealing money from her purse, possibly to buy cigarettes. The teachers at school had also become concerned about his increasingly unruly behavior and advised his parents to seek professional help.
Young people with the kind of problems described in this brief vignette represent the largest single group of patients seen in child and adolescent mental health settings; they are usually labeled as suffering from oppositional defiant disorder (ODD), conduct disorder (CD), or disruptive behavior disorder (DBD). The cost of these conditions to the individuals themselves, their families, and society is high (1, 2). Leaving aside personal and family suffering and the burden to health services, most juvenile delinquent acts are perpetrated by individuals with CD (3). The number of publications dealing with DBDs is enormous and growing exponentially.
Some professionals describe these children as having a “behavior disorder,” which often has negative connotations or implies that the condition is nonpsychiatric, that is, that children are “bad” rather than “mad.” However, the DBDs have diagnostic validity because i) symptoms intercorrelate highly, suggesting coherent syndromes rather than an aggregate of various types of deviance; ii) a genetic component is increasingly being documented; iii) these problems are recognized in every society and historical period; and iv) although prevalence may vary with time and place, this also applies to almost all mental and physical disorders (4). While a categorical diagnosis is useful to facilitate communication and clinical decisionmaking, evidence suggests that a dimensional understanding of these disorders might be more helpful, particularly in research.
These “bad kids” have a particularly negative outcome in adulthood. They are more likely to have psychiatric disorders, to show difficulties at work, and to be involved in violent relationships and marriages (5,6,7,8). In turn, these behaviors increase the chances of similar problems in their offspring, often abused or neglected by them, thus perpetuating the disorder (9). The challenge lies in finding ways to break this unfortunate cycle.
A pioneer in this area already suggested in 1935 of the children in question that “a strict definition or delimitation of the [component] groups is difficult because they tend to merge with each other” (10). Indeed, individuals with ODD and CD share many characteristics. For example, their conduct is socially unacceptable, they cause disruption or distress to others more than to themselves (i.e., they “externalize” their problems), and they are more likely to be male and to find it difficult to
learn from experience. Yet, they also differ widely from one another: They may be aggressive or not, break the law or not, feel guilt and empathy or not, may be sensation-seeking or not. Although ODD and CD represent well characterized, reliable behavioral syndromes, they do not fit easily into a traditional “illness” model because children with these problems are heterogeneous in relation to etiology, natural history, response to treatment, and outcome. This is similar to what happens with heart disease, where multiple factors contribute to the pathology and other illness characteristics, but the construct has heuristic value. Because of their heterogeneity, the usefulness of a diagnosis of ODD or CD for treatment planning is limited; an emphasis on identifying each child’s problems and modifying the individual risk factors is likely to be the best way of managing these disorders, as is the case with heart disease.
learn from experience. Yet, they also differ widely from one another: They may be aggressive or not, break the law or not, feel guilt and empathy or not, may be sensation-seeking or not. Although ODD and CD represent well characterized, reliable behavioral syndromes, they do not fit easily into a traditional “illness” model because children with these problems are heterogeneous in relation to etiology, natural history, response to treatment, and outcome. This is similar to what happens with heart disease, where multiple factors contribute to the pathology and other illness characteristics, but the construct has heuristic value. Because of their heterogeneity, the usefulness of a diagnosis of ODD or CD for treatment planning is limited; an emphasis on identifying each child’s problems and modifying the individual risk factors is likely to be the best way of managing these disorders, as is the case with heart disease.
Although the issue of whether these diagnoses represent independent conditions is unresolved (11, 12), ODD and CD are described together to highlight their similarities and differences and to avoid unnecessary repetitions. The content of this chapter overlaps and should be read in conjunction with topics discussed in other parts of the book, particularly those on Aggression in Children (Chapter 5.2.3), Fire Behavior in Children and Adolescents (Chapter 5.2.4), and the section on Juvenile Delinquency (included in Chapter 6.3.3).
