Catie Nielson and Mr. Mikle South, Psychology

Introduction

Autism spectrum disorders (ASD) are characterized by social communication deficits and restricted repetitive behaviors. Studies have found co­‐morbidity of ASD and social anxiety disorder (SAD), but it is unclear how the anxiety experienced in ASD is socially mediated (White, Oswald, Ollendick and Scahill; 2009). In this study, we measured psychophysiological reactivity during socially stressful (performance­‐evaluated) trials compared to unevaluated trials. We aimed to understand how anxiety in people with ASD is mediated by fear of negative social evaluation, a key factor in SAD. We hypothesized that the ASD group would show elevated stress during evaluation, whether the task was hard or easy, while the control group (CON) would be affected by evaluation only during difficult trials.

Methods

Twenty adults aged 18‐29 diagnosed with ASD were compared to 20 age-­ and IQ-­ matched healthy controls on modified Stroop and Multi‐Sensory Integration (MSIT) tasks developed by Gianaros and colleagues (Sheu, Jennings and Gianaros, 2012). Each task had hard (incongruent) trials and easy (congruent) trials. During incongruent Stroop trials, they had to focus on the color of a word and not the text of the word, and choose from different colored options. During congruent trials, the words for each trial were all the same color, making the decision much easier. MSIT incongruent and congruent trials were similar, but used numbers and positions instead of colors.

Participants were instructed at the beginning of each alternating block, whether they would or would not be evaluated for that block. During evaluative blocks, a research assistant watched the participant from another room through a window and the computer reported their accuracy levels. Each block, either evaluated or unevaluated, consisted of first, a minute of incongruent trials, followed by a minute of congruent trials. These alternating hard and easy trials were separated by 20 seconds of fixation cross and instructions.

We measured stress response with impedence cardiography, heart rate and skin conductance response (SCR). We examined within subjects differences over hard and easy trials, evaluated and unevaluated blocks, and between subjects with ASD and CON groups. Participants also took a battery of surveys, including the Fear of the Negative Evaluation (FNE) and the Social Anxiety Questionnaire (SAQ).

Results

We found that adults with ASD had higher physiological responses, relative to controls, during evaluated blocks regardless of their difficulty, which supported our hypothesis. Heart rate (HR) was elevated in ASD (p < 0.05) in all conditions, compared to CON, and was elevated in evaluated easy (congruent) trials as compared to unevaluated difficult (incongruent) trials (Figure 1). SCR followed the same patterns (p < 0.01), and is not shown here.
We also found parasympathetic activity during recovery periods was reduced in the ASD group. Additionally, response to social evaluation was significantly correlated with scores on the FNE and SAQ in both groups.

Discussion

This is the first study we know of to use measures of impedance cardiography in an autism sample. Findings of increased sympathetic activity during performance evaluation suggest that fear of evaluation in may underlie a great deal of everyday anxiety in people with ASD. This theory is supported by positive correlations between stress response during evaluation and the FNE and SAQ.

The decreased parasympathetic activity during recovery in ASD relative to CON, although not implicated in our hypothesis, is in line with other recent studies we have conducted, and suggests that ASD adults are out‐of‐sync with fear versus safety contexts (South, forthcoming). These cooperative findings suggest that once people with ASD become afraid, they have a difficult time learning when not to be afraid, leading to a perpetual heightened stress level. This heightened stress level was reflected in the initial differences between ASD and CON HR as well as SCR during the tasks. This elevated overall stress level may mitigate any extra stress that performance evaluation brings.

Conclusion

Anxiety in ASD is likely mediated by many factors, but fear of performance evaluation seems to play a strong role. Further investigation is needed, however, to understand how SAD and ASD are linked. In future experiments, we plan to disentangle social evaluation from non­‐social evaluation, in order to better understand this link and the nature of anxiety in ASD. For now, we suggest that interventions for anxiety in ASD should focus on helping individuals to recognize physiological responses to stress and develop specific coping skills for such situations.