Inhibitory control in early childhood aggression subtypes: Mediation by irritability

• In Event: Developmental Psychopathology Perspectives on Externalizing Behavior Problems Among Young Children

Abstract

Aggression, defined as actions intended to hurt or harm another (Dodge et al., 2006), has correlated but distinct physical and relational forms, which can be expressed proactively to achieve instrumental goals or reactively in response to perceived threat (Card & Little, 2006). Aggression is especially common in early childhood (Ostrov & Keating, 2004), when it is also potentially particularly developmentally impactful (Sroufe, 2013). Inhibitory control (IC) and irritability also show developmental change at this time, despite being relatively stable at later ages (Camacho et al., 2019; Gagne & Goldsmith, 2011). Lower IC is associated with higher levels of trait anger in early childhood (e.g., Gagne & Goldsmith, 2011), and has been implicated in irritability’s longitudinal course (Wakschlag et al., 2018). Both IC and irritability have differential associations with aggression subtypes (Brotman et al., 2017; Murray-Close et al., 2016), but these distinctions have been inconsistent in young children (Poland et al., 2016; Perhamus & Ostrov, 2020). This is potentially due to high co-occurrence of aggression subtypes in early childhood, which may be addressed through the use of severity and directionality scores (Essex et al., 2003). The present study tests irritability as a mediator of IC’s associations with aggression subtypes, hypothesizing poor IC would predict increases in irritability, which in turn would predict increases in aggression overall (severity), and a predominance of reactive over proactive subtypes (directionality). Moderation by gender was also examined.

The sample consists of 300 preschoolers (Mage at T1 = 44.70 months, SD = 4.38 months, n = 132 female), with data collected at three time points over two academic years [spring of year 1 (T1), fall of year 2 (T2), and spring of year 2 (T3)]. IC and irritability were measured using teacher report. Aggression was measured using observer reports. All measures were internally consistent (s > .70). Physical and relational severity scores were created using an average of z-scores across functions ([zreactive + zproactive]/2), whereas directionality scores are the half difference across functions ([zreactive- zproactive]/2; Ellis et al., 2003). Path analyses were conducted in Mplus,Version 8.4, and 5000 bootstrap samples with bias-corrected confidence intervals were used to test indirect effects. Gender moderation was tested through sequential examination of modification indices (Yoon & Millsap, 2007).

Standardized coefficients for the severity model are depicted in Figure 1. Consistent with hypotheses, IC predicted increases in irritability, which in turn predicted increases in both physical and relational severity. IC’s and irritability’s associations with physical severity were moderated by gender. Irritability mediated IC’s relations in the predicted direction for physical severity for girls only [95% CI (-.28, -.002)], and relational severity for both genders [95% CI (-.13, -.01)]. Standardized parameters for directionality models are presented in Figure 2. As in severity models, lower T1 IC predicted increases in irritability, which in turn predicted increasing predominance of reactive over proactive physical and relational aggression. The predicted indirect effect was significant for physical (95% CI [ -.14, -.002]) but not relational [95% CI (-.15, .01)] directionality. No gender moderation emerged for the directionality model.

Authors

• Gretchen R. Perhamus, State University of New York at Buffalo (SUNY Buffalo)
  Presenting Author

• Jamie M. Ostrov, State University of New York at Buffalo (SUNY Buffalo)
  Non-Presenting Author