Historical Note
Understanding of the DBDs has been closely linked to the study of delinquency, largely conceptualized in the past as a failure in moral development or “moral insanity.” Lombroso, an Italian psychiatrist, gained prominence at the end of the 19th century with his theory that some individuals were born criminal. About 50 years later Cleckley (13) coined the concept of “psychopathy.” According to him, psychopaths lacked remorse, were unable to have close relationships, were egocentric, and showed poverty of affect but, contrary to the prevailing opinion, were not necessarily aggressive or criminal. Although this concept of psychopathy was initially applied to antisocial adults, it was soon extended to young people. Psychoanalytic theories were postulating that the behavior had its origins in childhood and was the result of a poorly developed superego, thus opening the way to therapeutic interventions. Interest in juvenile delinquents, which had begun with Aichhorn’s Wayward Youth (10), impelled Bowlby (14) to examine these youth’s early experiences and ultimately led to the development of attachment theory. Robins (5) demonstrated the continuity between childhood conduct problems and adult antisocial personality, stressing their long-term implications. Levy (15) first described the syndrome of childhood oppositionality, which he characterized as the refusal to conform to the ordinary requirements of authority, or willful contrariness. The second half of the twentieth century saw the inclusion of these disorders in the psychiatric taxonomies, their splitting off from attention deficit hyperactivity disorder (ADHD), growing elucidation of their determinants and comorbidity, their implications for public health, and the development of preventive measures.
Classification
Definition
DSM-I (16) included no childhood disorders. DSM-II (17) listed a category of Unsocialized Aggressive Reaction of Childhood (or Adolescence), Group Delinquent Reaction and Runaway Reaction as the closest equivalents to CD. The category of CD was officially introduced in DSM-III (18) to describe children who showed a persistent pattern of behavior in which the basic rights of others or major age-appropriate societal norms or rules are violated, the definition still used in both DSM-IV (19) and ICD-10 (20). DSM-III also listed Oppositional Disorder to characterize children who show persistently disobedient, negativistic and provocative opposition to authority figures, manifested by violations of minor rules, temper tantrums, argumentativeness, provocative behavior and stubbornness. This description remains in DSM-IV (19) as Oppositional Defiant Disorder.
DSM-IV and ICD-10 Diagnostic Criteria
Diagnostic criteria for ODD and CD according to DSM-IV TR (21) and ICD-10 (20) are summarized in Table 5.2.2.1. Both taxonomies require the absence of CD to allow a diagnosis of ODD (i.e., implying a hierarchical relationship). ICD-10 explicitly conceptualizes ODD as part of the same dimension as CD, the former being a milder version of the latter (thus, the term conduct disorder often means ODD and CD in countries where ICD-10 terminology is used). In spite of the efforts made to homogenize the main classification systems, a conceptual divide remains. Yet, diagnostic criteria are very similar (Table 5.2.2.1) and differences largely reflect the time requirements for symptoms to be present, exclusion criteria, and the way both taxonomies deal with comorbidity. Using ICD-10 criteria seems to result in more children qualifying for ODD than when using DSM-IV criteria (40% more cases in one study) (22). This is because many children have symptoms of both ODD and CD but not enough to meet criteria for either diagnosis according to DSM-IV (for example two symptoms of each). Children diagnosed as suffering from ODD by ICD-10 but not by DSM-IV do not seem to differ in any respect from those diagnosed by DSM-IV, suggesting that DSM-IV criteria may be too strict or too narrow (22).
Empirical Approaches to Taxonomy
Empirical classifications (23) based on statistical analysis of symptoms have found an aggressive syndrome, which comprises symptoms of both ODD and CD, like destroying objects, bullying, fighting, and vandalism, and a delinquent or rule-breaking syndrome, characterized by stealing, lying, and truancy, which entails features of CD alone. Several types of aggressive behavior have also been identified (see Chapter 5.2.3) Dimensional taxonomies overlap with categorical diagnoses and are widely used to supplement each other and for research (23). The pros and cons of each are discussed in Chapter 4.1.
Subtypes of CD
Subtyping is important given the heterogeneity of CD. Subgroups may have implications for prevention, treatment, and prognosis. However, much disagreement still exists, exemplified by changes to the subtypes of CD listed in each edition of the DSM. DSM-III (18) divided CD into four subgroups according to whether children were socialized or undersocialized, and aggressive or nonaggressive. This was changed in DSM-III-R to “solitary type” and “group type.” DSM-IV (21) classifies CD according to age of onset (childhood or adolescent onset type, depending on whether there were symptoms prior to the age of 10 years). ICD-10 (20) describes three subtypes of CD: confined to the family context, unsocialized, and socialized.
Some researchers have emphasized the importance of overt (characterized by confrontation and fighting) and covert
(typified by deception, such as stealing and lying) symptoms (24), which overlaps with the aggressive–nonaggressive distinction. There are data showing that two different types of covert antisocial behavior may exist: property violations (e.g., stealing) and status offenses (truancy, running away) (24, 25). Other scholars suggest that life-course persistent (beginning during childhood and persisting past adolescence) and adolescence limited may be a useful distinction (26).
