(PS4-53) Measuring Irritability in Autistic Youth Using the Self-report Affective Reactivity Index

Background: Irritability is a primary clinical concern in autistic youth (Mayes et al., 2019). Historically, parent reports have been prioritized over self-reports when assessing for irritability in autistic populations (Kalvin et al., 2021). Information about reliability of irritability self-report is needed to incorporate autistic perspectives into research and clinical care. This project examined internal consistency of the self-reported Affective Reactivity Index (ARI; Stringaris et al., 2012) in autistic youth, assessed parent-child correlation in ARI scores, and investigated the associations among core characteristics of autism and self-reported irritable mood.

Method: The sample consisted of 100 autistic young people (aged 8-19 years, 25% female, 77% White), who participated in baseline assessments for two clinical trials. Youth and their parents completed the ARI, and parents completed the Social Responsiveness Scale-2 (SRS-2; Constantino and Gruber, 2012). Cronbach’s alpha was used to assess internal consistency. Pearson’s R correlation was calculated between parent-report and self-report ratings, and a paired means t-test was used to assess differences between parent-report and self-report t-scores. In a hierarchical regression with ARI self-report total scores as the dependent variable, covariates (age, sex and IQ) were added in the first step, and t-scores of the SRS-2 Social Communication and Interaction (SCI) and Repetitive Behavior (RRB) subscales were added in the second step.

Results: The self-reported ARI scale had strong internal consistency (𝛼=0.84). Additionally, parent and self-report ARI total scores were significantly correlated (r=.679, p< .001) and not significantly different from each other (t=-1.297, p=.198). Hierarchical regression found that covariates of age, sex, and IQ (added at step 1) accounted for 12.9% of the variance in ARI self-report total score (R2=.129, p=.004), and the SCI and RRB subscales (added at step 2) explained an additional 12.8% of the variance (R2∆ =.128, p< .001). Of note, SRS-2 RRB subscale was significantly associated with self-reported ARI (𝛽=.081, p=.011), accounting for all other model predictors; however, SRS-2 SCI was not significantly related (𝛽=.056, p=.139).

Implications: Our data seems to support the usefulness of ARI self-report in measuring irritable mood in autistic populations, but more comprehensive psychometric analyses are needed to confirm this conclusion. Only RRBs had a significant relationship with self-reports of irritable mood. This finding could reflect that irritability may partially stem from interruption of preferred interests or behaviors. Such findings could inform the development of child-focused interventions for irritability in autistic children.