Understanding the covariation among childhood externalizing symptoms: genetic and environmental influences on conduct disorder, attention deficit hyperactivity disorder, and oppositional defiant disorder symptoms.

Conduct disorder (CD), attention deficit hyperactivity disorder (ADHD), and oppositional defiant disorder (ODD) are three of the most common childhood externalizing behavioral disorders. ODD typically occurs in early childhood and is characterized by behaviors such as arguing with adults, losing one's temper, and angry or intentionally annoying behavior. CD often develops later than ODD, in early adolescence, and is characterized by behaviors including stealing, lying, fire setting, truancy from school, and property destruction. Although children with ODD often are diagnosed with CD when they reach adolescence, not all individuals with CD have had a previous diagnosis of ODD (Lahey, McBurnett, & Loeber, 2000). ADHD involves a pattern of behavior in which children are fidgety and restless, and have difficulty remaining in their seats, waiting their turns, or sustaining attention on a particular task.

Comorbidity among these behavioral disorders has been reported in both epidemiological and clinical samples (Biederman, Newcorn, & Sprich, 1991; Jensen, Martin, & Cantwell, 1997; Simonoff et al., 1997). That observation has generated considerable debate about the appropriateness of the current diagnostic system for these childhood externalizing disorders. Some research groups have proposed that individuals with comorbid disorders may represent distinct subtypes (Biederman et al., 1991; Faraone, Biederman, & Monuteaux, 2000). For example, it has been suggested that ADHD with CD may represent a more severe form of ADHD, while ADHD with ODD represents an intermediate phenotype between ADHD with CD and ADHD alone (Biederman et al., 1991). Clearly, there is some utility in making distinctions among the externalizing disorders, because they appear to be associated with different correlates and outcomes. CD is strongly associated with both concurrent and future alcohol use, more so than the other childhood externalizing disorders (Kuperman et al., 2001a, 2001b; Molina, Bukstein, & Lynch, 2002; Moss & Lynch, 2001). CD is also more strongly associated with future criminal and antisocial behavior (Crowley, Milkulich, MacDonald, Young, & Zerbe, 1998). ADHD tends to be more closely related to academic failure and cognitive deficits (Fergusson, Horwood, & Lynskey, 1993).

The causes for the interrelationships among ADHD, ODD, and CD are not fully understood. Twin studies are able to tease apart the extent to which phenotypic association is due to shared genetic and/or environmental factors. Understanding the extent to which different etiological factors contribute to the overlap among disorders has several important implications. Were the same genes found to influence these different disorders, their comorbidity might be evidence of a shared biological predisposition for several behavior problems. Alternatively, if the disorders overlap largely due to environmental reasons, it would have important implications for potential prevention and intervention efforts. Understanding the extent to which the same and/or different genes and environments contribute to these disorders also influences the way in which we classify and group these disorders; if "different" externalizing disorders are actually the result of the same genes, it may suggest that they should be considered a joint construct with varying symptomatic presentation. The extent to which the same genes contribute to each of these disorders also has implications for studies attempting to identify the specific genes involved in the disorders.

Several twin studies have suggested that common genetic factors contribute to the overlap among these disorders (Nadder, Silberg, Eaves, Maes, & Meyer, 1998; Silberg et al., 1996; Thapar, Harrington, & McGuffin, 2001; Waldman, Rhee, Levy, & Hay, 2001); however, not all twin studies have reached that conclusion (Burt, Krueger, McGue, & Iacono, 2001). In addition, there is disagreement about the extent of overlap among genetic influences on the externalizing disorders. Some studies have found that the genetic correlation between CD and ODD is sufficiently high that they have considered these symptoms to be part of a joint construct (Eaves et al., 2000; Nadder, Rutter, Silberg, Maes, & Eaves, 2002). Very few genetically informative studies to date have separately analyzed CD and ODD symptoms, and one of the few that did was the study suggesting that shared environmental influences contribute most strongly to the covariation among ADHD, CD, and ODD (Burt et al., 2001). Data from a twin study conducted in the United Kingdom found that ADHD symptoms and CD symptoms could be explained by a single genetic liability (Thapar, Hervas, & McGuffin, 1995); whereas, analyses of Australian twin data suggested that, although there was considerable genetic overlap, there were also significant genetic influences unique to each disorder. Thus, there is inconsistency in the literature regarding both the sources of covariation among the externalizing disorders, and the degree to which each of these disorders is influenced by unique genetic and/or environmental factors.

To help resolve these issues, this study investigated genetic and environmental contributions to the covariation among these externalizing disorders. Symptom counts for ADHD, CD, and ODD were all analyzed separately. We analyzed self-report interview data from 631 pairs of twins, recruited as part of a population-based Finnish twin study, FinnTwin12, with complete data on all externalizing disorders. The interviews were age-standardized; all twins were age 14 at the time of interview. The inclusion of both male and female same-sex twins allowed us to examine potential gender differences in the overlap among externalizing disorders.

METHODS

Sample

FinnTwin12 (FT12) is a population-based sample consisting of five consecutive birth cohorts of twins born in Finland from 1983 to 1987. All twins were identified through Finland's Central Population Registry, yielding comprehensive and unbiased ascertainment (Kaprio, Koskenvuo, & Rose, 1990; Kaprio, Pulkkinen, & Rose, 2002; Pulkkinen, Kaprio, & Rose, 1999). All twin families identified in the population registry with both cotwins living and resident in the country were included, unless both twins lived apart from both biological parents or one or both twins were in an institutional care facility and unable to comply with study demands. Baseline assessment was conducted late in the year in which each successive twin birth cohort reached 11 years. A total of 2724 families (87% of all identified eligible families) returned the initial family questionnaire, and from these families, 2567 twin pairs completed baseline questionnaires. The sample of participant twin-families resided throughout the whole of the country, and it is representative of the Finnish population from which it was ascertained; 14% of the twins' fathers (and 15% of their mothers) had earned a university degree.

From this epidemiological sample, a subset of 2070 twins (1035 pairs) was selected for more intensive study (Kaprio et al., 2002). This subsample is described in detail elsewhere (Rose et al., 2004). Briefly, the intensive sample of FT12 consists of a pilot sample drawn from the 1983 cohort (13%), which was randomly selected from a geographically limited region of the main population areas in Finland, a random sample selected from all eligible twin pairs in the remaining four birth cohorts (59%), and an enriched sample (28%) which added twin families in which one or both of the twins' parents exceeded a cutoff ([greater than or equal to]6 of 11 items) on our lifetime version of the Malmo-modified Michigan Alcohol Screening Test (Mm-MAST; Seppa, Sillanaukee, & Koivula, 1990), an 11-item diagnostic screen for alcohol-related problems, included in both parents' questionnaires administered at baseline. …