Prospective Associations between Anger and Later Internalizing Symptoms in Young Children with ASD.

Background: Early negative emotionality is a risk factor for later difficulties in the general population, including internalizing symptoms and anxiety (Dougherty et al., 2015; Hernandez et al., 2015). Although intensity of anger in response to real-world challenges in toddlers with ASD was found to be greater than that of typically-developing (TD) and DD peers (Macari et al., 2018), little is known about prospective relationships between anger and later internalizing symptoms including anxiety, which affect a large proportion of preschool- and school-age children with ASD (Gadow et al., 2004; Simonoff et al., 2008). The present study aims to address these gaps in knowledge.

Objectives: To examine the association between directly-observed and parent-reported anger in the second year and internalizing symptoms including anxiety at age 3 in children with and without ASD.

Methods: Participants included 94 toddlers (Time1: M_age=22mo, range: 13-30mo; Time2: M_age=39mo, range: 33-47mo): 53 ASD and 41 TD. At Time1, frequency of anger was assessed via parent report (Early Childhood Behavior Questionnaire (ECBQ), Frustration scale) and intensity of anger was assessed directly (Laboratory Temperament Assessment Battery (Lab-TAB; Goldsmith & Rothbart, 1999). Intensity of emotional expression of anger (iEE-Anger) across facial and vocal channels was coded during three Lab-TAB episodes designed to elicit anger. At Time2, parents completed the Early Childhood Inventory (ECI-4; Internalizing composite) and the Preschool Anxiety Scales-Revised (PAS-R; Generalized Anxiety score).

Results: In the ASD group, Time1 iEE-Anger was correlated with the PAS-R Generalized Anxiety score at Time2 (r(30)=.377, p=.040), while iEE-Anger did not predict either Time2 measure in the TD group. Parent-reported ECBQ Frustration at Time1 in the ASD group was positively associated with PAS-R Generalized Anxiety (r(53)=.401, p=.003). In contrast, parent-reported ECBQ Frustration in the TD group predicted ECI-Internalizing composite scores (r(41)=.550, p<.0001) but did not predict anxiety.

Conclusions: This study suggests a significant role for anger in the development of anxiety and internalizing symptoms. Results indicated that expression of anger at age 2 measured both directly and via parent report predicted psychopathology a year later. In the ASD group, both directly-assessed intensity and parent-reported frequency of anger at age 2 were associated with later generalized anxiety symptoms. This contrasted with TD toddlers, for whom parent-reported frequency of anger predicted later internalizing symptoms. Taken together, these preliminary results indicate that frequency and intensity of anger in toddlerhood is related to later anxiety in toddlers with ASD, but for TD children, frequency of anger instead predicts a range of internalizing symptoms. Thus, prospective relationships indicate that pathways to these affective and behavioral vulnerabilities may not be the same in children with TD and ASD. The data highlight the importance of a multi-method approach in understanding the development of internalizing symptoms and anxiety in young children with ASD.