A rather different possibility is that we underestimated ODD→CD because some of the children with CD in our dataset had a history of ODD in early childhood that had resolved prior to entry into GSMS. We have no way of testing this possibility directly. We did examine whether ODD→CD progression was more common at younger ages within the GSMS timeframe, however. Developmental variation was highlighted as the youngest cohort of GSMS (entering age 9) showed a stronger ODD→CD link than did youth who entered the study at older ages (age 11 or 13). Even in the youngest cohort, less than 1 in 5 cases of CD was preceded by ODD. Therefore we found some evidence that ODD→CD progression is more common earlier in the GSMS age range. To our knowledge this is the first analysis to test whether ODD→CD transition varies by age. Younger samples will be required to test whether this pattern extends earlier into childhood. If replicable this might indicate that ODD is a stronger precursor to full disorder CD that onsets in childhood rather than adolescence. This possibility should be addressed in further research.
