Sources of Continuity and Change in Callous-Unemotional Traits in Early Childhood

Sources of Continuity and Change in Callous-Unemotional Traits in Early Childhood Introduction • Callous-unemotional traits (CU; e.g., lack of guilt, empathy and affect) can be predictive of later psychopathy and other detrimental outcomes (Viding & McCrory, 2012). • Research on the etiology of CU in children and youth demonstrates moderate to strong heritability (40-78%) with the remaining variance explained by nonshared environmental influences (Viding & McCrory, 2012). • Longitudinal studies find that stability in CU is due to genetic effects while change is due to both genetic and nonshared environmental factors (Blonigen et al., 2006; Fontaine et al., 2010). • Following Willoughby (2011), a CU scale was formed by compositing the following 5 items from the CBCL (for a score of 0 – 10): Figure 1. Twin Intraclass and Cross Correlations • “Doesn’t seem to feel guilty after misbehaving” • There was evidence of substantial genetic change. 58.5% of genetic influence at age 3 was independent of age 2 (i.e., .38/.65 x 100). • “Punishment doesn’t change his/her behavior” • “Seems unresponsive to affection” • “Shows little affect toward people” • The nonshared environment also contributed to change across age, with all nonshared effects at age 3 independent of those at age 2. • “Shows too little fear of getting hurt” • The scale measures CU behaviors as a trait on a continuum, not as a clinical cutoff. • Previous research with young children has found that this CU measure is reliable and demonstrates divergent validity and stability across age. • Nothing is known about the etiology and stability of CU in early childhood. This is surprising given that characteristics critical to the development of CU, such as empathy, are coming online at this age. Results Descriptive Statistics Model-Fitting Results • A bivariate model was fit to the raw data to estimate genetic (A), shared environmental (C) and nonshared environmental (E) contributions CU traits at each age and to the covariance across age. • The full model was a good fit for the data, but several nonsignificant paths could be dropped for a more parsimonious model without a significant detriment in fit (see Figure 2 for reduced model). Table 1. CU means by age and gender Goals of current study were to explore the: Figure 2. Reduced Model • Stability of CU across ages 2 and 3. • Sources of continuity and change in CU across ages 2 and 3. Sample • 314 twin pairs (MZ=145; DZ=169) participating in the Boston University Twin Project. • Zygosity was determined via DNA analyses using DNA obtained from cheek swab samples. • Approximately equal number of males and females (53% Males, 47% Females). Conclusions • The etiology and developmental continuity and change in CU in early childhood mirrors that of older children and adolescents. • However, this is not to say that the same genetic effects operate across toddlerhood and older childhood, as we find evidence of novel genetic and nonshared environmental effects at age 3. • Genetic and environmental influence on CU at both ages. Methods • Continuity in CU was due solely to genetic influences. The genetic correlation was .65 (95% CI: .55-.74), indicating moderately high genetic overlap across age. • To some extent, the genetic influences on CU at age 2 are different from those at age 3. This has implications for molecular genetic research, as the need to consider developmental period is critical. • The range of CU scale scores in our non-clinical sample was 0 – 7. • As seen in Table 1, CU scores significantly declined across age (F=5.15, df=304, p=.02). • The gender effect was nonsignificant, but means were in a direction consistent with the literature with males having higher CU scores than females. • There was no significant interaction between age and gender (F=.74, p=.39). • CU was moderately stable across age (r=.45, p<.0001). Parenting Measures Twin Correlations • Primary caregiver completed the Child Behavior Checklist (CBCL) 1½-5 for each twin at ages 2 and 3. • As seen in Figure 1, for both ages, MZ correlations exceed DZ, suggesting genetic influences. • Variance components (Table 3) from the reduced model revealed little differential heritability across age. Table 3. Variance Components • Nonshared environment also contributes to change in CU across age. Identifying particularly important nonshared environmental factors could help inform interventions. • Longitudinal studies across larger age ranges are necessary to better understand developmental change and continuity in CU.