부산대학교 도서관

Callous-unemotional (CU) traits have proved to be a robust and informative construct; identifying a subgroup of children with conduct problems who show more severe and persistent antisocial behaviour. The majority of this work has focused on mid to late childhood and adolescent samples, yet the study of CU traits in early childhood allows identification of developmental pathways to CU traits and may inform the development of preventative interventions. The three empirical studies included in this doctoral thesis use a longitudinal epidemiological sample (Wirral Child Health and Development Study; WCHADS) followed from pregnancy up to age 7 years to examine important questions regarding: 1) the measurement of CU traits in early childhood 2) the contribution of the early parenting relationship to child CU traits; specifically maternal sensitivity to infant distress, with possible mediation by child attachment status, and 3) a candidate sex dependant mechanism for the translation of CU traits into physical aggressive from early to mid-childhood. The first study uses the extensive sample of consecutively recruited first time mothers and the second two studies focus on a subsample stratified by psychosocial risk. The aim of the first study (Chapter 2; n = 775) was to adapt a CU traits measure for use with preschool children. The CU measure derived showed acceptable psychometric properties, factorial invariance by sex and good stability to 5 years. Validity was supported by cross-sectional associations with physical aggression for both boys and girls and incremental prediction to aggression at age 5 in girls only. The second study (Chapter 3; n = 272) examined the longitudinal contribution of maternal parenting behaviours (sensitivity to distress and to non-distress, positive regard, intrusiveness) at 7 months and attachment status at 14 months to child CU traits assessed from age 2.5 to 5 years. Latent variable modelling yielded a single parenting factor which, in line with predictions, significantly predicted reduced CU traits. The effect was mainly explained by sensitivity to infant distress and positive regard towards the infant. These two indicators evidenced a significant interaction, such that the combination of low positive regard and low sensitivity to distress predicted increased child CU traits. Neither attachment security nor disorganization predicted CU traits, so there was no evidence for mediation by attachment status. The final study (Chapter 4; n = 276) examined a hypothesised sex-specific mechanism for the translation of CU traits to aggression via HPA –axis reactivity to stress. Age 5 cortisol reactivity was found to significantly moderate the association between age 5 CU traits and age 7 teacher and mother reported aggression, evidenced by a significant 3-way interaction with sex. There was a significant two-way interaction in boys, such that higher CU traits and lower cortisol reactivity predicted increased physical aggression. Overall, this thesis provides support for the valid measurement of CU traits over the early preschool period. Sensitivity to infant distress, alongside positive regard/warmth, predicted reduced CU traits suggesting that early interventions might also focus on enhancing maternal responsiveness to distress. Findings supported the role of cortisol reactivity to social stress in the translation of CU traits to aggression and critically this was sex specific.