(typified by deception, such as stealing and lying) symptoms (24), which overlaps with the aggressive–nonaggressive distinction. There are data showing that two different types of covert antisocial behavior may exist: property violations (e.g., stealing) and status offenses (truancy, running away) (24, 25). Other scholars suggest that life-course persistent (beginning during childhood and persisting past adolescence) and adolescence limited may be a useful distinction (26).
|
Comorbidity
The vexatious issue of comorbidity (27), which presents numerous, still unresolved theoretical and practical questions (discussed in Chapter 4.1) is highly relevant for ODD and CD since both often cooccur with other diagnoses. The most frequent comorbidities are with ADHD [about 10 times more often than expected (27)], major depression [about seven times (27)], and substance abuse [in adolescents, about four times (28)]. According to ICD-10 (20), CD is not diagnosed if ADHD is present (it would warrant a diagnosis of hyperkinetic conduct disorder), or if CD is associated with emotional disorders (mixed disorder of conduct and emotions). In similar circumstances, DSM-IV (21) requires multiple diagnoses to be made (e.g., ODD and ADHD, CD and major depression). Evidence about the best way of conceptualizing these problems (e.g., ODD and ADHD: two coexisting disorders vs. one condition different from both ODD and ADHD) is lacking. Boys with CD and ADHD have an earlier age of onset of disruptive behavior symptoms and worse outcome than those without ADHD (27), and ADHD symptoms in children with ODD may increase the likelihood of progression to CD. Less is known about individuals with DBDs and comorbid emotional disorders besides depression, but anxiety and somatoform disorders also appear to be more frequent than expected (29).
Reliability
The diagnoses of ODD and CD have acceptable interrater and test-retest reliability, comparable to the reliability of most psychiatric diagnoses in young people (30,31,32). Agreement
varies according to the informant and age of the child, being usually higher when parent or teacher reports or multiple informants are used, for aggressive than nonaggressive behaviors, and in older children or adolescents (33). Reliability is also higher in clinic than in community samples due to base rate issues. For example, test-retest agreement of CD using the NIMH Diagnostic Interview Schedule for Children Version IV in a clinic sample was κ 0.70 for parents as informants, 0.86 for children, and 0.71 when using data from both. The parallel results in a community sample were 0.56, 0.64, and 0.66 (32). Reliability of ODD is lower, in the range of κ = 0.4 – 0.6 (30, 32).
varies according to the informant and age of the child, being usually higher when parent or teacher reports or multiple informants are used, for aggressive than nonaggressive behaviors, and in older children or adolescents (33). Reliability is also higher in clinic than in community samples due to base rate issues. For example, test-retest agreement of CD using the NIMH Diagnostic Interview Schedule for Children Version IV in a clinic sample was κ 0.70 for parents as informants, 0.86 for children, and 0.71 when using data from both. The parallel results in a community sample were 0.56, 0.64, and 0.66 (32). Reliability of ODD is lower, in the range of κ = 0.4 – 0.6 (30, 32).
Validity
Developmental Considerations
Children’s prosocial impulses already become apparent in the first year of life, for example through cooperative interactions and sharing. Learning how to deal with and tolerate frustration are important aspects of the socialization process. A degree of defiance and noncompliance is normal in toddlers, probably reflecting the child’s assertiveness and search for autonomy or ignorance of what parents are prepared to tolerate. Notwithstanding this, toddler’s behavior may already be indicative of problems when it is too intense, persistent, or pervasive. However, there are considerable individual variations and distinguishing behaviors that are within the normal range from problematic defiance or noncompliance is difficult at that age. Prosocial behaviors usually increase up to the age of three years; a temporary decline then begins to emerge. Defiance and noncompliance, particularly in boys, may also increase about the age of two or three years. For the development of aggression, please see Chapter 5.2.3.
ODD symptoms appear earlier than CD symptoms (e.g., stubbornness at a median age of 3 years, defiance and temper tantrums at 5, argumentativeness at 6, compared with lying at 8, bullying at 9, and stealing at 12 years) (34). Aggressive behavior (hitting, biting, smashing objects) is common in 4- to 8-year-olds and decreases with age, although severely aggressive acts typically start after puberty. Covert antisocial actions such as property and status violations (stealing, truancy, running away) increase as children become older, being more prevalent during adolescence. Early adolescence is often associated with an increase in rebellious behavior. Teachers’ reports indicate that most oppositional symptoms, such as arguing, screaming, disobedience, and defiance peak between 8 and 11 years and then decline in frequency (35).
Stability and Change
Disruptive behaviours are quite prevalent in children but often extinguish as they grow older. However, many data show that CD symptoms are more enduring than changeable. Research corrected for measurement error, which is considerable, suggests that stability may be even greater than previously thought (36). This may be partly explained by symptoms of conduct disorder becoming increasingly varied with the passage of time while showing a growing resistance to change (37).
