Abstract
Different scoring methods have been proposed to identify adolescents with high levels of callous-unemotional (CU) traits as measured by the Inventory of Callous-Unemotional traits (ICU). This is the first study to scrutinize the usefulness of the normative, empirical, and DSM-5 with Limited Prosocial Emotions (LPE) scoring methods in criminal justice-involved girls (N = 192). Girls (Mage = 15.74) completed the ICU and other self-report questionnaires. A diagnostic interview was used to determine the presence of conduct disorder (CD) and other psychiatric disorders. Official arrest and rearrest data were also collected. The three scoring methods designated girls with (versus without) CU traits who scored higher on various indices of past and future antisocial behavior, exhibited higher rates of (childhood-onset) CD and higher levels of psychopathic traits, and displayed lower levels of agreeableness and conscientiousness. It was only when the normative and empirical scoring methods were applied that girls with CU traits had higher rates of substance use disorders and were less anxious, depressed, and/or neurotic. In a subsample of girls with CD (N = 119), the three scoring methods captured a more severely antisocial subgroup. Other expected group differences emerged, though mainly when the normative scoring method was used. Results provided support for the clinical utility of the three scoring methods but also showed that the normative scoring method most aligns with theoretical expectations. Research is warranted to explore if this normative scoring method serves as a better categorically-defined specifier for CD than the specifier that is incorporated in the DSM-5.