Abstract
Although research into the continuity of disorder from childhood to adolescence is sparse, results from both longitudinal and cross sectional studies suggest that the prevalence of disorder increases for girls but may remain more stable for boys. In this paper, the methodologies of two assessment phases of the Dunedin longitudinal study have been equated to estimate the continuity of DSM-III disorder from ages 11 to 15. Although the overall prevalence of disorder doubled between the ages, this was primarily because of an increase in nonaggressive conduct disorder and major depressive episode. The sex ratios in disorder had largely reversed from a male predominance at 11 to a female predominance at 15. In terms of persistence, over 40% of those with disorder at age 11 were also identified at age 15. However, over 80% of those identified with disorder at 15 did not have a history of disorder at 11. Significant sex differences were also found in the continuity of internalizing and externalizing disorders, with externalizing disorders showing more continuity for boys, and internalizing for girls. Logistic regression models were employed to evaluate the roles family background, academic and social competence, and early histories of behavior problems may play in the determination of disorder continuity.