Oppositional defiant disorder (ODD) is the single most prevalent psychiatric illness in children and adolescent in the UK(1) and is a very strong predictor of psychiatric disorders in adulthood (2). The strongest associations of ODD are with conduct disorder, antisociality, and ADHD (1, 3-5); however, ODD is also strongly correlated with emotional disorders both cross-sectionally and longitudinally (4-6). This wide range of associations and predictions suggests that ODD plays a pivotal role in the development of psychiatric illnesses (4). Despite intense research interest in ODD, it remains unclear what the underlying causes for this breadth of associations may be. A better understanding of ODD could potentially help explain issues of particular importance to child psychiatry. Firstly, it would shed light on the puzzling rates of comorbidity in children and adolescents (5, 7), and, secondly, it could help understand developmental trajectories from early life into adulthood.

We have recently been able to demonstrate in a large national sample (the Office for National Statistics Child and Adolescent Mental Health Studies from 1999 and 2004), that ODD may comprise of three dimensions that carry relatively distinct associations with other childhood psychiatric disorders (Stringaris & Goodman, submitted). We have provisionally termed these three dimensions Irritable, Headstrong, and Hurtful, conforming to their respective predictions. The Irritable dimension, composed of items indexing temper outbursts and anger, has strong associations with emotional disorders, whilst the other two dimensions predict more strongly to conduct disorders. In addition, the Headstrong dimension appears to be most strongly associated with ADHD and the Hurtful dimension to be a strong predictor of pre-meditated aggressive offending. As a result of these findings, which preliminary evidence suggests are replicated in longitudinal analyses, we are proposing a new model for ODD. Thereby, the three dimensions we have identified represent relatively distinct contributions to ODD that we hypothesize will arise from different sets of predisposing factors.

To study these questions will require a longitudinal study with measures of ODD symptomatology in middle/late childhood, and hypothesized predictors earlier in development. ALSPAC is the only study in the United Kingdom and one of the few in the world that would allow us to address these issues.

AIMS

The aim of the present study is to identify underlying factors for oppositionality in children and adolescents. Based on our findings so far, our main hypothesis is that the three dimensions we have identified as contributing to ODD are predicated on distinct predisposing factors. We expect the distinct correlates to lie in the domains of child temperament, family mental illness, parent child relationship, and academic attainment. Our hypotheses are as follows:

Temperament: we predict the Headstrong and Hurtful dimensions to be best predicted by "difficult" temperament. In contrast, we hypothesize that the temperament underlying the Irritable dimension will be mixed, encompassing both "inhibited" and "difficult" temperaments.

Family Mental Illness: we expect that all three dimensions will be predicted by a family history of antisociality. However, we predict that a family history of emotional disorders, such as depression and anxiety, will be differentially associated with the Irritable dimension.

Family Relationships: our prediction is that family relationships are important for both the Headstrong and the Irritable dimension. In contrast, we predict that such environmental effects will be of limited influence for the Hurtful dimension. This is due to evidence we have gathered so far showing that this dimension may be linked to callous and unemotional traits; these have been shown to be mostly due to additive genetic effects.

Cognitive Attainment: similar to our predictions about other environmental effects, we expect that frustration due to low academic achievement will be most relevant for the development of the Irritable and Headstrong dimension and significantly less so for the Hurtful dimension.

ANALYTIC DESIGN

This will be a developmental follow-back study, that is, we will establish symptom dimensions at the ages of 8 and 10 years and will then seek to establish their antecedents as detailed above. The study could thus be completed entirely on the basis of existing ALSPAC data. The instrument we will use to construct the dimensions will be parent-reported ODD (awkward and troublesome) symptoms collected as part of the DAWBA assessments completed at XX and YY months (this we have also used in the ONS study).

IMPLICATIONS

We expect our findings to help better understand the development of mental illness in childhood and adolescence. In particular, we expect our results to be particularly relevant for the central question of comorbidity between mental disorders in childhood. Furthermore, if childhood oppositionality encompasses more than one dimension this could mean that differential interventions might be indicated.