ODD and CD
The relationship between ODD and CD is complex. In some children, ODD symptoms begin in infancy, persist during childhood, and evolve into CD, often after puberty. Other young people show noncompliance and defiance for short periods or do not progress to CD; this may occur more often in females than in males (38). Oppositional behavior is present only at home in some children, while symptoms occur in most settings in others. ODD often starts in the family context and generalizes to other settings over time.
Adult Outcomes
The continuity between childhood CD and adult antisocial personality disorder (ASPD) has long been known (5,6,39), so much so that DSM-IV (21) requires evidence of CD prior to the age of 15 years for a diagnosis of ASPD. The validity of this requirement, particularly in females, has been questioned (40). A diagnosis of ASPD is only allowed after the age of 18 years (DSM-IV) (36), but there is support for the view that such a diagnosis may be appropriate in some adolescents (41).
Data are mounting showing that childhood DBDs are associated not only with ASPD but with a wide range of other psychiatric disorders in adulthood (e.g., substance abuse, major depression, psychosis), as well as with many adverse outcomes such as suicidal behavior, delinquency, educational difficulties, unemployment, and teenage pregnancy (7,8,42,43,44,45,46,47). The association, which applies equally to males and females (8), reflects not only the already noted stability of disruptive behaviors, but also the fact that childhood DBDs often trigger a chain of events that increase the likelihood of such unfavorable outcomes (for example, early defiance may lead to harsh parental discipline, aggressiveness, and peer rejection, which may in turn be followed by association with deviant peers, antisocial acts, substance use, conflict with the law, and mental illness). It has been argued that much of the research on which these findings are based has significant limitations because studies tend to focus on one specific outcome, such as psychiatric disorder or substance use, and because they often examine consequences over a limited followup period. However, when these weaknesses are circumvented, childhood DBDs are still associated with a wide range of unfavorable adult outcomes, which attests to the robustness of the links (8). Further, studies by and large show a dose–response relationship: The higher the number and variety of disruptive behaviors, the worse the adult outcomes. This said, most adolescents with CD do not develop ASPD in adulthood. A large decline in delinquent and antisocial activities is a commonly reported phenomenon in early adulthood (3, 48). It is not clear if this parallels a reduction in DBDs or if their manifestations change with age (e.g., whether behaviors extinguish, mutate into other psychiatric problems, or delinquent acts become more covert). This decline may reflect the existence of a desisting adolescent-limited— as opposite to a life-course persistent— CD (26). Early onset, severity, and exposure to risk factors would predict the latter group (48).
Most epidemiological studies in adults have neglected to examine the prevalence of childhood-onset disorders, the exception being a recent survey of people older than 18 years in the United States, which reported a 12- month prevalence of 1% each for ODD and CD (49). This shows DBDs can be found in adults although it is not clear whether they appear de novo or are a continuation of childhood problems.
There is little information about the adult consequences of children with ODD who do not develop CD. Passive-aggressive personality disorder has been hypothesized as one of the outcomes but there are no data supporting this (6).
Epidemiology
Prevalence
Estimates of the prevalence of ODD and CD vary depending on the population, diagnostic criteria, instrument, period considered (point or lifetime), and informant. Recent surveys using DSM-IV criteria have produced reasonably consistent results, summarized in Table 5.2.2.2. Overall, about 5% of children aged 6 to 18 years met DSM-IV criteria for ODD or CD in the previous 3 or 6 months, slightly lower rates than those reported in earlier studies using DSM-III or DSM-IIIR criteria (50).
However, the proportion that has met criteria for either ODD or CD at any time before 16 years of age (lifetime rate) is higher respectively for girls and boys; ODD: 9.1%, 13.4%; CD: 3.8%, 14.1% (51). As noted, prevalence of ODD is elevated when using ICD-10 compared with DSM-IV criteria (22).
However, the proportion that has met criteria for either ODD or CD at any time before 16 years of age (lifetime rate) is higher respectively for girls and boys; ODD: 9.1%, 13.4%; CD: 3.8%, 14.1% (51). As noted, prevalence of ODD is elevated when using ICD-10 compared with DSM-IV criteria (22).
TABLE 5.2.2.2 POINT PREVALENCE (PERCENT) OF DSM-IV OPPOSITIONAL DEFIANT DISORDER (ODD) AND CONDUCT DISORDER (CD) IN RECENT EPIDEMIOLOGICAL STUDIES USING DSM-IV CRITERIA | ||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| ||||||||||||||||||||||||||||||||||||||||||||
