This chapter reports a comparison of Dunedin Study males and females who exhibited childhood-onset versus adolescent-onset antisocial behaviour. We tested whether childhood-onset delinquents had childhoods of inadequate parenting, neurocognitive problems, and temperament and behaviour problems, whereas adolescent-onset delinquents did not have these pathological backgrounds. We also queried whether females as well as males showed this differential pattern of risk, asking whether childhood-onset females had high-risk backgrounds, but adolescent-onset females did not. Finally, we ask whether Dunedin Study members on the life-course persistent antisocial path had more problem outcomes as young adults compared to those on the adolescence-limited delinquent path.

Between 1985 and 1988, when the members of the Dunedin Study grew from age 13 to age 15, we observed that many Study members who had not shown antisocial behaviour problems began to take up delinquent activities as they made the transition from childhood to adolescence. Our first descriptive report of this phenomenon contrasted those adolescent-onset delinquents with their counterparts who had been showing antisocial behaviour problems since early childhood (Moffitt, 1990b). We reported that childhood-onset delinquents were characterized by abnormal levels of individual and contextual risk factors (hyperactivity, low IQ and family adversity, among others) whereas adolescent-onset delinquents were not. The adolescent-onset delinquents' backgrounds were not inordinately healthy, but they were not pathological either; they were simply average.