The Role of Somatic Symptoms on Emotion Recognition in Young Children with ASD

Background: Children with autism spectrum disorder (ASD) often display challenges in facial emotion recognition (FER; Lozier, Vanmeter, & Marsh, 2014). While previous literature suggests individuals with ASD demonstrate challenges in detecting negative emotions compared to typically developing (TD) peers (Ashwin, Chapman, Colle, & Baron-Cohen, 2006), recent literature suggests this extends to detection of happiness (Sato et al., 2018). Internalizing symptoms have been shown to be related to FER challenges (Székely, Tiemeier, Jaddoe, Hofman, Verhulst, & Herba, 2014), however, anxiety and depressive symptoms constitute much of the FER literature across youth samples (i.e., Demenescu, Kortekaas, den Boer, & Aleman, 2010). Given that somatic symptoms in childhood predict later anxiety and depressive disorders (Shanahan, Zucker, Copeland, Bondy, Egger, & Costello, 2015), somatic symptoms are a subset of internalizing symptoms that warrant further investigation. We are unaware of any studies that investigate the unique role of somatic symptoms on FER in young children with ASD.

Objectives: Our objective was to develop a greater understanding of factors associated with accurate recognition of happiness among children with ASD. We hypothesized that somatic symptoms would moderate the relation between developmental status and accuracy response time.

Methods: Participants were 152 children (ages 3:0 to 6:11) and their parents. Eighty-four TD children (42% female) and 68 children with ASD (19% female) were examined. Parent ratings from the Behavioral Assessment System for Children - Second Edition (BASC-2; Reynolds & Kamphaus, 2004) were used to measure somatic symptoms. Children completed a dynamic emotion recognition task via laptops in a laboratory setting. Two trials of twenty slides were presented; scores reflect the mean slide number of first accurate recognition. Higher scores indicate slower recognition.

Results: A moderation analysis was conducted using the SPSS 25 macro PROCESS (Hayes, 2013) to evaluate whether the association between developmental status and recognition of happiness was moderated by parent-reported child somatic symptoms. Children’s age was controlled for in the analysis. Results indicated significant main effect of status on emotion recognition (B = 8.490, CI₉₅ [3.035, 13.945], p = .002). The main effect of somatic symptoms on emotion recognition was not significant (B = .038, CI₉₅ [-0.048, 0.120], p = .386). The contribution of the interaction between status and somatic symptoms was significant (B = -0.121, CI₉₅ [-0.233, -0.008], ΔR2 = .02, F(1,147) = 4.512, p = .035. At low-to-moderate levels of somatic symptoms, children with ASD were slower to recognize happiness, while at high levels of somatic symptoms, children with ASD recognized happiness quickly and performed as well as their TD peers.

Conclusions: The current study found children with ASD were slower to recognize happiness compared to their TD peers at low-to-moderate levels of somatic symptoms. However, high levels of somatic symptoms appeared to sensitize children with ASD in their recognition of happiness, with performance similar to TD peers. While studies have shown this sensitizing effect on FER for broadband internalizing symptoms (Rosen & Lerner, 2016), our study provides evidence that somatic symptoms may be a unique contributor to emotion recognition in children with ASD.