Persistent Intolerance of Uncertainty: A Mechanism for Anxiety in Children with ASD?

Background: The inability to tolerate uncertainty is associated with some aspects of anxiety (Birrell, 2008). Intolerance of uncertainty (IUC) in environment or routine may underlie high levels of co-morbid anxiety seen in autism spectrum disorders, including associations between anxiety and repetitive behavior (ASD; Rodgers et al., 2012). To our knowledge, however, there are no known experimental manipulations of IUC in ASD samples.

Objectives: We explored how psychophysiological response would differ according to various combinations of uncertainty and threat (certain safety, certain threat, and uncertain threat) in ASD and matched control (CON) groups. We hypothesized that uncertainty would be particularly aversive for the ASD group and that measures of IUC would predict this response.

Methods: Here we report data for two experiments: the first with a group of children ages 8-11 (26 ASD, 31 CON) and the second with older teens (ages 15-18; 25 ASD and 25 CON). For the younger group, the top card on a computerized deck revealed either a feather that would spare a virtual balloon after a 30-second timer, a pushpin that would pop it, or a question mark that left the outcome in doubt until the timer was finished. We measured skin conductance response (SCR) during each 30-second block and also during a 30-second recovery period of watching a soothing underwater scene. The older teens where shown varied computer contexts where they knew that a startling airpuff to the neck would not occur, occur during a particular cue, or occur in that context but independent of the cue.  The dependent measure was EMG eyeblink response to white-noise startle probes occurring during each condition

Results: Experiment 1 with the children showed similar SCR response in each group during the threat phase, but critically a failure in the ASD group to calm during the recovery period following the uncertain condition. In the ASD group, SCR response to the task was specifically associated with the parent-reported Intolerance of Uncertainty Scale, and not with more general anxiety scales. There were similar strong positive associations with SCR response, IUC scores, and autism symptom severity from the Social Responsiveness Scale. No such associations were seen in the CON group. In Experiment 2 the ASD group failed to habituate as quickly over the course of the task and response in contexts following an uncertain context was exaggerated relative to controls.

Conclusions: These data suggest that while both ASD and CON samples are bothered by uncertainty, the ASD group has specific difficulty in calming down after such situations. This inability to recover from uncertainty may represent one mechanism for anxiety and resistance to change in ASD. We are building a model that considers both cognitive and physiological inputs to anxiety in ASD that may have a different balance in ASD versus anxiety in children without autism symptoms. These studies demonstrate the potential value for designing explicit experimental tests of IUC that can be manipulated to explore, for example, the relative response to reward versus threat or variations in timing of arousal and recovery.