Danger Lurking? Dissociation of Psychophysiology and Behavior In Response to Provoked Anxiety

Background: In addition to core impairments in social communication, many individuals diagnosed with an autism spectrum disorder (ASD) experience severe and debilitating symptoms of anxiety. Characterization of both shared and distinct neural mechanisms in ASD and anxiety may give insight into the neurodevelopmental course of ASD and improve the specificity of intervention techniques. Our aim was to measure both physiological and behavioral responses to an anxiety-provoking situation where a stranger intrudes on the experiment. In monkey studies and in the wild, such situations tend to be accompanied by increased vigilance but decreased physical activity in order to avoid detection.

Objectives:   We recently demonstrated a disconnect between everyday anxiety and risk-taking in an ASD sample (South, Dana et al., in press). In using a richer, ecologically-valid procedure we hypothesized that our more anxious ASD sample would not be able to contain physical activity in response to the stranger (see Kalin, 2003). We predicted that we would again find evidence for atypical physiological arousal in ASD.

Methods: The Intruder Paradigm for Humans was adopted from studies of emotion regulation in monkeys (Kalin, 2003). Three distinct phases occur while the participant attends to a computerized cognitive task: 1) the attending research assistant leaves the room for a short time, leaving the participant alone; 2) after the RA returns, an ill-dressed male confederate enters the room and begins an argument with the RA until both leave; 3) the confederate intruder re-enters the room but does not engage the participant. Disposable electrodes are used to collect skin conductance response for the duration of the experiment. Participants included 39 children and adolescents diagnosed with an ASD and 34 typically-developing (TYP) individuals matched on age and IQ.

Results: The ASD group was significantly more anxious than the TYP group according to parent- and child-report symptom measures. Baseline skin conductance levels were equivalent across groups. Repeated-measures ANOVA analyses of the skin conductance response (SCR) across all phases of the experiment demonstrated significant a main effect for phase (Isolation < Intrusion > Recovery); and a significant main effect of diagnostic group, due to substantially less psychophysiological responsiveness in the ASD group during all phases of the experiment. In contrast, coding of the video taped experiment demonstrated significantly more body, head, and vocal activity for the ASD group.

Conclusions: This is our second recent finding of significant dissociation of behavioral and physiological states in ASD. This disconnect of internal and external states may reflect atypical connectivity between the amygdala and other key emotion-regulation regions of the brain, and how this confusion is manifested in everyday emotion regulation in ASD. Experiments that can explicitly link results in human ASD samples with existing studies in animal models have strong translational value. We suggest that individual differences in the capacity to monitor internal emotional states (see South, Larson et al., 2010), as found in this adapted animal paradigm, may provide an useful marker for neurobiological and behavioral measurements of heterogeneity in ASD, and offer more specific targets for treatment.