Risk factors, histories of parental antisocial behavior and mental health problems were closely similar in CD and ODD, as were elevated levels of exposure to social adversity and problems in family functioning. Although few studies have compared the correlates of ODD and CD, this pattern of findings is consistent with previous reports based on GSMS (Rowe et al., 2002) and other datasets (Kolko et al., 2008), and runs counter to the view that ODD is associated with lower levels of such risks as compared with CD. There were two exceptions to this pattern. First, boys were at greater risk of CD (whether alone or in combination with ODD), but there were no sex differences in rates of ‘pure’ ODD. This finding is also largely consistent with reports from other studies, although many have reported at least some tendency for more boys to meet criteria for ODD than girls. We suspect that these study differences may largely reflect differing treatments of ODD/CD comorbidity: by definition, our ‘ODD only’ group excluded cases with comorbid CD, whereas studies examining overall prevalence rates frequently
