Incongruous Emotions during Fear-Eliciting Tasks in Children with Autism Spectrum Disorder

Background:  Many school-aged children with Autism Spectrum Disorder (ASD) are reported to have unusual fears (e.g., toilets, elevators) and are significantly less likely to be afraid of animals and harm (e.g., knives) than typically-developing (TD) and developmentally-delayed (DD) peers (Mayes et. al., 2013; Evans et al., 2005). Fear of animals, specifically spiders, is one of the most common and biologically-preprogrammed fears in the general population (Davey, 1992). Lacking such fears is both unusual and potentially maladaptive. It is unclear if atypical fear responses are present during the early stages of ASD. 

Objectives:  To examine emotional reactivity in TD, DD and ASD toddlers in response to potentially frightening stimuli.  

Methods:  50 Toddlers (20 ASD, 15 DD, 15 TD) between the ages of 15 and 26 months (M = 20 months) were administered three fear-eliciting tasks from the Laboratory Temperament Assessment Battery, Locomotor Version (LabTAB; Goldsmith & Rothbart, 1999): (1) Spider: a remote-controlled spider crawled toward participants, (2) Masks: an experimenter entered a room wearing masks, and (3) Dinosaur: a mechanical dinosaur emerged from a box. Each task consisted of three to four trials lasting 10 seconds each. Videotaped sessions were coded offline by blind coders for intensity of emotional responses during each trial using four-to-six-point Likert scales. Responses were averaged across trials, and composites were computed for Negative Affect (facial, bodily and vocal distress and escape behaviors) and Joy (facial, bodily and vocal joy). Emotions were then labeled as congruous with the task (e.g., Negative Affect) or incongruous (e.g., Joy). Proportions of congruent and incongruent emotions were computed by dividing each composite score by the sum of negative and joy composite scores during each task. All participants received the Autism Diagnostic Observation Schedule 2- Toddler Module (ADOS2-T: Lord et. al., 2012) to evaluate severity of autism symptoms. 

Results:  A linear mixed model revealed a significant Diagnostic Group*Task interaction (F(4,75.67) = 2.9, p = 0.027) for incongruous emotions. Pairwise comparisons revealed that the ASD group produced significantly more joy and less negative affect during Spider compared to the DD (p = .004) and TD groups (p = .02); however, there were no significant differences between groups for the Masks or Dinosaur tasks. While the majority of the TD (73%) and DD (80%) groups expressed fear during the Spider task, many toddlers with ASD (n = 11, 55%) expressed joy instead. For children with ASD, there was a significant positive correlation between the proportion of joy expressed during Spider and ADOS2-T total scores (r=.405, p=.006). 

Conclusions:  Preliminary findings suggest that toddlers with ASD are more likely to exhibit an incongruous emotion (joy) in response to a novel and potentially frightening stimulus compared to DD and TD peers. Their expression of joy did not generalize to other stimuli (Masks, Dinosaur). Incongruous emotion expression was associated with higher autism symptom severity in the ASD group. The presence of incongruous emotions in response to frightening stimuli in toddlers with ASD may have implications for later idiosyncratic fear expression, comorbidities, safety concerns, and challenging behaviors.