These differing findings suggest that conclusions on ODD→CD continuity may be sensitive to a variety of factors: age/developmental stage, sample composition, measurement techniques and analytic strategies. Not dissimilar variations have emerged in past reports from the Great Smoky Mountains Study (GSMS), the sample we report on further here. Using a ‘chain’ analysis where ODD at any annual assessment was used to predict CD at the immediately succeeding wave, Costello, Mustillo, Erkanli, Keeler & Angold (2003) found no prediction from ODD to CD between ages 9 and 16 years. Rowe et al. (2002), however, found that ODD at wave 1 predicted CD at any of three subsequent waves among boys. Links with CD were stronger than with other common outcome disorders, consistent with the specific relationship between ODD and CD assumed in DSM-IV. Results showed a markedly different pattern in girls. None of the girls who met criteria for ODD but not CD at wave 1 developed CD at later waves. Instead, persistence in ODD and heterotypic continuities to depression were common outcomes for girls. In a literature still largely focused on males, these findings emphasize the need for tests of sex-specific outcomes.
