The analyses to this point all suggested that ODD (whether assessed at the symptom or the diagnostic level) functioned as a precursor to, and risk factor for, CD. What they could not demonstrate, however, was how commonly this progression occurred. To assess this, we classified each child into one of six mutually exclusive onset ‘pathways’ reflecting the ordering of onsets of ODD and CD across study waves. As Table 1 shows, many disruptive youth only ever met criteria for one of these disorders between late childhood and mid-adolescence, and transitions between the two disorders were less common than anticipated. In particular, of the children who met criteria for CD a substantial proportion did not have ODD at any study wave. In children with ODD no diagnosis of CD during the study was also common (60.5%, 95% CI: 49.7, 70.2). Table 1 shows that the confidence intervals around percentages of children falling into each of the comorbid pathways overlap. Concurrent CD and ODD onset was the most common configuration. ODD→CD was slightly less common than the pathway where CD was diagnosed
